Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Special Report 230 Winds of Change Domestic Air Transport Since Deregulation TRANSPORTATION RESEARCH BOARD NATIONAL RESEARCH COUNCIL
1991 TRANSPORTATION RESEARCH BOARD EXECUTIVE COMMITTEE Chairman C. Michael Walton, Bess Harris Jones Centennial Professor of Natural Resource Policy Studies and Chairman, Department of Civil Engineering, The University of Texas at Austin Vice Chairman William W. Millar, Executive Director, Port Authority of Allegheny County, Pittsburgh, Pennsylvania Executive Director Thomas B. Deen, Transportation Research Board Adm. James B. Busey IV, Administrator, Federal Aviation Administration, U.S. Department of Transportation (cx officio) Gilbert E. Carmichael, Administrator, Federal Railroad Administration, U.S. Department of Transportation (cx officio) Brian W Clymer, Administrator, Urban Mass Transportation Administration, U.S. Department of Transportation (ex officio) - Jerry R. Curry, Administrator, National Highway Traffic Safety Administration, U.S. Department of Transportation (cx officio) Travis P. Dungan, Administrator, Research and Special Programs Administration, U.S. Depart- ment of Transportation (cx officio) Francis B. Francois, Executive Director, American Association of State Highway and Transporta- tion Officials, Washington. D.C. (cx officio) John Gray, President, National Asphalt Pavement Association, Lanham, Maryland (cx officio) Thomas H. Hanna, President and Chief Executive Officer, Motor Vehicle Manufacturers Associa- tion of the United States, Inc.. Detroit, Michigan (cx officio) Lt. Gen. Henry J. Hatch, Chief of Engineers and Commander, U.S. Army Corps of Engineers, Washington, D.C. (cx officio) Thomas D. Larson, Administrator, Federal Highway Administration, U.S. Department of Trans- portation (cx officio) Capt. Warren G. Leback, Administrator and Chairman, Maritime Subsidy Board, Maritime Ad- ministration, U.S. Department of Transportation (cx officio) George H. Way, Jr., Vice President for Research and Test Department. Association of American Railroads, Washington, D.C. (cx officio) Robert J. Aaronson, President, Air Transport Association of America. Washington, D.C. James M. Beggs, Chairman, SPACEHAB, Inc., Washington, D.C. J. Ron Brinson, President and Chief Executive Officer, Board of Commissioners of the Port of New Orleans, Louisiana L. Gary Byrd, Consultant, Alexandria, Virginia A. Ray Chamberlain, Executive Director, Colorado Department of Highways. Denver L. Stanley Crane, Gladwyne, Pennsylvania James C. DeLong, Director of Aviation, Philadelphia International Airport, Pennsylvania Randy Doi, Director, IVHS Systems, Motorola Inc., Northbrook, Illinois S. Earl Dove, President, Earl Dove Company, Dothan, Alabama Louis J. Gambaccini, General Manager, Southeastern Pennsylvania Transportation Authority (SEPTA), Philadelphia (Past Chairman, 1989) Thomas J. Harrelson, Secretary, North Carolina Department of Transportation. Raleigh Kermit H. Justice, Secretary of Transportation, State of Delaware, Dover Lester P. Lamm, President, Highway Users Federation, Washington, D.C. Denman K. McNear, Vice Chairman, Rio Grande Industries. San Francisco, California Adolf D. May, Jr., Professor and ViceChair, University of California Institute of Transportation Studies, Berkeley Wayne Muri, Chief Engineer, Missouri Highway and Transportation Department, Jefferson City (Past Chairman. 1990) Arnold W. Oliver, Engineer-Director, State Departmcntof Highways and Public Transportation, Austin, Texas John H. Riley, Commissioner of Transportation, Minnesota Department of Transportation, St. Paul Della M. Roy, Professor of Materials Science, Pennsylvania State University, University Park Joseph M. Sussman, Director, Center for Transportation Studies, Massachusetts Institute of Tech- nology, Cambridge John R. Tabb, Director, Chief Administrative Officer, Mississippi State Highway Department, Jackson Franklin E. White, Commissioner, New York State Department of Transportation, Albany Julian Wolpert, Henry G. Bryant Professor of Geography, Public Affairs and Urban Planning, Woodrow Wilson School of Public and International Affairs, Princeton University, New Jersey
Special Report 230 Winds of Change Domestic Air Transport Since Deregulation TRANSPORTATION RESEARCH BOARD NATIONAL RESEARCH COUNCIL Washington, D.C. 1991
Transportation Research Board Special Report 230 Subscriber Categories IA planning and administration V aviation Transportation Research Board publications are available by ordering directly from TRB. They may also be obtained on a regular basis through organizational or individual affiliation with TRB; affiliates or library subscribers are eligible for substantial discounts: For further information, write to the Transportation Research Board, National Research Council, 2101 Constitution Avenue, N.W., Washington, D.C. 20418. Printed in the United States of America NOTICE: The project that is the subject of this report was approved by the Governing Board of the National Research Council, whose members are drawn from the councils of the National Academy of Sciences, the National Academy of Engineering, and the Institute of Medicine. The members of the committee responsible for the report were chosen for their special competencies and with regard for appropriate balance. This report has been reviewed by a group other than the authors according to the pro- cedures approved by a Report Review Committee consisting of members of the National Academy of Sciences, the National Academy of Engineering, and the Institute of Medicine. This report was supported by a grant from the Alfred P. Sloan Foundation. Additional support was provided by funds from the National Research Council Fund, a pool of private, discretionary, nonfederal funds that is used to support a program of Academy-initiated studies of national issues in which science and technology figure significantly. The National Research Council Fund consists of contributions from a consortium of private foundations, including Carnegie Corporation of New York, the Charles E. Culpepper Foundation, the William and Flora Hewlett Foundation, the John D. and Catherine T. MacArthur Foun- dation, the Andrew W. Mellon Foundation, the Rockefeller Foundation, and the Alfred P. Sloan Foundation; the Academy Industry Program, which seeks annual contributions from companies concerned, with the health of U.S. science and technology and with public policy issues with technological content; and the National Academy of Sciences and the National Academy of Engineering endowments. The study was also sponsored by the Transportation Research Board. Library of Congress Cataloging-in-Publication Data National Research Council (U.S.). Transportation Research Board. Winds of change : domestic air transport since deregulation. p. cm. - (Special report / Transportation Research Board, National Research Council ; 230) ISBN 0-309-05104-5 I. AirlinesâUnited StatesâDeregulation. 2. Aeronautics, Commercialâ United States. 3. Aeronautics and stateâUnited States. 4. United States. Federal Aviation Agency. 1. Title. 11. Series: Special report (National Research Council (U.S.). Transportation Research Board) ; 230. HE9803.A4N37 1991 387.7'0973âdc2O 91-22773 CIP Cover design: Diane Ross
Committee for the Study of Air Passenger Service and Safety Since Deregulation JOEL L. FLEISHMAN, Chairman, Duke University, Durham, North Carolina GEORGE J. BEAN, Hillsborough County Aviation Authority, Tampa, Florida LANGHORNE M. BOND, Bond & Associates, Washington, D.C. MELVIN A. BRENNER, Melvin A. Brenner Associates, Weston, Connecticut JOHN J. FEARNSIDES, The MITRE Corporation, McLean, Virginia C0RNI5H F. HITCHCOCK, Public Citizen Litigation Group, Washington, D.C. ADIB KANAFANI, University of California, Berkeley TODD R. LA PORTE, University of California, Berkeley MICHAEL E. LEVINE, Yale University, New Haven, Connecticut JAMES W. MAR, Pacific Grove, California JOHN R. MEYER, Harvard University, Cambridge, Massachusetts ROBERT P. NEUSCHEL, Northwestern University, Evanston, Illinois CLINTON V. OSTER, JR., Indiana University, Bloomington JUDITH A. ROGALA, Flagship Express, Inc., Ypsilanti, Michigan KARL M. RUPPENTHAL, Walnut Creek, California Liaison Representatives DANIEL S. KLEIN, Air Transport Association of America, Washington, D.C. DALE E. MCDANIEL, Federal Aviation Administration, U.S. Department of Transportation SIEGBERT PORITZKY, Airport Operators Council International, Washington, D.C. DAVID TRAYNHAM, Subcommittee on Aviation, House Committee on Public Works and Transportation, Washington, D.C. SAMUEL WHITEHORN, Senate Committee on Commerce, Science, and Transportation, Washington, D.C. Transportation Research Board Staff ROBERT E. SKINNER, JR., Director for Special Projects STEPHEN R. G0DwIN, Study Director MARK A. DAYTON, Senior Program Officer THOMAS R. MENZIES, Research Associate NANCY A. ACKERMAN, Director of Publications LUANNE CRAYTON, Assistant Editor Consultants HERBERT N. JASPER JOHN S. STRONG
Preface A decade after passage of the Airline Deregulation Act of 1978 the Ex- ecutive Committee of the Transportation Research Board (TRB) concluded that a review was needed of the experience with airline deregulation. At that time, air passenger service and safety issues were being hotly debated in the media and in Congress, and the issues most in debate concerned whether airlines that provided most of the service to and from hub airports were able to exercise market power in charging for airline tickets and whether the increased concentration of the industry would reduce com- petition. Late in 1988 the Governing Board of the National Research Council approved a two-year study and helped support it with funds from the National Research Council Fund, a pool of foundation funds under the auspices of the National Research Council. The Sloan Foundation also supported the project, as did TRB. The committee appointed by the National Research Council included experts in aviation, economics, safety, airline and airport operation, and public policy. Some of the members have extensive experience in airline, airport, and federal agency management, others are scholars who have studied the economic and safety effects of airline deregulation or related public policy issues. To enhance its expertise, the committee invited the participation of liaison representatives from the Air Transport Asociation of America, the Federal Aviation Administration (FAA), the Airport op- erators Council International, and professional staff members of congres- sional committees. The Air Transport Association of America and the Federal Aviation Administration liaisons in particular met regularly with the committee throughout the study and commented on report drafts. The committee of experts assembled for this study was given a broad charge and asked to consider the following issues: Whether the current level of air passenger service is adequate in comparison with past and likely future trends,
Whether there are certain factors, working independently or inter- actively, that affect service, Whether safety has diminished since deregulation, and Whether policy changes by airlines, local agencies that own and operate airports, or the federal government are necessary to improve the quality of air passenger service. The committee defined its charge as an examination of the key issues affecting commercial aviation since deregulation. Using deregulation as a point in time for comparison, the first part of the study focuses on private sector issues: air carrier management and financing, effects on consumers (in terms of fares and flight frequency), and the nature of airline com- petition. In examining effects on consumers, the committee chose to em- phasize the trends in pricing and in flight offerings. In part because of concern about inadequate data and in part because of the limitations of time and resources, the committee chose not to examine service quality issues such as lost baggage, over-booked flights, or other customer service issues. In the second part of the study, the committee's report focuses mostly on public sector issues, namely safety, airport and airway capacity, and the ability of the federal government to discharge its responsibilities in these areas. Although the report focuses first on private and then on public sector issues, the two are obviously intertwined. For example, the chapter on safety necessarily addresses both private sector performance and public sector responsibilities. It is not meant to be a comprehensive assessment of aviation safety since deregulation but an examination of whether changes in safety are related to changes in air carrier operations since deregulation and whether the public sector's oversight of the private sector in this regard is adequate. Similarly, the chapter on capacity is not meant to be a comprehensive assessment of all capacity-enhancing technologies and practices but an assessment of how capacity might have been affected by deregulation and how public policies can encourage the private sector to use capacity in ways that will enhance competition and the efficient use of scarce resources. The safety and capacity issues raised in their respective chapters led the committee to examine how the Federal Aviation Administration has discharged its safety and capacity responsibilities since deregulation. In conducting this study, the committee drew on an extensive body of research conducted by many scholars and organizations on certain facets of commercial aviation since deregulation and supplemented the findings vi
from these studies with empirical analyses of its own. Even with the extensive literature, however, some of the most controversial issues raised during the study could not be completely answered on the basis of empirical analyses alone. The available data and methods employed, although con- sistently pointing in the same direction, nonetheless left room for differ- ences of opinion. The findings and recommendations in this report reflect both the findings in empirical research and the collective judgment and insight of the committee. Members did not reach complete agreement on some issues, and the differing views are reflected in the report. One member, Melvin A. Brenner, dissented from parts of the report. His dissenting statement appears at the end of the report. In taking a retrospective view on the economic and safety issues, the committee confined most of its analysis to the domestic commercial avia- tion system, primarily that of the 48 contiguous states, because that is where the effects of deregulation were most likely to be manifested and because both Alaska and Hawaii were treated differently in various pro- visions of the Airline Deregulation Act of 1978. During deregulation, however, some domestic carriers earned substantial revenues in interna- tional service, which grew more rapidly than domestic travel in recent years and tended to be more profitable because many of those markets are still regulated. Given the growing importance of international trade and competition, such issues are likely to play a more prominent role in commercial aviation in the future in ways that were not considered during the course of this study. For example, issues such as the appropriate level of foreign investment in U.S. airlines were not examined. Most of the studies of deregulation, as well as the analyses conducted by the committee, were retrospective or were comparisons based on then- current data. Meanwhile the commercial aviation industry has been chang- ing constantly. Not only has the international dimension become more important, but the industry has become increasingly concentrated. Looking backward from 1984, when many of the early studies were completed, it could be seen that many new airlines had entered the industry, competition was intense, and fares were falling. The optimism of these early evalu- ations must be tempered by subsequent events. Looking backward from 1991, it can be seen that the industry has become more concentrated with the largest airlines holding a larger share of passenger traffic than they had at the outset of deregulation. This more concentrated industry has also been very competitive, but continued to concentrate even as the committee completed its report. Indeed, between the last two meetings of the study committee, Pan American and Continental filed for protection vu
from creditors while reorganizing under Chapter 11 bankruptcy proceed- ings, and Eastern Airlines, after being in bankruptcy for 22 months, finally ceased operation. In its examination of the governmental role since deregulation, the committee focused most of its attention on the role of the FAA in providing ainvay and airport capacity and enforcing federal safety regulations. Whereas several evaluations of the FAA were conducted by others and resulted in similar diagnoses, the recommendations for institutional reform in these studies were not accompanied by detailed analyses of how these changes, if adopted, would affect the management, performance, and morale of the FAA work forces. This lack of a firm analytical base was a matter of debate within the committee. Several members favored specific institu- tional reforms because they were convinced that there would be benefits, whereas others were reluctant to support the reforms without a better empirical basis for predicting their effectiveness. The committee's rec- ommendation for a study of this issue is designed in part to provide a stronger analytical underpinning for choosing among options for reform. Although the committee found many benefits as a result of deregulation, can it assume that what is past is prologue? Given the rapid change in the industry, accompanied by the fuel price shocks caused by the 1991 Persian Gulf War, predictions must be tempered with caution. There are reasons for optimism: the successful deregulated airlines have better management, are more effective competitors, and are likely to hold their own in future global competition. The weakened financial posture of some airlines, exacerbated by the recent fuel price shocks and recession, however, is reason for concern. As the number of carriers continues to diminish, concern about reduced competition increases. Prudent government action was never in greater need; implementation of the committee's recom- mendations to improve competition will help. Over the longer term, the public responsibilities for providing safety and capacity would be enhanced by the committee's recommendation for a thorough study to determine how best to organize the FAA to discharge these responsibilities. The final report of the committee was reviewed by an independent group of reviewers in accordance with National Research Council report review procedures. VIII
Acknowledgments In order to obtain additional views on controversial issues examined during the course of the study, the committee invited presentations by several individuals, whose voluntary contributions to the study committee are gratefully acknowledged. John A. Burt, Executive Director of Acquisi- tions of the Federal Aviation Administration (FAA), briefed the committee on the FAA's plans for improving the management of technology acqui- sition. The committee was briefed on past and current policies that affect FAA personnel by both FAA managers in the Washington headquarters and first-line supervisors in the field. FAA managers Wanda Reyna, Lionel Driscoll, Duane Mason, and Burliegh Stokes provided the central-office view of staffing needs, problems, and agency plans. Mike Connor and William Pearman, first-line supervisors in air traffic control (ATC), dis- cussed management and morale issues that affect controllers and their managers. John Wagner and Sheldon La Mountain, first-line supervisors of electronics technicians, discussed the shortages of skilled personnel responsible for maintaining ATC technologies. The effect of airport terminal capacity constraints on the ease of airline entry into city-pair markets was debated before the committee by airline executives and airport managers. R. James Thorne, Vice President of Property of Northwest Airlines, and Jerry A. DePoy, Vice President of USAir, ably presented the views of major air carriers. Richard H. Judy, former Director of Miami International Airport, and James C. DeLong, Director of Aviation of Philadelphia International Airport, countered with the views of airport managers trying to provide access to all carriers that wish to serve a community. The committee also invited the perspectives of travel agents on computer reservation systems and the role of volume incentives on travel-agent bookings. David S. Paresky, President, Chairman, and CEO of Thomas Cook Travel, and Linda Paresky reviewed their extensive experience with Lx
air travel since deregulation and provided insight into the probable effects of policy changes on travel agents and consumer information. Many individuals assisted individual committee members and the proj- ect staff. Among them are Earl F. Weener, Barry C. Latter, and Thomas M. Barry of Boeing Commercial Airplanes; Leonard Wojick (at that time with the Flight Safety Foundation); Jonathon Howe and Pete West of the National Business Aircraft Association; Marty Langelan of the U.S. De- partment of Transportation; Alan Breitler, Nick Komons, Gene Mercer, and Canton Wine of the FAA; Dayl Cohen of the Massachusetts Port Authority; Richard Streit of the Research and Special Projects Adminis- tration; Francis Mulvey, James Noel, and Kim Coffman of the General Accounting Office; Richard Fletcher and Earl Doolin of Data Base Prod- ucts; and Joe Meier of Yield Data Services. The contributions of these individuals expanded the understanding of the study committee and staff. However, the findings, conclusions, and recommendations expressed in this report, along with any errors or omis- sions, are solely those of the study committee. The study received overall supervision from Robert E. Skinner, Jr., TRB Director of Special Projects. Stephen R. Godwin served as study director and, under the guidance of the committee, prepared the Executive Summary and Chapters 1 and 4 through 8. Committee member James Mar assisted in the preparation of Chapter 5 by writing the section on aging aircraft. Mark R. Dayton performed the original analyses of trends in air carrier finances and profitability and drafted the financial performance section of Chapter 2. John S. Strong, with the College of William and Mary, assisted Dayton and prepared Appendix A. Thomas R. Menzies performed the extensive, and in some cases original, empirical analyses on air fares and air service contained in Chapter 3 and drafted the chapter. Consultant Herbert N. Jasper wrote a commissioned paper on FAA or- ganizational alternatives and provided assistance in the preparation of Chapter 7. Special appreciation is expressed to Frances E. Holland and Marguerite Schneider for assisting in meetings logistics, coordinating the voluminous mailings and correspondence required over the course of the study, and providing word processing support for the numerous drafts.
Contents ExecutiveSummary .................................................. 1 PART I INTRODUCTION Deregulation of Commercial Aviation ..........................21 Brief History of Regulation, 22 Precedent and Rationale for Regulating Airlines, 23 The CAB Regulatory Approach, 26 Impetus for Deregulation, 28 Administrative and Legislative Reform, 29 Competition and Traffic Growth, 30 Effects of Deregulation, 31 Organization of Report, 38 Changes in Airline Operations and Financial Performance, 38 Passenger Fares and Airline Service, 39 Barriers to Competition, 40 Commercial Aviation Safety, 41 Airport and Airway Capacity Limits, 41 Constraints on the Performance of the FAA, 42 Summary, 43 PART II EFFECTS AND RESPONSES IN THE PRIVATE DOMAIN Changes in Airline Operations and Financial Performance .. 49 Air Carrier Management and Operations, 49 Route Strategies, 50 Asset Management, 52 Marketing Innovations, 53 Labor Costs and Relations, 55 Productivity, 56 Mergers and Acquisitions, 56 Industry Financial Performance, 57 Analysis Methodology, 59 Operating Results, 64 Capital Structure, 67 Summary, 79
3 Passenger Fares and Airline Service ............................85 Chapter Organization, 86 Data Sources, 86 General Trends, 87 Passenger Traffic, 87 Average Fares, 89 Fare Variety, 92 Fares by Market Distance and Density, 97 Effects of Competition on Fares and Service, 102 Systemwide Benefits, 103 Consequences of Hub Dominance, 107 Rural and Small Communities, 118 Changes in Service, 121 Changes in Fares, 125 Summary, 129 General Trends, 129 Effects of Competition on Fares and Service, 129 Rural and Small Communities, 130 4 Barriers to Competition ........................................ 134 Financial Risks, 135 Airport Capacity Constraints, 136 Terminal and Gate Capacity, 136 Contractual Agreements, 138 Dominated Hubs, 141 Slot-Controlled Airports, 142 Environmental Constraints, 143 Airport Restrictions, 144 Aircraft Availability, 144 Airline Marketing Strategies, 145 CRSs, 146 Travel-Agent Incentives, 152 Frequent Flier Programs, 154 Code Sharing, 155 Mergers and Acquisitions, 157 Summary, 161 PART III EFFECTS AND RESPONSES IN THE PUBLIC DOMAIN Commercial Aviation Safety .................................... 169 Safety Trends, 170 Accidents and Fatalities During Deregulation, 172 Accident and Fatality Rates, 173 Deregulation and Accidents, 175
Methodological Issues, 175 Carrier Finances and Accidents, 177 Maintenance Expenditures and Safety, 177 New Entrants Versus Established Carriers, 178 Pilot Experience, 179 Commuter Substitution, 181 FAA Performance and Safety, 182 Aging Aircraft, 185 Safety Indicators, 191 Near-Midair Collisions, 192 Pilot Performance, 192 Flight Data Recorders for Crew Monitoring, 193 Indicators of Controller Performance, 195 Aircraft Performance and Maintenance, 196 Summary, 197 Airport and Airway Capacity Limits ..........................202 Past and Current Demand for Airports, 203 Airport Use Since Deregulation, 203 Delay as a Measure of Capacity, 210 Projected Increases in Demand, 215 Limited Supply Expansion, 218 Few New Airports Likely, 220 Few New Runways at Existing Airports, 220 Better Use of Existing Capacity, 221 Modifications of ATC Procedures, 221 Replacement of Outmoded ATC Technology, 222 Systems Approach to ATC, 224 Marketplace Strategies, 225 Summary, 233 7 Constraints on the Performance of the Federal Aviation Administration ................................................... 238 Financing, 240 Appropriations, 240 Revenue Sources, 243 Personnel, 244 ATC, 244 Safety Inspections, 253 Shortages in Personnel with Special Skills, 254 Changing the FAA Culture, 255 Summary, 256 Acquisition of Advanced Technology, 257 Inadequate Planning and Management, 257 Complexity of High-Technology Acquisition, 258
Management, 260 Micromanagement, 261 Stable Leadership, 262 Summary, 263 PART IV SUMMARY 8 Conclusions and Recommendations ............................ 271 Changes in Airline Operations and Financial Performance, 272 Management Innovation, 272 Financial Performance, 273 Passenger Fares and Airline Service, 276 Benefits of Competition, 278 Hub-and-Spoke Networks and Competition, 279 Rural and Small Communities, 280 Barriers to Competition, 282 Risk of Entry, 282 Airport Terminal Capacity Limits, 283 Slot-Controlled Airports, 284 Airline Marketing Strategies, 285 Mergers and Acquisitions, 290 Recommendations to Enhance Competition, 290 Commercial Aviation Safety, 292 Airport and Airway Capacity, 294 Airport Congestion, 295 Airway Capacity, 297 Recommendations for Expansion and Better Use of Airport and Airway Capacity, 299 Constraints on the Performance of the FAA, 299 Institutional Constraints, 301 Unanswered Questions for the Future, 304 Options for Reforming the FAA, 305 APPENDIXES A Measuring Financial Risk in the Airline Industry ...........308 John S. Strong B Organizational Options for the Federal Aviation Administration .................................................. 313 Herbert N. Jasper C Airport Identification Codes ...................................384 DDissenting Statement ........................................... 388 Melvin A. Brenner Study Committee Biographical Information ......................393
Executive Summary Commercial aviation was one of the first industries affected by the controversial regulatory reforms that began in the 1970s. Beginning in 1975, administrative reforms of the Civil Aeronautics Board (CAB) gave carriers greater freedom in discounting prices and serving new mar- kets. The Airline Deregulation Act of 1978 removed restrictions on entry, pricing, and routes. Still unresolved in policy and practice, however, is the question of the appropriate role of government. How should it attempt to ensure fair competition and adequate safety, protect consumers, and provide airport and airway capacity? In the interest of informing the public debate about deregulation, the Executive Committee of the Transportation Research Board convened a committee of 15 experts to review air passenger service and safety since deregulation. The committee, under the chairmanship of Joel L. Fleishman, Senior Vice President of Duke University, included individuals expert in airline and airport operations and management, economics, safety, airway and airport capacity, and public policy. The findings of the committee, along with its recommendations, are summarized in the following pages. EMERGING INDUSTRY STRUCTURE The number of new carriers entering the industry and increased competition during the early years of deregulation seemed to confirm the expectations of some that the deregulated airline industry would approximate the clas- sical model of multiple-firm competition. In recent years, following a spate of mergers in the mid-1980s, the industry has become dominated increasingly by a few large firms, whose share of revenue passenger miles approximates the share held by the largest carriers before deregulation.
2 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION Past and Current Benefits The emerging industry suggests evolution toward oligopoly, but also dis- plays a substantial amount of competition. At the outset of deregulation only about 20 percent of city-pair markets had three or more competitors; this share has grown to 40 percent. City-pair markets with three or more competitors serve about two-thirds of passenger trips. To date, this com- petition has generated many benefits for consumers. The freedom given to managers by deregulation to realign their routes and costs has resulted in a more efficient industry (Chapters 1 and 2). Average fares have in- creased more slowly than costs, although most bargains occur in markets in which three or more carriers compete for service, especially when one of those competitors is a low-cost, new-entrant airline (Chapter 3). The development of hub-and-spoke networks, spurred by deregulation, has expanded the frequency and availability of service in most parts of the country (Chapter 3). In addition, rural areas, which used to receive sub- sidized jet-carrier service, continue to receive air service, though now via smaller turboprop aircraft. Substantial improvements in the frequency and timeliness of service have occurred in most rural areas (Chapter 3). Even while the industry has evolved and service has expanded, accident and fatal accident rates have been lower than during the period before deregu- lation (Chapter 5). Additional Forces Toward Concentration Several trends suggest that the industry will continue to concentrate, which at some point might threaten the benefits achieved during deregulation. Additional industry concentration may occur because of the nature of current competition among the largest carriers offering nationwide service. Being an effective major competitor requires, among other things, pro- viding service to a broad network of cities. Only a limited number of carriers will be able to achieve market presence in a sufficient number of large cities to build a national network (Chapter 4). In addition to the trend toward concentration in the national market, few hub airports are likely to support more than one hubbing carrier because of the substantial risk and cost associated with competing with a carrier at its hub, lack of gates and terminal space at many hub airports, and insufficient local traffic support. These conditions create an almost inevitable drift toward single-carrier dominance at many hubs, which
Executive Summary3 further reduces competition in regional markets and gives the dominating carriers an opportunity to exercise market power (Chapters 3 and 4). Despite the concentration that has occurred, competition remains keen, so much so that there is little evidence of monopoly profits. In contrast, the lackluster financial performance of the industry as a whole raises concern that too few carriers will be able to generate adequate revenues to reinvest for long-term growth (Chapter 2). This concern is heightened by the confluence of three major influences on carrier balance sheets. Just as the major carriers are entering a massive reequipment cycle, the down- turn in the business cycle, and the fuel price shocks and declining traffic caused by the Persian Gulf War threaten to weaken further the industry's financial posture. These problems, which are compounded for some car- riers by the large increase in debt they took on during the financial re- structuring in the 1980s, may lead to the demise of additional major carriers. It is impossible to predict with certainty how the pending industry shakeout will affect competition and balance sheets. The loss of one or two carriers, for example, may reduce excess capacity in the industry and improve cash flow and profits for the remaining carriers. The loss of too many carriers, however, could reduce the amount of competition to below a level necessary to discipline pricing. Although the existence of three carriers in an individual market may ensure adequate competition to dis- cipline pricing, the presence of only three carriers providing nationwide service would probably not be adequate. If as many as five or six major carriers survive the pending shakeout, the committee believes that pricing will remain competitive. The number of carriers is important to compe- tition, but no less important is the ability of the remaining carriers, both large and small carriers, to compete with one another in the maximum number of markets. Barriers to Entry and Competition Preserving or creating sufficient conditions for entry and competition to ensure that the evolving industry remains competitive is fundamental to preserving the benefits of deregulation. At the outset of deregulation, easy entry into city-pair markets was expected to discipline pricing because it was thought that any carrier that charged monopoly fares would be quickly contested by other carriers. During that period the shortage in airport terminal capacity, partly caused by the rapid expansion of hubs, reduced
4 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION the ease of entry in many markets, and contestability was further dimin- ished by innovative marketing strategies, including computer reservation systems (CRSs), that were developed and applied by some carriers to establish and build on the competitive advantage that they have at their hubs (Chapter 4). Risk of Entry A competitor that wishes to mount a significant challenge to another carrier at its hub is faced with a considerable financial outlay. The cost of pro- viding a competitive level of service at a hub is substantial; it requires outlays for advertising, personnel, and aircraft operations during start-up when the competitor attempts to win service away from the hub carrier. The risk of being unable to recover these outlays is the largest single deterrent to entry at hub airports. Indeed, instead of engaging in head-to- head competition at a hub, most carriers have opted for creating hubs at formerly underused airports, from which they compete for hub traffic from spoke cities. As the record shows, this strategy is also risky. Airport Capacity At many large airports that serve major markets it is also difficult and expensive to obtain gates and related terminal space. With little excess capacity at strategic airport terminals in the short run, the carriers operating hubs gained control over existing gates and related facilities as they ex- panded their hubs. Most major carriers, however, have been able to obtain gates at airports for the markets they wish to serve, but this has required financing new facilities in many cases. Other competitors, including small, low-cost carriers that wish to gain access to many large airports in major markets, find few gates available at times of day that would allow them to provide competitive service. Those that are available often come at a cost that threatens these competitors' low-price marketing strategy. In short, entry into many city-pair markets, although possible, is both expensive and risky. Although domination by one or occasionally two carriers appears likely at most hub airports, some degree of competition can be preserved at dominated hubs if sufficient access to gates remains available to allow entry by competitors wishing to establish a "spoke" operation at a competitor's hub.
Executive Summary 5 Marketing Advantages Through CRSs, incentives to travel agents, frequent flier programs, and other marketing practices, major carriers can influence consumer choices in ways that smaller, new-entrant airlines and larger airlines without these advantages cannot. CRSs, which are used to book more than 90 percent of domestic air travel, combined with volume incentives paid to travel agents, help provide the owners of these systems with extra ticket sales. CRSs enable carriers to earn revenues from their rivals by charging book- ing fees that are as much as twice the cost of providing the service in most cases and give them an edge in monitoring travel agents and in managing discount programs. Frequent flier programs, which are popular with business travelers, have helped build considerable brand name loyalty and make it difficult for smaller carriers to attract business travelers on the basis of low fares. Because they compose roughly one-half of the total travel market, business travelers are key to the success of air carriers. They are important to any airline's revenue because they frequently cannot take advantage of the deepest discount fares, even when the organization for which they travel is astute enough to negotiate special discounts directly with the airline. Code-sharing arrangements between major airlines and small regional and commuter carriers, in which small carriers share the major carrier's computer code in CRSs, signal to travel agents the existence of integrated fares and schedules between carriers serving small cities and the hubs of major airlines. Such relationships between large and small carriers predate code sharing and were also listed in CRSs before code sharing. The preference that CRSs give to code-sharing relationships, however, gives them a marketing advantage over other carrier relationships that do not include shared codes. Consumers receive some benefits as an indirect consequence of code sharing because the larger carrier generally requires a minimum level of service and safety as a condition of permitting the commuter carrier to use its code. Commuter carriers receive the benefits of traffic "directed" to them by major airlines, and these vertical links help support competitive air service from small cities via alternate hubs. There is one disadvantage, however: code sharing with a major carrier in commuter markets large enough to support more than one commuter can become a necessary condition for entry in those markets, which reduces the potential for competition. Being able to exploit all the competitive advantages listed above is not necessary for survival in the deregulated airline industry. Some new en- trants, for example, that do not have a CRS have survived and even
6 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION prospered in some years by finding niche markets. One former intrastate carrier has done well by providing low-fare, short-haul, point-to-point service. The existence of these barriers, however, appears to explain partly how the airlines that existed before deregulation, with their higher cost structures, managed to compete successfully with low-cost new entrants. (Some of the "holdover" carriers also lowered their costs by negotiating more productive work rules, lower wages, or by otherwise rationalizing their operations.) Recommendations To Improve Competition Proposals that seek to reduce impediments to competition must include a consideration of whether the recommended change will provide more benefits than costs. In the judgment of the committee, implementation of the three recommendations discussed next will reduce the tendency of the evolving industry structure to restrict competition without imposing costs in excess of benefits. Antitrust Policy A more forthright federal antitrust policy is required than has been prac- ticed during much of the 1980s. The U.S. Department of Justice (DOJ) has been more active on antitrust issues since authority for this task shifted from the U.S. Department of Transportation (DOT) to the DOJ in 1989. The study committee supports the more activist role adopted by the DOJ, but recognizes that further mergers or asset acquisitions leading to greater industry concentration are likely. Given the prospects for further concen- tration, the committee recommends that the DOJ oppose mergers or acquisitions in which the carriers offer substantial parallel service in city- pair markets or share a hub airport. However, the DOJ should not necessarily oppose mergers or asset acquisitions of carriers with complementary or end-to-end routes. Such mergers or asset acquisitions often may not harm competition. Given the importance of having at least three effective competitors in city-pair markets involving a connection at a hub, the maintenance of this level of choice for consumers should be used as a test for the adequacy of competition in "over hub" traffic when merger and acquisition proposals are considered, including acquisitions of individual assets, such as gates.
Executive Summary 7 CRSs CRSs offer considerable competitive advantages to the carriers that own them. These advantages would be reduced by improving the competition between the existing CRSs. Although carriers compete vigorously with one another in the selling of these systems, once an agent is under a lease, switching systems is not practical for the duration of the lease. Lease clauses prohibit the ue of the leased equipment to access other CRSs, and few agents can afford to operate more than one CRS. The committee recommends that travel agents be allowed to use their own equipment or desk equipment leased from the host carrier to access multiple CRSs. This would require a DOT regulation to prohibit contract or lease terms that restrict the ability of travel agents to use equipment that is connected to one CRS to switch freely among CRSs, along with continuing the prohi- bition against display and function bias and extending it beyond CRS owners to any software used in the interface allowing multiple access. Implementation of this recommendation would improve service to the traveling public. The advantages that CRS-owning carriers have over their competitors would be reduced by this "multiaccess" provision because agents could readily access alternative systems to compare simultaneously in an integrated display the best offerings for their clients and to determine the latest fares and seat availability (assuming that the existing prohibition on bias would be extended to others as it became possible to access CRSs through other software.) Airlines that do not own a CRS could encourage travel agents to book their flights through CRSs that charge them lower booking fees. The system would be most competitive if the booking fees were charged to travel agents, who could switch between systems, rather than to airlines that for competitive reasons cannot afford to opt out of participating in individual CRSs. Consumer Information Travel agents can influence consumer choices, and consumers are typically unaware of the financial relationships that exist between many travel agents and airlines. At present, when an agent tells a customer that a certain flight most closely matches the customer's preferences, the customer has no way of knowing whether the agent's recommendation is influenced by a desire to earn an extra commission. The committee recommends improving consumer information by re- quiring agents to disclose the incentive commissions (commission over-
8 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION rides) that they receive from carriers. This could be accomplished by a DOT regulation that has the effect of making consumers aware of the existence of override programs and the identity of the carriers favored by the agent. SAFETY Commercial aviation safety has improved steadily throughout the post- World War II era. The fatal accident rate has declined since deregulation, and indeed, during two years since 1978 (1980 and 1984) no fatal accidents occurred in regularly scheduled passenger service provided by large airlines. In addition, in 1986 only one person was killed in accidents in- volving these carriers. Thus, the three safest years in the history of scheduled passenger service provided by jet carriers have occurred since deregulation. The changes ushered into the industry with deregulation could con- ceivably have increased risk to public safety for three reasons: (a) de- creased emphasis on maintenance caused by increased pressure on costs; (b) increased reliance on less-experienced flight crews, resulting from increased entry and the booming demand for air travel; and (c) inexperience of new-entrant airlines. Studies in which various dimensions of these concerns have been examined, however, have not established any statis- tically significant relationship between carrier actions based on these rea- sons and accident and fatal accident rates. Concern has been expressed by many that the economic pressure on carriers brought about by deregulation may have increased risk, even if these risks are not yet manifested in increased actual accidents. Mea- sures of risk have therefore been proposed, both to examine the record since deregulation and to measure current trends. Reporting biases in nonaccident safety indicators such as those that track near midair col- lisions, however, prevent them from being used to evaluate changes in risk during deregulation. Congress has urged the Federal Aviation Ad- ministration (FAA) to develop better safety indicators, which is likely to be a difficult task. For communities to which deregulation brought a change in air service from jet to turboprop service, risk may well have increased during the early years of deregulation. The withdrawal of jet service from small communities had been occurring for many years before 1978, but dereg- ulation probably accelerated that trend. Fortunately, the difference in risk between jet and turboprop service has narrowed considerably over time,
Executive Summary 9 partly because of more stringent safety regulations applied to those carriers in 1978. By 1990 the fatal accident rate of commuter carriers and jet carriers offering regularly scheduled service was roughly equivalent. The aging of the aircraft fleet is only indirectly related to deregulation. Fleet aging was already under way before the late 1970s, but the reliance on older aircraft by new entrants and the boom in air travel during the 1980s may have created more demand for aging aircraft than otherwise would have been the case. The stringent maintenance and inspection sched- ules for aging aircraft suggested by an industry task force and incorporated by the FAA into regulation, however, appear to reduce concerns about the risk of catastrophic accidents caused by the failure of aged airframes. The more rigorous maintenance schedules for aging aircraft will increase the cost to air carriers and the demands on the FAA's safety inspection work force, which has not kept pace with industry growth over the last decade. For example, the FAA's special in-depth inspection programs at several carriers during the late 1980s revealed that its routine inspection program, which relies on inspections of paperwork and airline self- regulation to ensure safety instead of hands-on inspection of aircraft and maintenance practices, is not sufficient to ensure carrier compliance with maintenance regulations. Although the safety record in commercial aviation has improved since deregulation, continued vigorous efforts are required in carrier oversight and safety regulation to ensure that safety is not compromised. The com- mittee recommends thai FAA staffing and inspection procedures be made sufficient to ensure adequate maintenance for safety. In addition, the FAA should continue to reduce dfferences in equipment standards and oper- ating requirements between national and regional (commuter) carriers. AIRPORT AND AIRWAY CAPACITY The accelerated growth in demand for air travel during deregulation has begun to conflict with capacity limits at some airports and the airspace around them (Chapter 6). Deregulation gave airline managers more flex- ibility in employing assets to serve consumer demand, and managers have used this flexibility to provide capacity at formerly underused airports. The public sector, however, has lagged behind the demand for air travel, both in the case of specific airports and in the provision of air traffic control (ATC). The worst shortages in capacity occur in relatively few places, but they occur at major airports that serve a large share of total demand. The prospects for several new airports around major metropolitan
10 WINOS OP CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION areas to help meet this demand and to reduce congestion do not appear promising in the near term because of competing demands for land use and the opposition to increased noise. Matching Demand and Supply for Airports with Market Mechanisms Because much of the congestion at major airports occurs at peak periods of the day and week, peak-period pricing of runways could be used to allocate scarce capacity more efficiently. To be effective, peak prices for runway landing fees would have to increase substantially over their current level. The actual fee imposed would depend on the value placed on the landing by the users with the least value for landing during the peak, because these users would be the first to shift to avoid paying the cost. Higher fees for peak-hour use of airport runways would shift some traffic to the off-peak, thereby reducing some congestion and delay; would further encourage the development of underused airports; and would promote reliance on larger aircraft to increase passenger throughput without con- tributing to airspace congestion. Congestion pricing would also provide airport operators with a new source of additional revenues, which could be quite substantial. A potential problem with runway pricing is the extent to which airports, perhaps under pressure from community groups opposed to airplane noise, would use congestion pricing to restrict aircraft operations or the extent to which airports would avoid reinvesting the revenues earned to expand capacity. One strategy for limiting this kind of problem would be to require that funds collected through runway pricing be transferred to a central, off-budget authority of some kind. This authority would return the rev- enues to airports that use runway pricing for any expenditures designed to enhance capacity. Expanding Airway Capacity through Technology and Research The FAA's most prominent response to the increased demand since dereg- ulation is its plan to replace and upgrade the ATC systeili. The National Airspace System Plan (NAS Plan) was originally promulgated in 1981 as a 10-year, $12-billion program to replace outmoded technologyâliterally to replace vacuum-tube electronics and radars, some of which were surplus
Executive Summary I I World War II technology, with solid-state systems. During the past decade the NAS Plan has evolved into an ongoing multiyear capital improvement program currently estimated to cost $27 billion; these new technologies, which incorporate the NAS Plan, are not scheduled to be in place for another 15 years or more. The FAA has been criticized for not planning its capital needs well and for not managing the program effectively. However, the increase in total cost and the longer implementation schedule for key elements of the plan reflect the FAA's growing awareness that the original plan would not provide sufficient additional operational capacity to meet growing demand, nor would it take advantage of technology developments since the plan was envisioned. Given the complexity of this massive program for the acquisition of advanced technology, the uncertainties in federal funding, the uncertainties of some of the unresolved technical issues, and the prog- ress of the NAS Plan, further delays and cost increases would not be surprising. With the evolution of the FAA's capital improvement plan during the last decade and with key new technologies not yet in place, the airspace around some major airports has become increasingly saturated, causing ripple effects throughout much of the ATC system. The simulation models being developed by the FAA indicate that the national airspace system as currently configured is rapidly approaching capacity in critical sectors of the airspace. In the past the FAA could rely on local solutions to airspace congestion, but with the growing saturation of the system, changes in technical or procedural actions at one point may have the unintended consequences of moving the bottleneckâor delays and noiseâto some other part of the system. Simulation models, operations research, and other research tech- niques will becpme more important to the FAA in grappling with this problem. The FAA will need to develop and apply these techniques, first to put management decisions on a systems performance basis and, ulti- mately, to allow it to manage capacity more effectively. Recommendations Congestion Pricing Congestion pricing of runways could reduce congestion, encourage the use of underused airports, and provide additional revenues for enhancing capac-
12 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION ity. The DOT should permit and encourage airports to experiment with congestion pricing and invite evaluation of the effectiveness of these efforts by independent researchers. in the development of these proposals, the DOT should consider how to avoid the potential exercise of monopoly power by airports of airlines and their customers and how the revenues earned by congestion pricing will be used to provide needed additional capacity. Research The FAA should emphasize research on simulation modeling of airport and airspace capacity and related research. Greater use of such techniques would lead to the establishment of performance measures that would help the FAA make better use of existing airport and airspace capacity. The FAA should support and learn to use such research to manage capacity more effectively. IMPROVING FAA PERFORMANCE Notwithstanding deregulation, the FAA has the dominant role in providing ATC and ensuring air travel safety. Despite being hampered by the firing of most of its senior air traffic controllers in 1981, as well as by personnel ceilings and financial austerity attributable to the federal budget deficit, the FAA has been able to maintain a high level of safety since deregulation. To maintain this high level of safety in ATC, however, the FAA imposes considerable delay on commercial flights, particularly during inclement weather. Given the institutional constraints within which it must operate, the FAA will find it increasingly difficult to meet the growing demands placed on it by the commercial aviation industry, which raises concern for the future about the FAA's ability to provide adequate capacity while continuing to ensure safety. Institutional Constraints Many of the FAA's problems originate with the structural and statutory constraints within which the agency operates. As hav&all federal agencies, the FAA has been caught up in the efforts to respond to the budget deficit, especially by personnel ceilings imposed on the federal work force. Unlike most other federal agencies, however, the FAA in many respects is an
Executive Summary 13 operating agency in the sense that it is responsible for operating the ATC system 24 hours a day, 365 days a year. Its mission includes such important tasks as ensuring passenger safety in aircraft design and in the operations of air carriers. This mission requires management capability and technical expertise of the highest level. Foremost among the constraints of the FAA's institutional status and its management authorities and systems are the following: Inadequate flexibility in funding, personnel systems, and procurement; Frequent shifts in leadership because of the short tenure of admin- istrators; and Vulnerability to counterproductive intervention in technical and ad- ministrative matters (micromanagement) by the Office of the DOT Sec- retary, the Office of Management and Budget, and Congress. These constraints reduce the ability of the FAA to attract and retain the personnel needed for its high-technology mission, hamper its abilities to respond to a changing operational environment, and deprive the agency of stable leadership during a period that requires rapid change and innovation. Concerns About Future Performance The committee is confident that the FAA will maintain its commitment to safety, but the increased demand for air travel is likely to result in increased delay as controllers hold more flights on the ground until they can be merged safely into the traffic stream. As these delays mount, the FAA will come under increased pressure to change its practices in ways that could reduce safety, for example, by allowing a return to holding patterns for arriving aircraft. There is even reason to be concerned that the FAA will not have the expertise needed to keep the ATC technology fully operational. More effective performance of ATC in the future, however, will not be provided by simply hiring more controllers, although shortages in a few high-cost cities and questions about the appropriate numbers and expertise of maintenance technicians have been difficult to resolve. As outlined previously, the FAA must adopt more of a systemwide approach to managing ATC. Doing so requires both a reorientation in the FAA's approach and a more rapid development of the research and technology that the FAA will need to manage this task effectively and safely. This is one of the key points at which the constraints of being a government
14 WINDS OP CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION agency impede progress. The stable, long-term leadership needed to man- age this transition to a more complex, more technology-dependent mission is unlikely to be achieved with the frequent turnover of administrators that has been typiéal in the past. The more rapid acquisition of advanced technologies required to adapt the NAS Plan to unanticipated capacity needs is unlikely to be accomplished in time to meet demand using a detailed adherence to the federal procurement process. The technical and managerial expertise to accomplish this more sophisticated and demanding mission may not be available to the FAA within the personnel and salary constraints of the civil service system, even after the recent efforts of Congress to allow for pay differentials in high-cost areas of the country are taken into account. Finally, the funding needed for technology and personnel is likely to run afoul of the need to constrain federal spending in future years. Governmental constraints may also impede the conduct of the FAA's safety missions. For example, because of limited staffing, the FAA's maintenance inspectors tend to focus their attention on maintenance rec- ords instead of on inspection of actual maintenance. Whereas the FAA has the authority to perform more on-site inspections (and more unan- nounced inspections), the agency does not have sufficient personnel to make this a regular practice. The increased staffing levels for maintenance inspectors supported by the administration will help, but the demands are increasing faster than the staffing levels, particularly when the need to inspect aging aircraft is taken into account. Another cause for concern in other areas of the FAA's safety oversight responsibilities is its inability, within the pay limits imposed on the federal work force, to attract and retain the personnel required to carry out its mission, including test pilots, aeronautical engineers, maintenance inspectors, and electronics techni- cians. Given the likely limits on federal expenditures for the foreseeable future, it is difficult to be assured that staffing levels and the expertise of the staff will be sufficient to meet the challenges ahead. Because of the severity of these problems and the risks posed to safety, the proposals made in recent years for institutional reform of the FAA should be con- sidered seriously, and appropriate action should be taken. Options for Reform Three basic choices are available for institutional reform; the options range from fully public to private models:
Executive Summary 15 Return the FAA to its former independent status. (a) Assign the ATC functions of the FAA to a public corporation (a governmental enterprise) while the rest of the FAA's functions remain in the DOT, or (b) convert the entire FAA to a public corporation that is responsible only to the President and Congress or to the DOT Secretary. (a) Assign the ATC functions of the FAA to a congressionally chartered private corporation, or (b) assign the ATC functions to a congres- sionally chartered private corporation and imbed within this corpora- tion a legally independent, governmental unit which would be responsible for setting and enforcing safety standards, and for performing many of FAA's safety oversight and regulation duties other than those directly related to ATC. The first option, that of making the FAA independent, might afford some improvements in agency management but would not likely address needed fundamental reforms in pay, procurement, budgeting, and financ- ing. Significant progress in this regard would be made by converting the Airport and Airway Trust Fund into an enterprise fund, by allowing the FAA to charge users for its services, and by giving the agency's managers greaier discretion over personnel. In addition, the reforms in procurement suggested by the Packard Commission and others could be adopted, if the FAA were freed from restrictive federal procurement legislation. Options 2 and 3 incorporate these proposed reforms. Separating the FAA's ATC service from the rest of the agency and assigning it to either a public or private corporation, as suggested by options 2(a) and 3(a), would improve the efficiency of the ATC service. Separating the ATC functions of the FAA from many of its other safety and regulatory functions, however, could compromise safety. Although it has not been proven that separating the different functions of the FAA would compromise safety, many observers quite familiar with the FAA have argued that it would. Options 2(b) and 3(b) propose to retain this linkage by either keeping the separate functions of the FAA in one entity [Option 2(b)] or in close proximity [Option 3(b)]. Option 2(b), the public corporation option involving the entire FAA, is attractive because it would appear to provide more managerial discretion for all of the FAA in dealing with problems in funding, personnel, and procurement. This option would include the kinds of public oversight that are exercised over most federal corporations. Divorcing the corporation entirely from the DOT, however, would risk some loss of the ability to develop a comprehensive national transportation policy. It would also
16 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION require the establishment of a process or an agency to review the rates charged to its customersâanalogous to the Postal Rate Commission. To help address these shortcomings, Option 2(b) includes the option of es- tablishing a policy and reporting link to the Secretary of Transportation, whose authorities would not be delegable to lower-level officials. Whereas the private entity suggested in Option 3(b) would have the greatest flexibility in developing management and pay incentives that would improve performance of ATC functions, many would be uncom- fortable with complete privatization of the ATC service. The proposal to imbed a public safety oversight function within this private entity might help ease this concern. Options 2(b) and 3(b) would give the FAA the operational discretion and authority needed to address its most pressing problems. Congressional oversight would be retained, but the forms of accountability would differ. A more thorough delineation is needed, however, of how well existing corporate forms within the federal government exercise the greater dis- cretion they have. Also of interest is a comparison of how public ac- countability would be exercised in the two options. Without more study of this kind, it is unclear which option would provide for superior overall performance. The focus of options 2(b) and 3(b) is on improving the effectiveness of the FAA's headquarters operation and general staffing (both expertise and total number), which the committee believes deserve most emphasis. Little information is available, however, on whether either option would in any way place at risk the operational integrity of the agency. Although the FAA's performance can and should be faulted in some areas, the performance of its staff in the field indicates an admirable commitment to safety; the effect that any reorganization might have on this commitment deserves careful consideration. The committee is concerned about the ability of the FAA in its current form to meet future challenges posed by continued air traffic growth. This concern extends beyond the efficient provision of airspace capac- ity; the committee is not assured that the FAA in its current form, despite the best intentions of its managers and staff, will be able to continue to maintain the high level of safety in the aviation system that the American public enjoys and has come to expect. Of the options listed above, 2(b) and 3(b) would provide the operational authority and discretion needed to improve performance without severing links be- tween regulatory and operational functions, which may compromise safety. The committee recommends apublicly mandated study of change in the organization of the FAA by an independent group or organization
Executive Summary 17 that focuses on the relative merits and drawbacks of options 2(b) and 3(b), with a report to the President and Congress within 2 years after the study gets under way. SUMMARY OF RECOMMENDATIONS In order to preserve and enhance competition, the committee recommends the following: The DOJ should oppose mergers or asset acquisitions in which the carriers offer substantial parallel service or share a hub airport; however, the DOJ should not necessarily oppose mergers or asset acquisitions of carriers with complementary or end-to-end routes. The DOT should issue a regulation that would allow travel agents to use their own equipment or desk equipment leased from the host carrier to access multiple CRSs and extend the prohibition on bias to any software used in multiple access. The DOT should improve consumer information by requiring agents to disclose the incentive commissions (commission overrides) they receive from carriers. To enhance safety, the committee's recommendations are as follows: The FAA staffing and inspection procedures should be made suffi- cient to ensure that maintenance is adequate. The FAA should continue to reduce differences in equipment stan- dards and operating requirements between national and regional (Com- muter) carriers. To use existing capacity more effectively, the committee recommends the following: The DOT should permit and encourage airports to experiment with congestion pricing for runway use while guarding against the exercise of monopoly power by airports and preserving incentives to invest the rev- enues earned in new capacity. The FAA should emphasize research on simulation modeling of air- port and airspace capacity and related research and learn to use such research to more effectively manage capacity.
18 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION The committee's recommendation for better future air travel capacity and safety is as follows: A publicly mandated study of change in the organization of the FAA should be conducted, the focus of which should be on the choice between a public and private corporate model. A report should be submitted to the President and Congress within 2 years of initiation of the study.
PART Introduction
ii Deregulation of Commercial Aviation The commercial aviation industry in the United States has grown at an almost phenomenal rate since the end of World War II. In 1945 the major airlines flew 3.3 billion revenue passenger miles (RPMs) By the mid-1970s, when deregulation was beginning to develop political currency, the major carriers flew 130 billion RPMs. By 1988, after a decade of deregulation, the number of domestic RPMs had reached 330 billion (DOT 1984-1990; CAB 1973). Although the original aviators were daredevils (31 of the first 40 pilots in airmail service were killed on duty), during its growth the aviation industry established a remarkable safety record (Lederer and Enders 1988). Today's commercial airline pilots have the same average life expectancy as the rest of the population (Lederer and Enders 1988). Commercial aviation affects the entire country. Since World War II, the industry has provided air service throughout the United States and has become a $254-billion sector that directly employs more than 2 million people in airline and airport operations (Wilbur Smith Associates 1989). Airline deregulation serves as a useful point from which to review recent trends in commercial aviation service and safety because it significantly reduced the role of government in regulating the service provided by the industry. Nevertheless, the government role is still quite important. The federal government continues to provide air traffic control (ATC) services, regulate safety, allocate funds for airport construction and expansion, and have responsibility for antitrust enforcement and consumer protection. Local and state governments and independent regional authorities are also directly involved as airport operators. Events independent of deregulation have affected air passenger service since deregulation, for example, the fuel crises of 1979, the air traffic controllers' strike of 1981, the severe recession from 1980 through 1982, 21
22 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION increased financial leverage, and at the time of this writing, the recession and fuel price shocks caused by the 1991 Persian Gulf War. Many changes in aviation, however, are more directly a result of deregulation. As ex- pected, the competition introduced by deregulation brought discount fares and more flight options for air passengers. Deregulation also ushered in many unanticipated changes: the importance of hub-and-spoke networks, the growing influence of computer reservation systems (CRSs), travel agent commission overrides, frequent flier programs, and the increased dominance of individual airlines at some hub airports. The following section provides a brief review of the forty-year history of air carrier regulation. The premises that led to regulation are outlined, and the criticisms that led to deregulation are summarized. In the next section the early experience with deregulation is reviewed, and some key points that carry forward throughout the remainder of the report are high- lighted. Although it is difficult to isolate the specific effects of deregu- lation, this policy shift began fundamental changes in air carrier management and operations. These changes can be observed and their consequences traced. The basic changes in air carrier management are reviewed in Chapter 2. Their effects on consumers are analyzed in Chapter 3. Impe- diments to the performance of the deregulated marketplace that threaten the benefits estimated in Chapter 3 are examined in Chapter 4. The effects of changes in air carrier operations on safety are examined in Chapter 5, which also introduces an assessment of the adequacy of the public sector, primarily the Federal Aviation Administration (FAA), to respond to changes since deregulation. This latter issue is taken up again in Chapter 6, in which the focus is on the public sector's responsibility for providing capacity. The key point that emerges from chapters 5 and 6 is that the federal government's responseâwhich is carried out by the FAAâhas not kept pace with the public's demand on the system. The basic impe- diments to the performance of the FAA are outlined in Chapter 7. The committee's recommendations to improve air passenger service and safety are presented in Chapter 8. BRIEF HISTORY OF REGULATION Government involvement in the aviation industry began almost at the birth of manned flight. The U.S. Post Office began promoting aviation in 1918 to transport mail (Morgan 1981). Mail subsidies provided a stable cash flow for airline entrepreneurs and served as nursemaid to a newborn industry. As the passenger market grew, however, small carriers began
Deregulation of Commercial Aviation23 offering competitive services on the most profitable routes (the Postmaster General could not regulate entry). At the onset of the Great Depression, demand fell and the subsidized carriers that were losing revenues to the unregulated carriers began lobbying for protection (Levine 1965). A bid- rigging scheme in which the Postmaster General connived to guarantee markets for a few airlines caused a major scandal and led Congress to develop a different regulatory approach. The debate that followed occurred in an era of public fear of destructive competition that could lead to monopoly and was fueled by the economic hardship of the depression. The policies and regulatory approach ultimately adopted in 1938 legislation were, in many respects, patterned on the regulatory structure of the truck- ing industry, which had been regulated out of fear of destructive com- petition, with elements drawn from regulation of the railroad industry, which also had developed out of fear of both monopoly and destructive competition. Precedent and Rationale for Regulating Airlines The Interstate Commerce Act of 1888 formed the legal framework for regulating the country's first truly big private enterprisesâthe railroads (Meyer et al. 1959). Congress established a commission form of regulation (in the public utility context) in an attempt to protect the public from price gouging that resulted from the monopoly position railroads had on many routes and to protect the railroads from savaging each other through de- structive pricing on competitive routes. Later, as the industrialization of the late 19th and early 20th centuries led to giant enterprises in other industries such as steel and oil, Congress relied more heavily on antitrust laws to regulate unfair competition. As Meyer et al. (1959) pointed out, had the railroad industry emerged after antitrust legislation was developed, the public utility model of regulation, which applied to the aviation, rail, and trucking industries until recent years, might never have been adopted. Economic Rationale for Regulation The main economic argument supporting regulation of the railroads was that the industry resembled a natural monopoly (Meyer et al. 1959). Because the cost of entry is exceedingly high, in many markets only a single carrier can profit from providing service. In addition, the size required to operate results in economies of scale that make it possible for
24 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION one firm to provide service at a lower cost than could two or more firms, and thereby dominate the market. Even the railroad industry, however, does not fit the monopoly model perfectly. By the late 1 800s many cities were served by competing railroads (Meyer et al. 1959). Railroads also faced stiff competition in some parts of the country from canal, inland waterway, and Great Lakes steamers. Well after the creation of the Interstate Commerce Commission (ICC), dramatic growth in the use of the auto-truck in the 1920s created a new form of competition for the railroads. Indeed, many have argued that a major driving force for the creation of the ICC was not just relief from monopoly pricing but relief from competition, and that the push for ICC regulation of the trucking industry was partly motivated by the fear of destructive motor carrier competition held by rail and motor carrier interests. Another political force leading to regulation of the railroads came from alliances of agricultural and business interests pressing for relief from arbitrary pricing (Meyer et al. 1959). Railroads cut prices below costs on competitive routes and covered costs by charging higher prices on routes they monopolized. The ICC relied on an adjudicatory process in an attempt to establish fair rates. This regulatory system was applied to motor carriers in 1934, when they began to compete with railroads and with each other. Congress also adopted the public utility form of regulation in 1938 when it created the Civil Aeronautics Authority, which was reorganized and renamed the Civil Aeronautics Board (CAB) in 1940. It did so partly because it was concerned that the small scale of the industry and the number of willing entrants would lead to excessive competition, which would ultimately have negative effects on service and safety, perhaps leading to monopoly (Levine 1965). Although the economic arguments for regulation of the airlines, par- ticularly fear of monopoly and industry fear of competition, are similar to those that led to the regulation of the railroads, supporters of deregu- lation argued that some important differences existed in the airline in- dustryâthe most notable difference being that the scale required for entry into the airline industry is substantially smaller than that for the railroad industry. Unlike a railroad, an air carrier does not have to make a massive, unrecoverable investment in right-of-way. Aircraft, though expensive, can be leased or mortgaged. Because airlines appeared to incur fewer unre- coverable costs on entry, there was less fear that investments will be wasted if an entrant could not survive once it enters a market. Instead of taking a complete loss, the entrant could move to other markets. With the experience of more than a decade of deregulation, however, it is clear that the investments in the hub-and-spoke system required for much airline
Deregulation of Commercial Aviation 25 entry can be very large. Airline entry involves substantial pre-operating expenses, which may be only partly recoverable, and may include ever more substantial operating losses while customers are drawn to the new service, much of which cannot be recovered on exit. Some carriers exiting a market are able to redeploy assets to other markets and recover part of their losses; even larger sums can be lost if the new entrant ceases op- erating, as most have since deregulation. Whether airlines reach economies of scale as they increase in size has long been studied and debated by economists, but if economies of scale exist, concern about monopoly has greaterjustification. The largest carrier, through economies of scale, would have the lowest unit cost and would therefore be able to price all competitors out of the market. In the San FranciscoâLos Angeles market, however, Levine (1965) found that new entrants such as Pacific Southwest were able to operate at unit costs well below those of national (trunk) carriers such as United.' Subsequent studies by Caves et al. (1984) and Gillen et al. (1985) did not find evidence of economies of scale in the U.S. or Canadian airline industries.2 Levine (1987), however, developed a theoretical justification for and found sub- stantial evidence of economies of scale and scope in the airline industry that have emerged since deregulation. In early discussions about aviation deregulation, economists argued that given the mobility of airline investments, any carrier attempting to charge monopoly prices would be contested by a new entrant offering services at a lower cost (MacAvoy and Snow 1977). This argument was later formalized as contestability theory (Bailey and Panzar 1981; Baumol et al. 1982). According to this theory, the threat of entry would serve the same disciplining function as actual entry. However, subsequent analysis of the performance of airline markets after deregulation finds little evidence of contestability in airline markets, partly because of the difficulty of gaining access to congested airports, but also because of the competitive features of hub-and-spoke systems, frequent flier programs, travel-agent incentive programs, and the operation of CRSs (Bailey et al. 1985; Levine 1987; Bailey and Williams 1988). Provision of a Social Good Another rationale for the regulation of both railroads and airlines is that transportation has important social effects. In particular, the provision of transportation service to rural areas at below-cost prices through various forms of subsidy has often been promoted on the grounds that transpor-
26 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION tation binds the country together and promotes economic development (Fromm 1965). Safety Regulation A final, and perhaps the most widely accepted, rationale for regulating transportation is that it presents public safety issues. Regulation of airlines for reasons of safety resulted from concerns about transaction costs and limited consumer information, and fear that ruinous competition would lead carriers to cut costs and expose the public to unnecessary risks. Fear of the effects of ruinous competition was an argument also used by op- ponents of airline deregulation. The CAB Regulatory Approach Congress gave the CAB authority over routes, fares, and market entry. The CAB also had safety oversight responsibilities, which were transferred to the Federal Aviation Agency in 1958. Throughout its history, however, the CAB usually exercised its economic regulatory authority in a piece- meal, ad hoc fashion, based largely on applications from individual airlines for adjustments (Bailey et al. 1985). It did make several attempts to regulate on a comprehensive, industrywide basis, but they were generally regarded as cumbersome, rigid, and ultimately unsuccessful. Entry The CAB rarely allowed new entrants into the industry. In 1938, 16 existing carriers were "grandfathered" into trunk carriers by Congress. In 1978 these same carriers (reduced to 11 through mergers) accounted for 94 percent of scheduled traffic (FAA 1989a). The 11 trunk airlines were United, American, Delta, Eastern, Trans World, Western, Braniff, Continental, National, Pan American, and Northwest. New carriers were permitted into the industry only under fairly restric- tive conditions. After World War II, the CAB permitted new carriers to serve routes with low traffic volumesâessentially small cities. However, it restricted the ability of those carriers to compete with the trunks until the late 1960s (Bailey et al. 1985). The so-called local service airlines developed out of this policy; by 1978 nine such carriers remained. The
Deregulation of Commercial Aviation 27 local service airlines at this time were Allegheny, Airwest, Hughes, Fron- tier, North Central, Ozark, Piedmont, Texas International, and Southern. The trunk carriers could have provided the same service at a higher subsidy (Morgan 1981). Subsidies to the trunk carriers to provide service to small communities were largely phased out during the 1950s as the local service airlines took over subsidized service. The CAB also allowed charter carriers to develop, but they were limited to providing only charter flights and were not allowed to compete for scheduled service. Similarly, all-cargo carriers began offering service, but the CAB did not allow them to compete in the passenger market. Intrastate ' carriers were exempt from CAB economic regulation. The CAB exempted from most of its economic regulations the commuter airlines (effectively limited to aircraft that seated no more than 19 passengers, later 60 pas- sengers). Substantial intrastate carriers emerged in Texas and California, and commuter carriers emerged in many markets, particularly those in which the CAB permitted certificated carriers to withdraw from required service. Routes The CAB tightly controlled the routes that individual airlines could serve and those that they could cease to serve. The CAB granted carriers routes to ensure service to communities but attempted to accomplish other policy goals as well. Attractive routes might be granted to a carrier in financial difficulty or to cross-subsidize service on a route with light traffic that the CAB required a carrier to serve (Meyer and Oster 1981). Route au- thority was granted through a cumbersome, expensive, time-consuming process. Once a carrier had established service, it was difficult for another carrier to enter the market and compete. The entrant had to demonstrate that the authority was "required by the public interest, convenience and necessity." The incumbent carrier could protest the entry by claiming that it would be harmed. The board had an unofficial policy that limited the number of carriers in individual markets, which essentially limited the amount of competition (Bailey et al. 1985). Fares Fares were regulated in two ways: (a) by direct approval or disapproval of a proposed fare increase and (b) by establishment of a fare or a narrow
28 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION range of fares for a given route (Meyer et al. 1959). Price competition under CAB regulation was atypical, but did occur. Although discount fares were subject to review and could be protested by other carriers, the CAB allowed some discount fares. During the late 1950s, for example, a night coach discount fare was allowed, as were discounts for families and transcontinental trips (Bailey et al. 1985). With the introduction of jet aircraft in the 1960s, carrier operating costs declined substantially as seat capacity increased. The CAB allowed more discounts during this period, and average fares declined. Even with the discounts, however, pricing adjustments permitted by the CAB did not keep pace with the impact of technology on industry costs. During the 1960s, when jets were being adopted by the trunk airlines, average seat-mile costs declined 21 percent, but average fares declined only 7 percent (Bailey et al. 1985). In markets in which carriers did compete, competition tended to be on the basis of service (scheduling and amenities) instead of price, which inflated costs as carriers attempted to offer the latest aircraft or most frequent service (Douglas and Miller 1974). A growing number of critics believed that the industry created and nurtured by the CAB was stable but inflexible and inefficient. Impetus for Deregulation The popularity of deregulation as a political issue during the late 1970s was preceded by a considerable body of academic research that was critical of regulation (Derthick and Quirk 1985). The justification for the regu- lation of transportation industriesâto promote stability and avoid exces- sive competitionâhad long been perceived as unwarranted (Keyes 1951; Meyer et al. 1959; Caves 1962; Levine 1965; Keeler 1972). Keyes (1951) could not find any justification for the CAB's restrictions on entry. Caves (1962) did not find evidence of economies of scale that would warrant the CAB's reluctance to allow competition. Meyer et al. (1959) predicted that cessation of CAB entry and fare regulation would increase competition on the heaviest routes but would not result in excess capacity; instead the industry was likely to remain dominated by a few large firms. Meyer et al. (1959) and Caves (1962) argued that the market structure of aviation, though concentrated, was little different from other near-oligopolies in the U.S. economy that operate without direct government regulation. By prohibiting price competition, the CAB promoted an industry that was inefficient.
Deregulation of Commercial Aviation 29 Levine (1965) took note of the discount fares available in intrastate markets in California that were not regulated by the CAB and concluded that allowing price competition nationwide would bring lower prices for consumers. (Intrastate carriers in California were regulated by the Public Utilities Commission, but it allowed free entry and regulated fares lightly.) Keeler (1972) estimated that consumers paid 48 to 84 percent more for regulated fares than for roughly comparable flights provided by unregu- lated carriers. Not everyone greeted the prospects for deregulation with enthusiasm. Among others, Secor Browne, a former chairman of the CAB, argued that because of the overcapacity in the airline industry and the strength of the biggest carriers, the strong airlines would simply "push the weak ones off the cliff" (Brenner 1988). Smaller carriers were worried that they would lose market share on routes on which they competed with the largest carriers. Without adequate passenger flow on those routes, the small carriers might suffer loss of passenger flow throughout their entire network (Pearson 1986). Most trunk carriers also opposed deregulation out of fear of losing control of preferential routes, but they were not united in their opposition (Derthick and Quirk 1985, 154). Many observers of the industry were concerned that small communities that received subsi- dized air service would lose service completely. They also worried that intense competition would inevitably affect the industry's overall com- mitment to passenger safety (Pearson 1986). ADMINISTRATIVE AND LEGISLATIVE REFORM Passage of the Airline Deregulation Act of 1978 is frequently cited as the point at which deregulation began, but the CAB actually had begun dereg- ulating as early as 1975. John Robson, Chairman of the CAB from 1975 to 1977 under President Ford, supported several administrative changes, which can be labeled "administrative deregulation," to increase com- petition (Morgan 1981, 44). In 1975 certain markets that had been shielded from competition were opened to other carriers, and more markets were opened in 1976 and 1977. In 1977, among its last acts under the leadership of Robson, the board gave carriers much greater discretion in pricing discount seats; for example, it permitted Texas Air's Peanuts fares and American's Super Saver fares (Morgan 1981). Under the chairmanship of Alfred Kahn, who was appointed by Pres- ident Carter, the CAB allowed carriers to offer additional forms of discount
30 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION fares. In early 1978 the CAB announced that it would no longer review fare reductions of up to 50 percent of the standard industry fare level (SIFL) and that it planned to allow freer route entry and exit (Morgan 1981). As the Deregulation Act of 1978 was being debated in Congress, carriers were enjoying the early returns of administrative deregulation profits, traffic, and load factors were all higher than 1977 levels. In late 1978 President Carter signed the Airline Deregulation Act of 1978. The law set the end of the CAB's authority over routes (December 31, 1981) and domestic fares (December 31, 1983) and dissolved the CAB itself (De- cember 31, 1985). By mid-1979, the CAB granted all domestic route ap- plications as a matter of course and approved virtually all fares within a wide zone. Subsidies to support essential air service to small communities were continued through 1988 (and were subsequently extended by Congress), with administrative responsibility given to the U.S. Department of Transportation (DOT). Antitrust oversight was given to the U.S. Department of Justice (DOJ), but was later shifted to the DOT in the 1984 legislation that ended the CAB (such authority shifted back to the DOJ in 1989). Carriers must still demonstrate to the DOT [and the Federal Aviation Administration (FAA)] that they can offer safe service and must demonstrate adequate working capital and the ability initially to protect funds deposited with them ("financial fitness"), but are otherwise free to offer service. Although antitrust authority has shifted to the DOJ, the Office of the Secretary of DOT continues to have responsibility inherited from the CAB for regulation of CRSs, essential air service determinations, and consumer protection matters (lost baggage, on- time performance, involuntary bumpings, and protection against fraudulent advertising). Competition and Traffic Growth With the easing of CAB regulations and passage of the Airline Deregu- lation Act, the industry entered an era of intense competition. Major airlines and the former local service carriers began competing with one another, charter carriers moved into scheduled service, former intrastate carriers such as Air Florida, Pacific Southwest Airlines, and Southwest moved into interstate markets, and numerous new firms began offering service. As the differences between the trunk airlines and the local service airlines and the large former intrastate carriers became blurred, and with the emergence of new entrants, the formal definitions of carriers changed. Under Part 121 of the Code of Federal Regulations (14 CFR 121) carriers that earn more than $1 billion annually are now called "majors." Carriers
Deregulation of Commercial Aviation 31 that earn less than $1 billion but $100 million or more are called "na- tionals." Those that earn less than $100 million are called "regionals." This nomenclature is used in this report in reference to the postderegulation industry, except that in some sections regional airlines are included in the same category as commuters. In many respects, particularly in the types of aircraft operated, the regionals and the large commuters (which account for most commuter traffic) are quite similar. The majors, nationals, and large regionals are listed in the text box. Between 1978 and early 1984 the number of airlines that were reporting financial data (the Form 41 reports, which excludes commuters) to the DOT increased from 43 to 87 (DOT 1990a). As a result of this new entry, discount fares proliferated, fare wars began, and total traffic increased dramatically as passengers hustled to take advantage of previously unheard-of coast-to- coast fares. The number of RPMs almost doubled. The established major airlines, with the exception of those that were failing, shared in the traffic growth, but the new entrants made substantial inroads into market share. Between 1978 and 1985, the share of total traffic of the incumbent trunk airlines declined from 94 to 77 percent (FAA 1989a, 56). Commuter air travel also grew markedly during this period. Commuter carriers operate aircraft with fewer than 60 seats and are governed by Part 135 of the Code of Federal Regulations (14 CFR 135). In 1978, 210 commuter (or regional) airlines offered service and flew about 200 million RPMs. By 1988 total commuter RPMs had increased severalfoldâto nearly 1.4 bil- lionâbut the number of carriers had declined to 176. The largest 50 com- muter airlines carried about 92 percent of all passengers (FAA 1989a, 87). By 1990, however, of the 148 new entrants that were reporting financial data to the DOT, only 44 remained. Only 14 of the new entrants classified as majors, nationals, and large regionals were still operating scheduled domestic passenger service in mid-1991 (see text box). The number of firms entering the industry each year fell from a high of 22 in 1979 to only 5 in 1987 and 3 in 1988 (DOT 1990a). In 1978, 210 commuters offered service compared with 176 in 1988 (FAA 1989a). The shrinking number of carriers has improved the market share of the largest airlines, such that the largest carriers now have a somewhat greater market share than before deregulation (although they are not all the same carriers). Effects of Deregulation Identification of the specific effects of deregulation on the airline industry is complicated by other major changes occurring at roughly the same time
32 WINOS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION COMMERCIAL AIRLINES OFFERING DOMESTIC SCHEDULED PASSENGER SERVICE AS OF JULY 1991 Majors America West' Pan American" American Southwest" Continental0 Trans World Delta United Northwest USAIr Nationals Air Wisconsin" Markair Alaska Midway°" Aloha Midwest Express" American Trans Air" Towerb Hawaiian Trump Shuttle Horizon"" Westairb Large Regionals Braniff International"' MOM Grand" Carnivalb Reeve Executive Airlines"," Trans States)d 01n Chapter II bankruptcy in 1991. (Eastern, a former major, ceased op- erating during early 1991.) bNew entrant as a major, national, or large regional after 1978. "Classified as majors on January I, 1990. "Formerly operated as a commuter carrier. "Formerly bankrupt, reinitiated service in July 1991. SouRcE: Classification and list of carriers provided by the Office of Aviation Statistics, DOT. deregulation was introduced and by long-term trends that influence air caniers. Shortly after deregulation began, the 1979 fuel crisis caused fuel costs to skyrocket and precipitated a recession that dampened demand through 1982. With additional constraints on air travel imposed by the shortage of air traffic controllers after the controllers' strike, the full effects of deregulation were probably not apparent until the mid-l980s. Several long-term trends also affected the industry, most notably the consistent growth in demand for air travel in the postâWorld War II era and the shift to jet technology in the 1960s, followed by the introduction of wide- bodied aircraft (e.g., the Boeing 747) in the 1970s.
Deregulation of Commercial Aviation 33 Billions 350 300 250 200 150 100 50 OL_ 1960 1965 1970 1975 1980 1985 1989 Year FIGURE 1-I Revenue passenger miles, scheduled domestic service of certificated carriers (vertical bars represent, from left to right, administrative deregulation, De- regulation Act, fuel crisis, PATCO strike) (CAB 1973, DOT 1974-1990). The selected indicators of industry output, productivity, and average prices shown in Figures 1-1-1-5 illustrate the difficulty of isolating the effects of deregulation from other major influences. During the 1960s, for example, the number of RPMs grew steadily (Figure 1-1). There is some indication that the growth rate declined in the early 1970s as demand slackened after the 1973 fuel crisis and during the subsequent recession. If one uses 1975-1976 to demark the beginning of administrative deregu- lation, it appears that the easing of regulation under Chairman Robson and the introduction of more discounts contributed to a burst of travel, which was eclipsed by the 1979 fuel shortage and price hikes of 1980. By the time the FAA began recovering from the controllers' strike in 1982, RPMs had begun another growth spurt, growing far faster than under regulation, and the growth rate slowed only with the flattening of domestic demand from 1988 to 1990. The effects of deregulation on productivity appear more distinct if the trend in RPMs is considered together with aircraft departures (Figure 1-2) and load factors (Figure 1-3). Although the total number of departures de- clined from 1970 to 1975, the number of RPMs increased because of larger aircraft, increased load factors, and longer trips. The rapid growth in the number of RPMs after 1975 exceeded that of departures because load factors increased to historic levels. The number of aircraft departures declined after
Thousands 7000 i 6000 5000 4000 3000 2000 1000 0 IIIIIIIIII IIIII IP III I I I 1960 1965 1970 1975 1980 1985 1989 Year FIGURE 1-2 Aircraft departures, scheduled domestic service of certificated air carriers (vertical bars represent, from left to right, administrative deregulation, De- regulation Act, fuel crisis, PATCO strike) (CAB 1973, DOT 1974-1990). Percent 80 70 60 50 40 30 20 1965 1970 1975 1980 1985 1989 Year FIGURE 1-3 Load factors for domestic scheduled service of certificated air carriers (vertical bars represent, from left to right, administrative deregulation, Deregulation Act, fuel crisis, PATCO strike) (CAB 1973, DOT 1974-1990).
Deregulation of Commercial Aviation 35 Index, 1972=100 350 300 250 200 150 100 50 0 - 1970 1975 1980 1985 1988 Year Output Cost Index â' Input Cost Index FIGURE 1-4 Trends in airline costs and output, 1970-1988 (vertical bars represent, from left to right, administrative deregulation, Deregulation Act, fuel crisis, PATCO strike) [based on work by Meyer and Oster (1987, Figure 6.1)]. the 1979 fuel crisis as carriers reduced service, but began booming with the considerable new entry that occurred in the early 1980s. Along with the productivity of greater load factors, a general indication of the increased productivity of the airline industry can be gained by comparing the trend in input costs with the trends in the cost of producing outputs (Figure 1-4). Input costs are based on the air carrier cost index, which is currently produced each year by the Air Transport Association of America. The cost of outputs is measured by total domestic air carrier expenditures divided by total output of available seat miles. Both trends are indexed to 1972 to allow for comparison. Figure 1-4 is derived from and extends Figure 6.1 in work by Meyer and Oster (1987), which showed the trend in these two cost indexes from 1972 to 1980. As Meyer and Oster (1987) noted, if one examines industry performance from 1976 to 1980, productivity improved because the cost of producing outputs in- creased far more slowly than the cost of inputs. As indicated by the extension of these trends though 1988, air carriers managed to stabilize the cost of producing outputs even as input costs continued to grow. The productivity gain attributable to deregulation appears to result from "a redeployment of the factors of production away from regulation-induced service patterns toward patterns determined by market demand" (Meyer and Oster 1987). Such a redeployment of assets would be expected to
36 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION Cents per mile 8 OL_ 1960 1965 1970 1975 1980 1985 1988 Year FIGURE 1-5 Average yield (1967$), scheduled service of domestic certificated carriers (vertical bars represent, from left to right, administrative deregulation, De- regulation Act, fuel crisis, PATCO strike) (DOT 1990). result in a one-time productivity gain; further improvements would depend on the ability of the industry to continue to capitalize on new technology. As a measure of the prices charged consumers, trends in the average fare can be approximated by viewing the trend in average yields. (Yield is total passenger revenue divided by RPMs traveled.) The trend in average yields shown in Figure 1-5 is expressed in constant (1967) dollars. Average yields actually increased in both constant and current dollars from 1957 to the early 1960s, after which a steady decline occurred, made possible largely by the efficiencies of jet aircraft. With the increased flexibility and discounting during administrative deregulation, fares began to fall. Yields increased sharply after 1979 as carriers sought to cover rapidly escalating expenses as fuel increased from roughly 20 to 30 percent of total costs (Bailey et al. 1985). After the fuel crisis and the strike by the Professional Air Traffic Controllers Organization (PATCO), yields began to decline again; they increased modestly after 1985. From an examination of the long-term trend in average yield one might infer that deregulation had no effect. Such an inference, however, would fail to take into account the multiple influences, some short-term and some long- term, that affect pricing. Some studies have attempted to unravel the early effects of deregulation from the effects of the 1979 fuel shortage using various kinds of multivariate models. Fuel price increases and the recession had a substantial short-term effect on the airline industry just as deregulation began.
Deregulation of Commercial Aviation 37 The average fuel price per gallon paid by carriers increased from about 35 cents in late 1978 to almost 70 cents by the end of 1979, which helps to explain why fares were increased during this period (Oster 1981, 162). Oster (1981) estimated that had average fuel prices not increased between March 1979 and March 1980, major and national carriers would have saved $1.9 billion, which was roughly 10 percent of total costs for 1980. Gomez-Ibanez et al. (1983) estimated that without the fuel price increase, major and national carriers would have shown net operating income gains between $150 million and $660 million in 1981. Morrison and Winston (1986) found a similar effect; that is, although the industry lost money after the fuel shortage, the losses would have been far greater without the flexibilities given to managers through deregulation. In contrast to these studies, however, Dempsey (1990) argued that once the long-term trend in average yields was adjusted to account for changes in fuel prices, the effect of deregulation on prices, though positive, had evaporated by 1988. Dempsey (1990) further argued that the adjusted trend in yields declined more before deregulation than afterward. This trend analysis, however, only adjusts for a single influence (fuel prices) and omits other significant influences on the trend in airline costs. For example, the introduction of wide-bodied jet aircraft (such as Boeing 747s, L-10i is, and DC-lOs) into the fleets of most airlines during the early 1970s greatly reduced the average cost of producing seat-miles; no com- parable technology gain occurred in the deregulation period to help explain the continued productivity improvements of those early years (Meyer and Oster 1981, 77). The industry undoubtedly continues to enjoy productivity benefits as a result of technology (including gains in engine efficiency and technological advances that allow reductions in crew size on the newest aircraft). It is difficult to assign relative shares between the productivity benefits of deregulation and those of technology, which were large before deregulation and continued to be significant afterward. Nonetheless, the productivity benefits of allowing managers to redeploy assets more effi- ciently, though a one-time benefit, have not disappeared; they have become part of the base on which additional productivity gains are added. To argue that the benefits of deregulation had dissipated by 1988 on the basis of a simple linear regression model predicting postderegulation prices, which only adjusts for fuel costs, is just too simplistic.5 Models sophisticated enough to control for all the significant short- and long-term influences on the air carrier industry over a multi-year period have not yet been devised and estimated and, given the complexity of the methodological issues (to say nothing of weaknesses in the data), are unlikely to ever gain widespread acceptance. It is possible to separate
38 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION short-term effects using statistical techniques in order to indicate the effects of deregulation during periods when few other influences are highly vari- able, but separating the effects over the long term depends more on judg- ment than statistical interpretation. Given the relative consistency of some trends, particularly total RPMs and average yields, it can be argued that deregulation did not cause a decisive change in industry output and average fares (Dempsey 1990; Brenner 1988). On the other hand, a plausible argument can be made that the productivity gains indicated in Figures 1-2-1-4 would not have occurred without regulatory reform because the industry had reached maturity. The productivity benefits of phasing a new era of jet technology into the airline fleet had largely been absorbed by the mid- l970s, and sub- sequent technological benefits have been less dramatic. Given the slow- moving regulatory environment, it is reasonable to conclude that the pro- ductivity benefits of redeploying assets away from administrative require- ments to market demand would not have occurred. It is also unlikely that carrier managers and labor forces would have experienced the same pressures to cut and control costs that have been experienced since deregulation. Without these changes, fewer of the cost reductions experienced by the industry would have been realized or passed along to consumers. This would have caused average yields to fall more slowly. Evidence for this can be seen from a comparison of average yields in Canada to those in the United States. Canadian airlines were not deregulated until 1987. Although the trend in yields in the two countries before U.S. deregulation was similar, in the early 1980s average yields in the United States declined considerably more than in Canada (Jordan 1991). Even so, it is clear from the highly aggregated trends presented in Figures 1-1-1-5, and the meth- odological difficulties associated with disaggregating them, that there will always be room for disagreement about the specific effects of deregulation. Despite the difficulty in measuring precisely the effects of deregulation, it is possible to focus on the specific changes made possible by deregulation and to arrive at reasonable judgments about, how these changes affect the industry and consumers. ORGANIZATION OF REPORT Changes in Airline Operations and Financial Performance Deregulation gave airline managers freedom that they had never had under the cumbersome administration of the CAB. At the same time, deregu-
Deregulation of Commercial Aviation 39 lation made competing in the marketplace, instead of providing a high level of service in a protected environment, the test of managerial success. The freedoms of the market, however, mean that successful top managers not only have to compete with each other, they also have to cope with such things as hostile takeovers and leveraged buy outs. Deregulation has helped usher in a new generation of airline managers, and demanded of them greater skill in managing carrier finances. Despite improved productivity, the profitability of the airline industry as a whole has lagged behind that of industry in general both before and after deregulation. Brenner (1988) expressed concern that low prof- its are the result of an inherent tendency toward excess capacity in the airline industry; that is, airlines feel compelled to offer more flights or better service than their competitors, which ultimately leads to offering service at prices below marginal costs. Low profits may not be in and of themselves a problem unless cost cutting compromises the ability of the industry to attract capital for reinvestment (Kahn 1988). Available indicators of the profitability and financial stability of the industry are summarized in Chapter 2. Passenger Fares and Airline Service U.S. Senate hearings in 1975 convinced many that prices regulated by the CAB and charged to consumers were too high (Bailey et al. 1985). The prospect of lower fares for consumers helped build the coalition of consumer groups and business interests that actively lobbied for deregu- lation. Chapter 3 includes an analysis of available data and a review of studies on the service and fares offered to consumers since deregulation. In the mid-1980s, when the number of new entrants and the com- petitiveness of the industry was at a peak, bargain fares proliferated. As the share of traffic controlled by major carriers has increased, how- ever, and as these carriers have built dominant positions at their hubs, concern has risen about the competitiveness of the industry and the fares and service being offered to consumers (GAO 1988; DOT 1989; Borenstein 1989). Trends in the prices paid by consumers, disaggre- gated by type of travel, city size, and level of competition, are reviewed in the first half of Chapter 3. One of the primary concerns about deregulation was the effect that it might have on the fares and service provided to residents of small com- munities. The available data on small community fares and service are analyzed and summarized in Chapter 3; however, the data collected from
40 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION commuter carriers are substantially less extensive than the data collected from major, national, and regional carriers, which limits the analysis. Barriers to Competition As described in Chapters 1, 2, and 3 of this report, the airline industry is much more competitive and efficient than it was under regulation. However, the very success of the major airlines in adapting their marketing tactics to the new environmentâas witnessed by their growing market shareâalso raises concerns about the future competitiveness of the in- dustry. The top five airlines now have a larger share of domestic RPMs than they had in 1978 (up from 69 percent of domestic RPMs in 1978 to 74 percent in 1988), but are also competing with each other in many more markets than before because of the extension of their hub-and-spoke net- works (DOT 1990a). The remaining new entrants now have only about 5 percent of the market and are finding it difficult to expand. All the major airlines have strengthened their positions at their hubs and have become the dominant carriers in several major markets. Ac- cording to contestability theory, in order for the benefits of deregulation to be realized potential competitors should be able to enter the market if monopoly fares are being charged; fear of entry by other carriers should keep the airline already serving the market from charging monopoly prices. The financial risks of competing with a carrier at its hub are considerable, and new entrants have found it expensive to gain access to many airports that serve major markets. The advantages enjoyed by the major airlines through their extensive network of affiliated travel agents and CRSs can make it difficult for new entrants and smaller airlines to enter new markets to increase their market share. It is also true that by listing with CRSs a small carrier can make its schedules and fares known to travel agents all over the country (albeit at considerable expense). Travel agents clearly benefit consumers by min- imizing the time required to find out about all the existing alternatives, which have increased in number and complexity since deregulation (Levine 1987). The problem arises if travel agents with an eye toward earning extra commissions steer passengers toward their CRS-affiliated airlines and away from lower-cost or more convenient alternatives. Computer reservation systems have been controversial and remain so (Levine 1987). The importance of CRSs, airline relationships with travel agents, and airline marketing strategies, along with the difficulty of entry at airports serving key markets, raises concern about the ability of low-
Deregulation of Commercial Aviation 41 cost new entrants to expand. This, combined with the precarious financial position of more than one major airline, makes further industry concen- tration a legitimate public policy concern. Potential barriers to entry and competition are reviewed in Chapter 4. Commercial Aviation Safety The Airline Deregulation Act of 1978 explicitly states that Congress had no intention of allowing safety to be deregulated. Nevertheless, those resisting deregulation during the debates of the 1970s expressed concern about the commitment to safety of the new entrepreneurs and their ability to provide it (Pearson 1986). Since deregulation, some have argued that the greater cost consciousness of airline managers will inevitably mean a lesser commitment to safety (Gray 1987; Lederer and Enders 1988). Some have also expressed concern that congestion in the airways, due to either reduced numbers of controllers or the slow rate of replacing the ATC system, will increase the probability of crashes (O'Brien 1988, OTA 1988). In Chapter 5 the trends in airline safety before and after deregulation are examined. The studies that have related carrier finances and mainte- nance practices are reviewed, along with other studies examining such effects as the role of pilot experience in accidents since deregulation. The performance of the FAA in providing ATC and in inspecting air carrier maintenance performance is also reviewed. Nonaccident safety indicators are examined to determine whether they can be used to examine changes in risk since deregulation. Airport and Airway Capacity Limits The hub-and-spoke service pattern, although beneficial to both airlines and passengers in many respects, has increased peak demands on airport runway and gate capacity because of the need to closely space arrivals and departures in order to minimize the time lost during connections. Some delay for air travelersâand perhaps more in coming yearsâis being caused by the physical limits of airports as they are being used in the deregulated environment. The remarkable growth in air travel during the last decade has not been matched by a similar growth in airport capacity. Although many of the nation's airports can handle current and expected peak travel demand, the
42 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION destinations of many travelers include major airports with capacity limits. Delays at these airports can cause ripple effects throughout the entire network. The FAA expects the number of congested airports to increase from 21 in 1989 to 33 by 1997; congestion is expected to increase at hub airports and airports used mainly for origin and destination travel, despite planned runway expansion at many of them (FAA 1989b). The new Denver airport, expected to become operational during the 1990s, is the only major commercial passenger airport to be opened since the early 1970s. The problems are not limited to runway capacity. Since deregulation, airway capacity has been slowly recovering from the controllers' strike of 1981. During recent years traffic growth throughout the system has begun saturating the airspace around some major airports. The new tech- nologies being acquired to replace the current ATC system are expected to ease some capacity problems, but the FAA will still have to develop the ability to manage demand as the airspace becomes increasingly sat- urated. In Chapter 6 the effect of limited airport and airway capacity on air passenger service is discussed, and options for expanding capacity are reviewed. Constraints on the Performance of the FAA The FAA is the federal agency responsible for certificating airlines, in- specting aircraft manufacturing and airline operations, and providing ATC. Although the regulatory authority of the FAA was not reduced with deregu- lation, its staffing levels fell substantially during the early 1980s. The initial strain was due to the firing of the air traffic controllers in 1981 and the slow process of hiring and training several thousand replacements. The FAA, however, allowed the number of technicians who maintain the ATC system to fall by attrition to below the necessary minimum and then found rebuilding difficult because of personnel ceilings imposed during the 1980s. The FAA had hoped that a massive overhaul of the ATC technologyâ in the form of the National Airspace System Plan (NAS Plan)âwould be phased in and fully in place by the year 2000. The productivity gains that the FAA expected would allow it to handle the demand for travel with fewer controllers and technicians (and less congestion and delay for passengers) have not occurred with the speed for which the FAA had hoped. The FAA also allowed the number of experienced inspectors to decline through attrition during the early years of deregulation at a time when the
Deregulation of Commercial Aviation 43 number of airlines and aircraft was increasing. Although the agency has been replacing air traffic controllers, technicians, and inspectors, a sub- stantial proportion of its personnel will become eligible to retire in the next few years. The adequacy of the FAA's staffing, which affects both safety and capacity, is reviewed in Chapter 7. Dissatisfaction with the FAA's performance mounted during the 1980s, and several groups and more than one comprehensive study began ad- vocating fundamental reforms. A summary and critique of these proposals is presented in a commissioned paper by Herbert N. Jasper, which appears as Appendix B. SUMMARY After more than a decade of experience with the deregulation of the commercial airline industry, the following five major issues of concern have emerged: The competitiveness of the industry (its effects on the fares and level of service provided to consumers today and the prospects of reduced competition from further industry concentration), The long-term financial stability of the industry, Possible discrimination against consumers of different types or in different parts of the country, The safety provided to the public by airlines and the FAA, and The ability of the federal government to respond to airport and airway capacity constraints. The findings of the study committee in each of these areas, along with its recommendations for improvement, are presented in Chapter 8. NOTES Levine's (1965) analysis of intrastate travel in California was based on a single intrastate market, which at the time had the largest daily passenger flows in the world. In small markets, the monopoly model continues to hold. That is, the passenger flows are so small that only a single carrier can cover costs and make a profit. These studies did, however, find economies of density, which refer to the ability of a carrier to offer additional flights without a proportional increase in cost when the expansion is limited to its existing network. These economies exist because
44 WINDS OP CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION certain costs (e.g., administration, advertising, passenger service, and terminal facilities) are relatively fixed. SIFL is an estimate developed by the CAB of appropriate fare levels based on carrier costs. CAB used SIFL in reviewing proposals for rate increases. The discussion of productivity in this paragraph is made in terms of unit costs, which, though related to productivity, are not the same. Productivity refers to the transformation of inputs into outputs, whereas unit production costs refer to the cost of inputs and outputs. These two concepts are basically the same over the short term, or when prices are held constant. For a discussion of the relationship between productivity and unit production costs, see work by Meyer and Oster (1981, 74-75). Dempsey (1990) also argues that the increased circuity caused by hub-and-spoke networks, which he suggests are higher "perhaps by as much as 30 percent," more than offsets any benefits of deregulation. The estimates of circuity for hub trips in Table 3-12, however, are on the order of 5 percent. REFERENCES ABBREVIATIONS CAB Civil Aeronautics Board DOT U.S. Department of Transportation FAA Federal Aviation Administration GAO General Accounting Office Bailey, E., and J. Panzar. 1981. The Contestability of Airline Markets during the Transition to Deregulation. Law and Contemporary Problems, Vol. 44, No. 125. Bailey, E., D. Graham, and D. Kaplan. 1985. Deregulating the Airlines. MIT Press, Cambridge, Mass. Bailey, E. E., and J. R. Williams. 1988. Sources of Economic Rent in the Deregulated Airline Industry. Journal of Law and Economics, Vol. 31, pp. 173-203. Baumol, W. J., J. C. Panzar, and R. W. Willig. 1982. Contestable Markets and the Theory of industrial Structure. Harcourt, Brace, Jovanovich, Inc., New York, N.Y. Borenstein, S. 1989. Hubs and High Fares: Airport Dominance and Market Power in the U.S. Airline Industry. Rand Journal of Economics, Vol. 20, No. 3, Autumn, pp. 344-365. Brenner, M. 1988. Airline DeregulationâA Case Study in Public Policy Failure. Transportation Law Journal, Vol. 16, pp. 180-199. CAB. 1973 ed. Handbook of Aviation Statistics. Caves, D., L. Christensen, and M. Tretheway. 1984. Economies of Density Versus Economies of Scale: Why Trunk and Local Service Airline Costs Differ. Rand Journal of Economics, Winter. Caves, R. 1962. Air Transport and its Regulators. Harvard University Press, Cam- bridge, Mass. Dempsey, P. 1990. Flying Blind: The Failure of Airline Deregulation. Economic Policy Institute, Washington, D.C. Derthick, M., and P. Quirk. 1985. The Politics of Deregulation. The Brookings Institution, Washington, D.C.
Deregulation of Commercial A via t ion 45 DOT (Research and Special Programs Administration). 1974-1990. Air Carrier Traffic Statistics. DOT. 1988. Stud)' of Airline Computer Reservation Systems. Report DOT-P-88-2. May. DOT. 1989. A Comparison of Air Fares and Services at St. Louis Airport Before and After Trans World Airlines Acquired Ozark Airlines. Report DOT P-37-89-3. DOT. I 990a. Secretary's Task Force on Competition in the U.S. Domestic Airline industry: industry and Route Structure. Feb. DOT. I 990b. Secretary's Task Force on Competition in the Domestic Airline industry: Pricing, Vol. II, Feb. Douglas, G., and J. Miller 111. 1974. Economic Regulation of Domestic Air Transport: Theory and Policy. The Brookings Institution, Washington, D.C. FAA. I 989a. FAA Aviation Forecasts Fiscal Years /989-2000. Report FAA-APO- 89-1. U.S. Department of Transportation. FAA. I 989b. Airport Capacity Enhancement Plan. Report DOT-TSC-FAA-89- I. U.S. Department of Transportation. Fromm, G. (ed.). 1965. Transport investment and Economic Development. The Brook- ings Institution, Washington, D.C. GAO. 1988. Airline Competition: Fare and Service Changes at St. Louis Since the TWA-Ozark Merger. RCED-88-2I713R. Washington, D.C., Sept. Gillen, D., T. Oum, and M. Tretheway. 1985. Airline Costs and Performance: implications for Public and industry Policies. Centre for Transportation Studies, University of British Columbia, Vancouver, Canada. Gomez-Ibanez, J., C. Oster, Jr., and D. Pickrell. 1983. Airline Deregulation: What's Behind the Recent Losses? Journal of Policy Analysis and Management, Vol. 3, No. I, pp. 74-89. Gray, R. 1987. Aviation Safety: Fact or Fiction. Technology Review, Aug./Sept. Jordan, W. 1991. U.S. Airline Performance from 1978 to 1988: The Bridge from Regulation to Deregulation. Journal of the Transportation Research Forum. Vol. 31, No. 2, pp. 317-325. Kahn, A. 1988. Airline Deregulation-A Mixed Bag, But a Clear Success Never- theless. Transportation Law Journal, Vol. 16, pp. 229-250. Keeler, T. 1972. Airline Regulation and Market Performance. Bell Journal of Eco- nomics. Autumn. Keyes, L. 1951. Federal Control of Entry into Air Transportation. Harvard University Press, Cambridge, Mass. Lederer, J., and J. Enders. 1988. Aviation Safety: The Global Conditions and Pros- pects. In Proc., Transportation Deregulation and Safety, Northwestern University, Evanston, Ill. Levine, M. E. 1965. Is Regulation Necessary? California Air Transportation and National Regulatory Policy. Yale Law Journal, Vol. 74, July, pp. 1416-1447. Levine, M. E. 1987. Airline Competition in Deregulated Markets: Theory, Firm Strategy and Public Policy. Yale Journal on Regulation. Vol. 4, Spring, pp. 393- 494. MacAvoy, P., and J. Snow (eds.). 1977. Regulation of Passenger Fares and Com- petition Among the Airlines. Ford Administration Papers on Regulatory Reform. American Enterprise Institute for Public Policy Research, Washington, D.C. Meyer, J., M. Peck, J. Stenason, and C. Zwick. 1959. The Economics of Competition in the Transportation industries. Harvard University Press, Cambridge, Mass. Meyer, J., and C. Oster (eds.). 1981. Airline Deregulation: The Earl)' Experience. Auburn House Publishing Co., Dover, Mass.
46 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION Meyer, J., and C. Oster. 1987. Deregulation and the Future of Intercity Passenger Travel. MIT Press, Cambridge, Mass. Morgan, I. 1981. Toward Deregulation. In Airline Deregulation: The Early Experience (J. Meyer and C. Oster, eds.), Auburn House Publishing Co., Dover, Mass. Morrison, S., and C. Winston. 1986. The Economic Effects of Airline Deregulation. The Brookings Institution, Washington, D.C. O'Brien. J. 1988. Deregulation and Safety: An Airline Pilot's Perspective. In Proc., Transportation Deregulation and Safely, The Transportation Center, Northwestern University, Evanston, III., pp. 353-388. Oster, C. 1981. Impact of Rising Fuel Prices. In Airline Deregulation: The Early Experience (J. Meyer and C. Oster, eds.), Auburn House Publishing Co., Dover, Mass., pp. 161-188. OTA. 1988. Safe Ski esfor Tomorrow: Aviation Safely in a Competitive Environment. Congress of the United States, July. Pearson, R. 1986. DeregulationâAn Insider's View. Midcontinent Perspectives. Mid- west Research Institute, Kansas City, Mo. Wilbur Smith Associates. 1989. The Economic Impact of Civil Aviation on the U.S. Economy. Partnership for Improved Air Travel, Washington, D.C.
PART II Effects and Responses in the Private Domain
2 Changes in Airline Operations and Financial Performance Deregulation began a transformation in air carrier asset management, labor relations, corporate organization, and airline management that is still occurring. Whereas in some cases air carriers shifted fairly quickly to the competitive environment engendered by deregulation, in other cases the residual effects of the previous regulatory regime, manifested in route structures and aircraft fleets or an inability to adapt to market conditions, have placed some carriers in peril. The main point of this chapter, de- scribed in the first section, is that deregulation has allowed carrier man- agers to begin allocating assets more efficiently, control costs, and price services according to consumer demands. Before turning to the assessment in Chapter 3 of how these changes have affected consumers, however, it is necessary to examine how increased competition has affected air car- riers' financial condition. As is revealed in the second section, the financial position of a few of the major carriers has weakened considerably, but these carriers also entered deregulation in a vulnerable position. The fi- nancial performance of other major carriers appears to be consistent with their prederegulation performance. A variety of influences is affecting all air carriers, however, which raises legitimate concerns about the long-run financial strength of more than just the weaker carriers. AIR CARRIER MANAGEMENT AND OPERATIONS Managing air carriers during the regulated era required a different set of skills than those most in demand since deregulation because of the control 49
50 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION that the Civil Aeronautics Board (CAB) exercised over routes, prices, and equipment. By the 1970s the CAB had effectively stopped granting new routes to the largest trunk carriers; hence they were restricted to serving their existing routes. Also, for any carrier, awards of additional routes required a lengthy and expensive regulatory process, with no guarantee of success. Trunk or local service airlines could compete on price to win market share, but only within a fairly limited sphere. The CAB also exercised considerable influence over decisions about the acquisition of aircraft (Morgan 1981). In this environment managers needed to be ex- perienced at operating within the confines of CAB regulations, if not adept at lobbying to change them. Many airline managers were indeed quite effective, but skills in marketing and cost control were less important than those in law and politics (Meyer and Oster 1987). Deregulation gave managers the ability to deploy assets and to price ser- vices according to market demand, a freedom exercised daily by managers throughout the rest of the economy. Most top airline executives, however, most of whom had staunchly resisted deregulation, were not prepared for the freedom given to them, nor were they particularly adept at exercising it. Some of the early strategic moves by carriers such as Pan American and Eastern, for example, were ineffective and failed to make them cost-competitive or to offer a sharply differentiated product (Meyer and Oster 1987). Because the CAB had protected carriers from failing, managers were also unaccus- tomed to taking risks that could result in the failure of the firm. Braniff, for example, expanded far too aggressively and was sent into bankruptcy (for the first time in 1982) by the first major downturn in the economy. Throughout deregulation, top and middle managers who remained from the regulated era have either been trained on the job or replaced by man- agers and owners more prepared for marketplace competition. Not all new managers, or new entrant entrepreneurs for that matter, have been suc- cessful. Some carriers have been significantly weakened (though carriers such as Pan American were weak before deregulation); some have failed (Eastern); and some have become stronger. In the main, however, dereg- ulation ushered in a new era of management. Some management inno- vations developed or expanded during deregulation have been successful at increasing productivity and controlling costs. Route Strategies The freedom to enter markets was the most profound regulatory reform, and carriers moved rapidly into many new markets. Whereas at the outset
Changes in Airline Operations and Financial Performance51 of deregulation the CAB had granted route authority in 24,000 city-pair markets, the number had increased to 106,000 within 18 months after passage of the 1978 legislation (Bailey et al. 1985). Although airline managers experimented initially with different strat- egies, most fairly quickly realigned their routes into hub-and-spoke net- works, which have become one of the most prominent features of air service since deregulation. Some carriers had long recognized the benefits of hubbing; Delta and Eastern'had used Atlanta as a hub since the 1960s and formerly local service airlines such as Piedmont and Allegheny had fairly well-developed systems. Because the CAB rarely granted new route authority, however, the options for building such a network had been limited. Indeed, under regulation, trunk carriers had sought to obtain route authority in long-haul, city-pair markets because the fares allowed in those markets were the most profitable. As discussed later, this led to an over- investment in aircraft that were most efficiently deployed in those markets but unsuited to hub-and-spoke networks. Hub-and-spoke networks are appealing to air carriers for several reasons. Most noteworthy, they allow carriers to provide service to a larger number of city pairs without a commensurate increase in cost than do point-to-point networks. For example, a carrier needs a minimum of 10 flights to serve 10 city pairs in a point-to-point route system. If operated through a hub, however, those same 10 flights can serve as many as 100 city pairs.1 Also, by concentrating the flow of passengers toward a central point, hubs make possible service between city pairs that do not have sufficient passenger flows to support nonstop service in a point-to-point system (Wheeler 1989; Brenner 1990). Passenger demand for nonstop service also gives the carrier an opportunity to charge higher-than-average fares on a route that it monopolizes (al- though as discussed in Chapter 3 such routes account for a small share of total trips) (Wheeler 1989). Finally, as a result of the extensive networks made possible by hubbing, carriers are also able to attract passengers and, with tight scheduling, are able to meet passengers' preference for single-carrier service (Wheeler 1989). This also gives the hub carrier a marketing advantage.2 The freedom to enter new markets also allowed carriers to adjust their route systems to balance their traffic flows. Eastern, for example, had mostly north-south traffic, which is highly influenced by seasonal demand, but was able to obtain greater year-round equipment utilization by entering east-west markets. Similarly, United, which operated primarily in east- west markets, added more Sun Belt cities to take advantage of increased demand for travel to those cities during the winter (Bailey et al. 1985).
52 WINOs OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION Carriers also withdrew from small markets that were unprofitable to serve with the jet aircraft that had become the industry's perceived standard during the 1960s and 1970s. These moves were hastened by carrier efforts to align service with cost as they eliminated the internal cross subsidies for small community service that had been fostered by the CAB's fare policy. Additional incentives to abandon routes with few passengers were provided by the fuel crisis of 1979, which further made jet service un- economical in small markets (Oster 1981). As discussed in more detail in Chapter 3, the withdrawal of jet service from routes with few passengers was usually replaced by service by one of the fast-growing commuter airlines. Commuters offered service in smaller, propeller-driven aircraft, the operating costs of which were better suited to small markets. Asset Management Giving managers more flexibility in the use of assets showed early benefits. Average daily use of aircraft during the mid-1970s was about 8.3 hours a day; this increased almost immediately after deregulation by about an hour a day (Meyer and Oster 1981; Bailey et al. 1985). Carriers also increased efficiency by increasing the number of seats available on some aircraft. Seating densities had been increasing on some aircraft before deregulation, but on some models, such as the 727-200's that proved attractive for hub-and-spoke feeder traffic, carriers actually reduced seating density during the early 1970s (Meyer and Oster 1981, 78). With planes flying only half-full and with an emphasis on service quality, managers had little incentive to use aircraft more efficiently. After the beginning of administrative deregulation, seating densities on the most widely used aircraft increased at a faster rate than previously (Meyer and Oster 1981, 78). (Subsequent efforts to increase aircraft utilization were influenced as well by the fuel price increases caused by the fuel crisis of 1979.) As an example of efforts to improve effièiency, carrier managements added more rows of seats to some narrow-bodied aircraft (727-100's and 200's and DC-9-30's). They also reduced the size of seats on such wide- bodied aircraft as L-l0l l's, DC-lO's, and 747's, in most cases to the size used on narrow-bodied aircraft. Seats that were typically 22 inches wide before deregulation were reduced to about 19 inches, but some carriers used even narrower seats on some aircraft (Blyskal and Hodge 1989; Yom 1990). Simply increasing seat-mile output would have been of little use if managers had not also become more adept at marketing strategies to
Changes in Airline Operations and Financial Performance 53 help fill the seats; fortunately, they became more adept at both (see Figure 1-3 and the following discussion on marketing). The shift to hub-and-spoke networks, along with the fuel crisis of 1979, placed a premium on relatively short-haul, fuel-efficient aircraft. For ex- ample, twin-engine, narrow-bodied aircraft accounted for less than one- fourth of total aircraft use by commercial air carriers in 1980, but by 1990 this share was about half (FAA 1989). Because CAB fare regulation rewarded trunk carriers in long-haul markets and because of the low seat- mile cost of wide-bodies, carriers had an incentive to purchase these aircraft for long-haul service. The realignment of route strategies into hub- and-spoke networks during deregulation found some carriers with an ex- cess supply of wide-bodied aircraft.3 Some became available to new entrant airlines at bargain prices, which may have facilitated some of the intense fare wars in coast-to-coast service during the early 1980s. In addition, large shares of the fleets of some carriers, such as American and Trans World, were rendered uneconomic by the 1979 fuel crisis (Oster 1981). Hence a good deal of fleet adjustment has been required to shed fuel- inefficient equipment and acquire new aircraft more suited to hub-and- spoke service. These adjustments by airline managers were facilitated by the increased ease in selecting and deploying assets after deregulation and by incentives to acquire and use more productive aircraft, but they did not occur all at once. Aircraft are expensive assets with long service lives. As a result, fleet adjustments have occurred during the entire deregulation period. Marketing Innovations Airline managers quickly learned the importance of marketing to attract passengers, and, since deregulation, marketing has taken on several key features: sophisticated discounting practices, closer relationships with travel agents through extensive computer reservation systems (CRSs), extra com- missions for agents, travel awards to frequent fliers, and closer affiliations with smaller carriers to allow them to market a more extensive service network. The use of discount airfares had been growing before deregulation, but the pricing strategies employed pale in comparison with both the availa- bility and the complexity of discounts in recent years. In the early years of deregulation the incumbent air carriers matched the low fare offerings of the new jet entrants in across-the-board discount fares, otherwise they were hard-pressed to compete with lower-cost, often nonunion carriers.
54 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION Many observers have concluded that the fare wars of the early 1980s got out of hand and resulted in below-cost pricing (Meyer and Oster 1987). With time, however, carriers developed yield management systems that, aided by the growing sophistication of CRSs, helped air carriers target discount seats that otherwise would have gone unsold. This practice al- lowed them to match the claims of new entrants in advertising without having to discount seats they would have sold at a higher fare.4 The growing complexity of fares elevated the role of travel agents and the importance of computerized systems to track the frequent changes in service and fares. In the mid-1970s travel agents began promoting the concept of a single computerized system to integrate the offerings of all carriers, and several carriers banded with the travel-agent industry in 1978 to promote a similar concept (Clippinger and Strong 1987). These efforts failed for a variety of reasons, and individual carriers began promoting their own systems. The systems of United and American quickly emerged as the front-runners, and by 1981 the two airlines claimed affiliations with nearly 80 percent of the automated travel agencies. This share has changed but little during a period in which almost all travel agents have acquired access to a CRS (Clippinger and Strong 1987; DOT 1990). CRSs, whose development required substantial investments of capital, have greatly increased the efficiency of distribution of a complex and frequently changing array of information. These systems have also made accessible to virtually all travel agents the offerings and fares of almost all airlinesâfrom the largest incumbent to the smallest new entrant. Despite their contribution to the efficiency of a market dependent on information, CRSs have been controversial, in the main because a small number of airlines own the information distribution systems on which their competitors depend for the display of their flights and fares. By training travel agents in the use of these systems and by leasing their equipment on favorable (and exclusive) terms, air carriers have de- veloped closer relationships with travel agents than had existed in the past. Carrier managements have also tried to influence agent bookings by of- fering extra incentives for sales greater than a certain amount; these in- centives are referred to as "commission overrides." Frequent flier programs have been perhaps the airlines' most successful marketing tool. When American, under a new management team with extensive experience in marketing, offered its frequent flier program in 1981, the other carriers dismissed it as a gimmick (Levine 1987). The incumbent carriers, who had higher labor costs than the new entrants, soon recognized the importance of retaining business travelers, who are less willing to take advantage of the deepest discount fares, but who might
Changes in Airline Operations and Financial Performance 55 opt for the low fares offered by their competitors. The importance of offering frequent flier benefits is heightened by the fact that roughly 5 to 6 percent of fliers account for about 40 percent of all trips taken annually. The use of frequent flier programs by all but one jet carrier is testimony to their importance in building customer loyalty, especially from business travelers . As carriers began to develop extensive hub-and-spoke networks, they began to rely more heavily on commuter airlines to feed traffic from small communities into their hubs. Commuter airlines are able to serve markets more inexpensively than the large, jet carriers. Their main cost advantage results from their use of smaller, prop or turboprop aircraft more suited to short haul, low density markets and from higher aircraft utilization. To a lesser extent, commuter carriers have had a cost advantage because of lower labor costs: their management structure is more streamlined; their workforce has less seniority; fewer are unionized; and they have fewer work rules (Meyer and Oster 1984, 64, 77, 210). Affiliations between carriers had always required the coordination of schedules and arrange- ments for handling joint fares and baggage transfers. The existence of such affiliations between large and small carriers predates deregulation, but large carriers soon learned the value of marketing these affiliations by allowing the commuter carrier to share their two-letter identifier (code) in CRSs. This practice both communicated to travel agents nationwide the existence of connections and provided hub carriers with more influence over the incoming traffic. This has developed into a continuum of rela- tionships, from joint operations (code sharing) to outright ownership. Labor Costs and Relations The competition fostered by deregulation dramatically changed labor re- lations in commercial aviation, a change that is still evolving. Before deregulation, labor costs were increasing sharply, but because of the reg- ulated system of protected markets, such costs could be passed on more easily to consumers. Many new-entrant carriers offered services with non- union labor and were encumbered by few of the work rules that had developed in the unionized firms. After deregulation the managers of the previously established carriers looked to labor, which accounted for roughly 40 percent of total costs, for cost savings in order to remain competitive with low-cost new entrants. Although many workers resisted the efforts of managers to control labor costs, virtually all labor groups ultimately agreed to compensation or work-rule changes or both. These concessions
56 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION had an immediate effect on labor costs. From 1972 to 1978 labor costs for major carriers increased at an average annual rate 3.8 percent greater than the consumer price index (CPI); this rate fell to 1 percent below the CPI from 1979 to 1982 (Meyer and Oster 1987, 97). Although labor costs are increasing in the airline industry, the increases have been at a slower rate than the CPI since deregulation. Although total employment has grown with the expansion of the industry and productivity has improved, labor relations at some carriers remain contentious, as witnessed by the strike in 1989 that pushed Eastern into bankruptcy and ultimately failure and by the repeated attempts of the unions of United and, to a lesser extent, Trans World, to gain control over their companies. Productivity Productivity benefits since deregulation have occurred for a variety of reasons reviewed previously, including greater use of capital, use of more fuel-efficient aircraft, moderation in labor cost increases, and service and marketing strategies to attract passengers and help fill aircraft. The general trends in productivity gains were discussed earlier (Figures 1-1-1-4). The cost of producing total output (available seat miles) has increased more slowly than input costs, and load factors have shifted to a significantly higher level, where they have remained.6 Mergers and Acquisitions The ease of entry under deregulation and the lack of production scale economies in air transportation would appear to make mergers unnecessary and inefficient. Nevertheless, several carriers pursued mergers or acqui- sitions during the early years of deregulation as expedients to extend networks and acquire aircraft, skilled personnel, and airport access quickly. For example, Pan American merged with National because as a formerly international carrier under regulation, Pan American did not have a do- mestic route network or hub airport, advantages that it assumed it would obtain by acquiring National (Meyer and Oster 1987). Republic Airlines was formed by the merger of three regional airlines to create a nationwide system. These strategic moves to build nationwide networks, however, did not prove competitive. Pan American and Republic had difficulty merging unionized labor forces with strong adherence to seniority systems,
Changes in Airline Operations and Financial Performance 57 and in neither case did management stake out dominant positions at major hubs. After the initial wave of mergers and acquisitions, many of the subsequent combinations seemed as much motivated by hopes either to achieve economies of scope or scale or to gain greater control over markets (Levine 1987; Meyer and Oster 1987). Whether some of the merged companies will emerge as strong com- petitors is uncertain. The difficulties inherent in merging work forces with seniority lists and in shedding inefficient operations have financially weak- ened several carriers. In addition, the increased indebtedness required to finance some mergers and acquisitions makes some carriers vulnerable in any major downturn of the economy. The trends in the financial posture of the industry and what they portend for the future are discussed next. INDUSTRY FINANCIAL PERFORMANCE The apparent financial performance of the major air carriers has been sufficiently weak since deregulation to raise questions about the long-run health of the industry. These concerns have focused on three areas: the thin profit margins of the industry, the increase in industry leverage, and the airlines' low return on equity capital. Of particular concern is what the airlines' financial condition implies for their ability to withstand periods of recession and to attract capital for long-term investment in replacement and expansion of their current fleets. The financial condition of the airline industry during the last 35 years is examined. Changes in the industry since deregulation are noted, and airline performance in the pre- and postderegulation eras is compared. Concern over the financial condition of the airline industry has increased and been validated for some by the carrier bankruptcies that occurred in 1990 and 1991. For the industry as a whole, losses in 1990 were the worst ever experienced by the industry up to that point. The current financial turmoil in the industry has been caused in part by escalating fuel costs, recession-induced weakening of domestic traffic, and a drop-off in inter- national traffic because of fears of terrorism as a result of the Persian Gulf war. The results in 1990 and early 1991 seem to bear out predictions that several weak carriers could not resist a downturn in demand or rapid changes in production costs. The financial position of the industry as a whole has weakened some- what during the last decade, although it was hardly robust before deregulation. Nevertheless, this decline is not nearly so severe as might be suspected from observing only a few measures of financial perfor-
58 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION mance or from looking at only a few troubled carriers. A comparison of pre- and postderegulation financial performance shows that industry operating margins have weakened and returns on equity have declined. Although some measures are not significantly weaker than in the pre- deregulation period, especially when account is taken of the controllers' strike and periods of recession, they have been significantly affected by the capital restructuring that took place in the industry during the 1980s. These measures include the industry's debt/equity ratio and return on investment (ROl). The low return on equity (ROE) and the increase in industry leverage are perhaps the main sources of concern for the industry's financial health, especially their implications for the structure of future long-term invest- ment. The return to capital in the airline industry has generally been low and volatile, being strongly cyclical with the business cycle. It is unlikely that such a low return could be sustained indefinitely. Furthermore, it could cause a further increase in the reliance on debt financing that could make the industry continually vulnerable to recessions or sustained periods of flat demand. On the other hand, the industry average performance in the postdereg- ulation period was significantly affected by the weak performance of a few struggling carriers. Even in this period of generally weak financial results, many of the major carriers earned significantly higher profits than the industry average and produced better returns on invested capital. To the extent that the failure of Eastern and the bankruptcies of Pan American and Continental lead to a restructuring of the industry, overall industry results are likely to improve significantly. Industry performance since deregulation makes it difficult to conclude that carriers have been profiting from overall fares that are substantially higher than competitive levels. This says nothing about pricing variations between markets and classes of passengers, however, only pricing and revenue on an industrywide basis. What air carrier financial performance since deregulation implies about future performance is uncertain. Financial results have been significantly affected by the nearly continuous structural change and expansion in the industry during the last decade. The realignment and consolidation that have occurred as a result of the 1990-1991 recession will likely reduce pressure on fares, revenue, and profitability from overcapacity in the industry. Whether this will lead to a more stable industry will depend on the pricing behavior of the surviving major firms. Outcomes may range from a tight oligopoly with excess profits and reduced service to an ag- gressively competitive industry in which the remaining firms fight for
Changes in Airline Operations and Financial Performance 59 market share throughout their networks, with profits and returns on in- vested capital remaining weak. Analysis Methodology Limitations of the Financial Analysis It is important to establish the limitations of this financial analysis before further examination of the industry's financial results. A general problem of financial analysis is that the ratios and key financial variables examined have different meanings and interpretations for different firms, industries, and periods in the economic cycle. Because of this and the vagaries of the data available, the actual state of the industry's and individual carriers' results may be difficult to establish and interpret without substantial qual- ification of the data. Furthermore, all financial managers have strong incentives to present their companies in the most favorable light to the extent that accounting and financial practices permit. For instance, the incentive under regulation may have been for the airlines to overstate their assets and liabilities in order to increase permissible fares and their actual returns. In a deregulated environment, the incentive may be the opposite (though individual cir- cumstances always vary), that is, to understate assets and thereby improve stated returns to investors. In spite of the above problems, it is still worth looking at the book returns of the airlines as shown in their Form 41 filings with the U.S. Department of Transportation (DOT) (previously with the CAB). Absent the alternative of an in-depth historical, financial analysis of each airline, this is the best available means of comparing the industry's performance over time. Where appropriate, considerations of market valuation and risk are addressed. The advent of deregulation changed the basic rules of the game for air carriers. As noted previously, deregulation increased the demands on management for marketing skills, strategic planning, cost control, and competition with other firms. Deregulation stemmed in part from a belief that airlines, like other firms, should earn their survival and growth in the market and not look to a public body or policy to guarantee it. It might then be appropriate to ask why, in a deregulated environment, one should be concerned with the financial condition of the airlines.
60 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION The financial condition of the industry directly affects individual firms' behavior in the short run and, ultimately, their structure and performance in the long run. In the short run, failing firms may resort to less-than- compensatory fares in order to generate sufficient cash to cover their fixed, short-run commitments, but not their long-run costs. This may threaten, in turn, the profitability and survival of other carriers in the long run. Although passengers benefit from low fares in the short run, for well-run carriers to survive, fares must be raised eventually to recoup losses and provide a sufficient return to keep capital in the industry. Furthermore, the disruption to air service caused by the failure of a particular carrier imposes real costs on passengers, both business and pleasure travelers. In the long run, whereas the survival of any particular firm is not important on a national policy basis, the failure of a significant number may lead to increased concentration and too few firms in the industry. The minimum number of firms necessary to ensure a competitive industry has been widely debated. As important as the number, however, is the degree to which the existing airlines serve the same markets and how vigorously they compete on price and service. A small number of nationwide firms that compete with each other at all the large commercial airports may provide much stronger competitive pressure to hold down costs and fares than a large number of carriers competing in less extensive networks. The fewer the number of firms, however, the easier it is for them to form and enforce a tight oligopoly in which industry output is lower and fares are higher than would be the case in a competitive market. This issue is taken up again in Chapter 4. Definitions and Organization of the Analysis In discussions of the pre- and postderegulation periods, the following time periods are generally used. For comparisons of the financial performance of the industry in this period, data from 1955 to 1989 have been collected.7 The starting point predates the jet era and includes most of the significant industry expansion and investment periods. The postderegulation period is defined as 1979-1989, and the prederegulation period is 1955-1978. The prederegulation period is subdivided into the periods 1955-1967 and 1968-1978. The latter subdivision picks up after the initial boom period in the 1960s from the introduction of jet aircraft and is equal in length (11 years) to the postderegulation period. An alternative breakdown of the pre- and postderegulation periods that is often proposed incorporates the years of significant administrative deregulation (1976-1978), before the passage of the Airline Deregulation Act in October 1978. Accordingly,
Changes in Airline Operations and Financial Performance 61 financial averages are calculated for 1955-1975 and 1976-1989 to capture the effects of this administrative deregulation. The effects of recession years must be accounted for in any comparison of results over time, particularly for an industry that is as strongly affected by the business cycle as is the airline industry. Recessions by quarter occurred as follows: the fourth quarter of 1957 through the second quarter of 1958 (1957:4-1958:2), 1960:2-1960:4, 1969:4-1970:3, 1973:4-1974:4, 1980:1-1980:2, and 198 1:3-1982:3. The effects of these recessions on airline performance tended to lag the recessions somewhat, so the recession effects on an annual basis appear in 1958, 1961, 1970, 1975, and 1981- 1983.8 When nonrecession data are reported, they are calculated by ex- cluding these years. Table 2-1 includes industry averages for all of the ratios and financial measures discussed in this chapter for each time period. Furthermore, they are reported with and without the recession years in each period. All of these breakdowns are reported to allay any concern that the choice of endpoints has dramatically altered the results. The only data that will be extensively used in this analysis, however, are the pre- and postderegu- lation averages and the 1968-1978 averages, all without the recession years. Finally, industry data and graphs will show results for all certificated air carriers before 1981 and for major, national, and large regional carriers from 1981 to 1989. The operating ratios in Table 2-1 are defined in the following way. The income margin is operating income as a percentage of operating revenue. It is sometimes referred to as the gross profit margin. The profit margin, following work by Brenner (1988), is operating income less interest ex- penses as a percentage of operating revenue. Net profit margin has its usual meaning of net income after taxes as a percentage of operating revenue. Although the greatest effects of changes in air carriers' financial struc- ture during the last decade have been on capital measures, the changes have also affected some of the operating ratios. The effects on capital are discussed later. The principal effects on operating ratios are on the profit and net profit margins because of the effect on interest expenses. To the extent that carriers have shifted debt and interest expense to holding com- panies and away from the operating company results, the profit and net profit margins overstate carriers' returns because both are calculated after interest expenses are deducted. The variability of individual carrier results, especially since deregula- tion, may make it tempting to ignore industry averages. Such averages may conceal a wide or skewed distribution in which some carriers are
62 WiNos OP CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION TABLE 2-1 AVERAGE FINANCIAL MEASURES FOR AIR CARRIER INDUSTRY FOR VARIOUS PERIODS Legislative Administrative Deregulation Deregulation Financial 1979- 1955- 1955- 1968- 1976- 1955- Measure 1989 1978 1967 1978 1989 1975 Excluding Recession Years" Operating results Income margin 2.8 6.6 8.0 4.9 3.4 6.9 Profit margin -0.3 4.2 6.0 2.1 0.5 4.5 Net profit margin 1.0 3.3 4.0 2.4 1.8 3.1 Debt and equity Debt/equity ratio 1.24 1.26 1.21 1.31 1.16 1.31 Return on equity (ROE) 4.4 9.3 10.6 7.6 7.0 8.5 ROE, all manufacturingb 13.0 12.0 11.4 12.7 13.4 11.6 Return on investment 8.0 7.4 7.5 7.2 8.8 6.7 All Years Operating results Income margin 1.9 5.8 7.2 4.1 2.5 5.9 Profit margin -1.5 3.4 5.1 1.3 -0.6 3.5 Net profit margin 0.4 2.7 3.4 1.7 1.2 2.5 Debt and equity Debt/equity ratio 1.34 1.30 1.24 1.37 1.26 1.35 Return on equity (ROE) 1.6 7.5 9.1 5.5 4.2 6.5 ROE, all manufacturingb 12.5 11.6 11.0 12.3 12.9 11.2 Return on investment 7.2 6.6 6.9 6.3 8.0 6.0 NOTE: Data (expressed as percentages except for the debt/equity ratio) are for total certificated air carriers through 1980 and for major, national, and large regional carriers thereafter. "As used here, recession years are 1958, 1961, 1970, 1975, 1981, 1982, and 1983. bROE averages for all manufacturing industries do not include 1989. Therefore, averages listed under 1979-1989 and 1976-1989 are only for the periods 1979-1988 and 1976-1988, respec- tively. SOURCE: Data from Handbook of Airline Statistics, CAB, and from Air Carrier Financial Sta- tistics, CAB and DOT. doing significantly better than the average, whereas others are doing sig- nificantly worse. The financial condition of all carriers, however, is af- fected by and affects the pricing and service behavior that is examined in this analysis. Although it is the weak carriers that, to an extent, set industry prices, the service and fare results since deregulation cannot be viewed in isolation from the overall industry financial performance.9 Nevertheless,
Changes in Airline Operations and Financial Performance 63 the results for the major carriers during the last 22 years are examined in addition to the industry averages to illustrate the variability of carrier financial performance. The analysis focuses on the largest carriers that existed before dereg- ulation and how they have fared afterwards. These carriers are grouped as weak or strong carriers. The carriers that are currently having financial difficulties (Eastern in liquidation, Pan American and Continental reor- ganizing under Chapter 11 bankruptcy, and Trans World deferring debt payments) are considered the relatively weak group of carriers. The strong group includes American, Delta, Northwest, United, USAir, and Piedmont (which merged with USAir in 1989). In 1990, two fast-growing air car- riersâSouthwest Airlines and America Westâwere reclassified as major carriers (their share of total revenue passenger miles, however, remains quite smallâabout 5 to 6 percent). They are not part of the financial analysis but may play important roles in some markets in the future. A brief description of each follows. Southwest Airlines started as a small, intrastate carrier. In 1972 it served three Texas cities with just 4 aircraft, but by 1990 its network had grown to 27 airports served with 94 aircraft (Aviation Week and Space Technology 1990a, 82). Southwest has consistently pursued a fare and service strategy at variance with the rest of the industry. It does not operate a hub-and- spoke network, participate in CRSs, serve in-flight meals, accept reser- vations, or serve long-haul markets. Nonetheless, it has been among the most profitable carriers, if not the most profitable carrier, in recent years. When other major carriers were absorbing their largest annual losses in 1990, Southwest posted a profit. The carrier has succeeded by identifying specific market segments with a demand for point-to-point, low-cost jet service and providing that service at a price that other carriers with higher cost structures cannot match. America West started as a niche carrier in 1983 with a focus on markets that could be served from its Phoenix and Las Vegas hubs. In contrast with Southwest, America West expanded rapidly into markets in the Mid- west and on the East Coast, where it competes directly with other major carriers. It has been among the most highly leveraged carriers. In 1990 America West began another expansion program in which it announced plans to expand its 100 aircraft fleet by an additional 74 aircraft (Aviation Week and Space Technology 1990b, 30.) These plans ran afoul of the recession and fuel price shocks of the Persian Gulf war, and in June 1991, America West filed for Chapter 11 bankruptcy. Although its bankruptcy was precipitated by a unique confluence of events, America West's attempt
64 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION to expand and take on the major airlines in markets other than the niches where it was most competitive has not proved to be a successful strategy for any new-entrant airline. Each of the following sections begins with a discussion of the relevant financial measures for the industry as a whole, followed by a discussion of those measures for individual carriers. The specific limitations of the available data and the analysis of the financial results are addressed in each section as well. Operating Results Airline traffic is strongly affected by general economic conditions; con- sequently operating results for the industry have always been cyclical. Even during nonrecession periods, airlines have historically had thin op- erating margins, which is perhaps not surprising given the nature of airline output. When the effects of recessions are accounted for, the, difference between the pre- and postderegulation periods is much less than has often been cited. Nevertheless, since deregulation, there has been a decided downward shift in industry operating margins. Care must be taken in using operating results to assess the financial conditions and prospects for any industry. The peculiar nature of an in- dustry's structure, output, and technology significantly affects operating margins. As a result, comparisons between industries and sectors of the economy on the basis of operating results alone can be meaningless or even misleading. Nevertheless, comparisons among firms within an in- dustry or for an entire industry over time can be quite useful. Furthermore, when used in conjunction with other financial indicators, operating results can provide important information about the contribution of current ac- tivities to the financial health of the firm or industry. The income and profit margins for the industry, both pre- and post- deregulation, are shown in Table 2-1. Because the results for the net profit margin discussed next tell much the same story as these measures, they are not further examined. However, because other analyses have used them for various comparisons, they are presented for reference. The net profit margin, shown in Figure 2-1, represents the broadest picture of industry results compared with operations because it shows the returns available to equity in relation to industry sales. The average net profit margin for the postderegulation period was 1 percent, compared with 2.4 percent during 1968-1978 and 3.3 percent during 1955-1978.
Changes in Airline Operations and Financial Performance 65 Percent 8i -2 1955 1960 1965 1970 1975 1980 1985 1989 Year FIGURE 2-1 Net profit margin for total certificated air carriers. Although the profit from each dollar of revenue is down, the downward trend began before deregulation. Of particular importance for the airline industry and readily apparent in Figure 2-1 is the cyclical nature of the net profit margin. The business cycle has a great impact on the profitability of airline service. Industry profits can evaporate quickly in a downturn, as traffic and revenues suffer. Sometimes comparisons of net profit margins are made between in- dustries. This is an inappropriate use of the net profit margin because the nature and importance of each industry's earnings per dollar of sales differ. For example, high-volume, less capital-intensive industries, such as gro- cery retail stores, have much smaller operating margins than low-volume, more capital-intensive industries. Such inappropriate comparisons have been purposely omitted in this report. Whereas operating results for the industry as a whole have declined since deregulation, the performance of the relatively strong major air carriers has more nearly matched their performance during the predereg- ulation period, though the trend is down slightly for them as well. Table 2-2 presents the results for the two groups of major carriers during the last 20 years. As is apparent from the data, the industry income, profit, and net profit margins have all been significantly affected by the results of the weaker carriers that have been struggling in the postderegulation period. The
66 WINoS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION TABLE 2-2 AVERAGE FINANCIAL MEASURES FOR MAJOR AIR CARRIERS Financial Measure Excluding Recession Years 1968-1978 1979-1989 Including Recession Years" 1968-1978 1979-1989 Operating results Income margin Strong 6.6 4.3 5.8 3.4 Weak 2.4 -0.7 1.7 -1.6 Total 4.7 2.7 3.9 1.7 Profit margin Strong 4.6 2.4 3.8 1.3 Weak -0.9 -5.6 -1.7 -6.3 Total 2.1 -0.2 1.3 -1.3 Net profit margin Strong 3.8 3.1 3.2 2.6 Weak 1.0 -2.9 0.2 -3.1 Total 2.5 1.1 1.8 0.6 Debt and equity Debt/equity ratio Strong 0.88 0.64 0.91 0.72 Weak' 1.96 3.45 2.09 3.28 Total 1.27 1.17 1.33 1.25 Return on equity Strong 9.7 9.8 8.2 8.4 Weakb 4.2 -9.9 0.8 -14.5 Total 7.6 7.8 5.5 5.0 Return on investment Strong 8.3 9.8 7.4 9.3 Weak 5.9 3.7 4.9 3.1 Total 7.1 7.9 6.2 7.2 NOTE: Strong carriers are American, Delta, Northwest, United, USAir, and Piedmont. Weak carriers are Continental, Eastern, Pan American, and Trans World. Data are expressed as per- centages. except for the debtIequity ratios. "As used here, recession years are 1970, 1975, 1981, 1982, and 1983. bFOr the weak carriers, the 1979-1989 period only includes the years 1979-1987 because none of these carriers had significant book equity in 1988 and 1989. SOURCE: Data from Air Carrier Financial Statistics, CAB and DOT. margins for the strong carriers have moreS nearly approached their pre- deregulation levels. Furthermore, the decline in all the margins has been less severe for these carriers than for the weaker ones. The downward trend of the industry average for the net profit margin in the last 20 years stems principally from the weak carriers (Figure 2-2). These results point out the thin nature and variability of the operating margins in the industry at its current scale and technology of production. Furthermore, they indicate that the large expansion in output since de-
Changes in Airline Operations and Financial Performance 67 Percent 10 -5 -10 15L 68 70 72 74 76 78 80 82 84 86 88 Year - Strong - Weak FIGURE 2-2 Net profit margins for strong and weak caniers. regulation has been accompanied by a less-than-proportionate rise in prof- its for the industry as a whole. Although a decided shift downward in operating margins has occurred, the performance of the strong air carriers has tended to remain more stable between periods. The foregoing discussion is by no means meant to attribute the decline in operating margins since deregulation to deregulation itself. The suc- cessive periods of recession, the fuel price run-up, and the Professional Air Traffic Controllers Organization strike all significantly disrupted in- dustry profitability. In fact, absent deregulation, the decline in this period might have been significantly greater)° To attribute cause and effect would require a much more extensive analysis. Capital Structure Changes in the capital structure of the airline industry since deregulation create some difficulties in making historical comparisons. The following analyses take note of these difficulties and indicate where the value of the ratios and industry results examined have been affected. The most important of these effects are changes in the sources of financing for air- craft and in the valuation and ownership of the airlines. The discussion covers first the effects of lease financing on the capital structure of the airlines and the problems associated with measuring long-term debt. Next the industry's leverage and returns on equity and investment during the
68 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION pre- and postderegulation periods are examined. Last, an examination is made of the industry's betas, which are market measures of industry financial risk. Leasing and Long-Term Debt Leasing consists of both operating leases and capital or financial leases. Operating leases are short term (generally not more than 5 years) and have varying degrees of flexibility for cancellation by the airline. They generally convey no residual value in the aircraft and are considered strictly as an operating cost from an accounting standpoint. Financial leases are long term, generally 12 to 25 years. Because of restrictions on termination, their long term, and the contractual commitment to pay the total value of the lease payments, they are considered a form of capital financing. The total cost of the lease payments is amortized over the life of the lease, and a portion of the rental payments is attributed to the implicit interest cost of the financing. Both forms of aircraft financing have become important since the mid- 1970s, roughly during the deregulation period. From virtually 100 percent debt and internal financing of aircraft in the 1960s and early 1970s, current estimates are that 50 percent of the aircraft of major carriers is under lease and that this figure might reach 70 percent during the 1990s. ' Of particular interest here is the effect that operating leases have on carriers' financial performance. Capital leases appear as long-term liabilities on carriers' balance sheets, and the interest expense and the amortization of principal are treated in a manner similar to interest and depreciation on debt-financed aircraft. The principal advantage offered by capital leases is the lower financing costs that arise from the difference in the ability of lessors and lessees to use depreciation and other tax benefits. In leases under which the airline does not retain the residual value of the plane at the end of the lease, the opportunity for long-term capital gains may be foregone, especially if the expectations of the lessor (reflected in the lease terms) and the lessee differ. Operating leases, on the other hand, do not appear on balance sheets and are sometimes criticized for providing offâbalance sheet financing of essential capital goods. The implication is that fixed capital costs are hidden, thereby presenting a rosier balance sheet picture of debt to equity investors. Nevertheless, the substitution of operating leases for other forms of financing offers advantages as well as disadvantages.
Changes in Airline Operations and Financial Performance 69 On the positive side, operating leases may provide a method of lowering the air carrier's overall cost of capital. Equity, debt, capital leases, and operating leases all have differing direct costs, risk premiums, depreciation and tax benefits, and degrees of flexibility. Relying too heavily on one source of financing would likely result in capital costs greater than those achievable with a mixed portfolio of capital sources. Part of the lower cost that operating leases may provide to an airline's financing derives from the flexibility offered by their relatively short term. The best way to describe this aspect is by way of an admittedly oversim- plified example. Suppose that an airline puts 30 percent of its operating fleet on equally staggered, 5-year noncancellable leases. In this case, 6 percent of its fleet is up for renewal each year. In the event of a fall-off in traffic, these aircraft can be returned to the lessor and a new lease delayed until after the downturn. To the extent that equity financing of aircraft is supplanted by the contractual commitments of debt and leases (both operating and capital), the flexibility and cushion of carriers in a downturn are reduced. In this example, if that 30 percent of the fleet were equity financed, the earnings requirements of these aircraft (essentially dividends and retained earnings) could be totally avoided in a downturn (though eventually, in an upturn, the aircraft must provide a return to compensate shareholders in the long run). In summary, the move to operating leases by carriers may assist in balancing a debt portfolio. However, when changes in individual carrier and industry leverage are examined over time, specific terms of operating leases must be taken into account, because apparent leverage may be reduced by operating leases, but the financial risk of the airline may not drop proportionately. Industry Leverage Significant concern about the financial health of the airline industry re- volves around its level of leverage. In particular, if a firm is already highly leveraged, it may not be able to finance operating losses during recessions or periods of reduced demand. Defining the level of a firm's leverage, however, can be difficult. With regard to debt, the true liabilities of an airline depend on the terms of the commitment to pay, that is, on the degree of the fixed commitment incurred by the airline in the financing of operating assets. To the extent that book liabilities of long-term debt and capital leases do not reflect all commitments to pay, the leverage of the firm is understated. In addition
70 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION to operating leases, another potential source of such understatement, more typical recently, is airline holding companies established for the purposes of diversification, mergers, or buy outs. To the extent that the holding company issues debt to finance operating assets of the air carrier, the book financial results, based on the operating company results only, would understate the debt being used to produce the operating company's net income. 12 Furthermore, if the holding company is financed almost entirely by debt (for example, if debt was issued to pay for the purchase of the operating company's equity in a leveraged buy out) and has no other sources of income beyond that provided by the airline operating company, then the book returns to equity are in truth funds that will be used to pay interest and debt expenses of the holding company. The major implication for this case is that the apparent cushion provided by equity in the airline's capital structure is in fact illusory. With regard to equity, the true value of assets employed by the airline in the production of its services may differ widely from those reported on a book basis. Assets can be misspecified if they go unrecognized or are over- or undervalued. As Levine (M. E. Levine, testimony before Sub- committee on Aviation, House Committee on Public Works and Trans- portation, Oct. 4, 1989) noted, many assets of the airlines are undervalued from a book perspective. These assets include many jet aircraft bought in the 1960s and 1970s that are fully depreciated and yet have significant economic lives remaining and, therefore, significant financial value as well. Other assets that are not recognized on airlines' books include pre- miums on aircraft delivery dates, slots at capacity restricted airports, and gains on leases on gates and equipment at airports. Premiums on newly delivered aircraft have reached 25 percent in the case of some aircraft, and slots have been valued at $1 million.13 The rental value of airport capacity, from a current or replacement cost standpoint, may significantly exceed the rental cost of that capacity because of the long term of the leases (15 to 30 years). The significant increase in demand for space at capacity-constrained airports has increased the value of facilities there as well. Whereas airline operating decisions are properly made on the current value of the capacity, cash outlays for capacity depend on its historical cost. To the extent that any of these assets are not recognized in book equity, the leverage of the firm will be overstated.'4 At any particular time, how onerous or beneficial hidden liabilities and unrecognized assets are to an airline will change depending on the cir- cumstances of the individual airline and expectations about or the state of the economy in general. These assets may be particularly vulnerable to wild swings in value. So at any given moment, whether an airline is really
Changes in Airline Operations and Financial Performance ii Ratio 5 $ Li 68 70 72 74 76 78 80 82 84 86 88 Year - Strong -i-- Weak -41â Total FIGURE 2-3 Debt/equity ratios (book basis). leveragedâwhether it can or cannot reach somewhere into its assets to come up with more money to meet its firm commitmentsâis difficult to determine. Given these problems, the following paragraphs begin with an exam- ination of book leverage for the industry and for the strong and weak groups of carriers. Then for the 5-year period 1985-1989, adjustments to take account of the effects mentioned above are made by using holding company debt and market values of airline equity. Average debt/equity ratios for the industry as a whole are reported in Table 2-1 for the different periods. The average for the postderegulation period was 1.24, down slightly from 1.26 for the prederegulation period and 1.31 for 1968-1978. Debt in these calculations includes both long- term debt and airline obligations under capital leases (which have grown to about one-third of total long-term debt). This measure, therefore, has been affected in recent years by the effects of holding companies and the hidden effects of operating lease transactions. It is unlikely, however, that the ratio would rise significantly above the long-run average even were these effects included.'5 The debt/equity results on a book basis for the strong and weak groups of major carriers are shown in Table 2-2. For the strong carriers, the debt/ equity ratio actually dropped in the postderegulation period, whereas the ratio for the weak carriers increased from an already high 1.96 to nearly 3.5. Figure 2-3 depicts the results summarized in Table 2-2 for the strong and weak carriers.
72 WINOs OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION TABLE 2-3 DEBT/EQUITY RATIOS BASED ON HOLDING COMPANY DEBT AND AVERAGE MARKET VALUES OF EQUITY Carrier 1985 1986 1987 1988 1989 Strong American 0.73 0.81 1.03 0.92 0.46 Delta 0.47 0.40 0.42 0.51 0.36 Northwest 0.72 1.11 0.61 0.67 NA United 1.53 0.82 0.43 0.99 0.31 USAir 0.87 0.61 1.20 0.92 0.52 Subtotal 0.84 0.73 0.67 0.81 0.40 Weak Continental 3.30 4.05 4.53 10.03 7.51 Pan American 0.66 0.89 1.58 2.41 2.11 Trans World 1.68 3.08 3.14 2.91 NA Subtotal 1.91 2.72 3.40 4.88 5.13 Total 1.10 1.13 1.13 1.41 0.76 NOTE: Average market values of equity are the average of the high and low values of the common stock for the year. NA = not applicable. SOURCE: From data supplied by John S. Strong. The high leverage of a few carriers (Continental, Trans World, and Pan American) is lost in the overall industry average, and the holding company debt of Northwest is not accounted for at all. Therefore, the debt/equity ratios of individual strong and weak carriers for 1985-1989 given in Table 2-3 are based on holding company debt and average market values of equity. By using market values of equity, the value that the market places on assets that are not reflected in the book values can be captured, and by using holding company debt, a more accurate estimate of the fixed commitments of the airlines is obtained. Figure 2-4 shows the results in Table 2-3 for the strong and weak carriers and the total for the entire group. The striking dichotomy between the strong carriers and the weak ones is apparent in both Table 2-3 and Figure 2-4. American, Delta, United, USAir, and Northwest (until the buy out in 1989) have generally kept debt to less than half of their capitalization.16 Continental, Pan American, and Trans World, on the other hand, rely heavily on debt to finance their operations. Whether high levels of debt can be sustained by these carriers is problematic. The bankruptcy of Continental was tied directly to the high cost of carrying its debt, and Trans World was forced to defer payments to some of its bondholders in the first quarter of 1991. Certainly, the high leverage of these carriers has made them vulnerable in the current recession and, if they survive, may make it difficult for them to renew their fleets in the reequipment cycle that should last for the next decade.
Changes in Airline Operations and Financial Performance 73 Ratio 6 r â o 1985 1986 1987 1988 1989 Year - Strong - Weak - Total FIGURE 2-4 Debt/equity ratios (holding debt/average market equity basis). Returns on Equity and Investment Returns on equity and total investment are key measures of an industry's and individual firms' abilities to attract capital for continued growth and replacement of existing assets. Whereas high leverage can kill a firm quickly when operating losses arise, continual low returns on invested capital can also lead to its demise, though generally more slowly. Eventually, access to equity markets would be lost, forcing a reliance on debt and the subsequent potential for an inability to cover fixed charges. Returns on equity and investment for both the industry as a whole and for the major carriers in- dividually during the last 20 years are examined in this section. Industry results as reported on a Form 41 basis are discussed below. The problems noted previously about holding companies and the valuation of assets affect the interpretation of ROEs. To the extent that assets go unrecognized on a book basis, but produce some of the returns reported, the ROE is overstated. To the extent that leased assets are being purchased at prices that include a competitive rate of return, below-average ROEs are understated. These results should be considered, therefore, as an am- biguous estimate of industry results. In the case of holding companies owning all the airline's equity and being financed themselves by debt, the reported ROE really has no meaning. The results reported here account for these cases.'7 The industry's ROE has never been strong and consistently has under- performed the average of all manufacturing industries except for the strong
74 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION Percent 25 -15 1955 1960 1965 1970 1975 1980 1985 1989 Year ROE ROI FIGURE 2-5 ROE and ROL for tota' certificated air carriers. year of 1978 and the expansion years of the mid-1960s)8 In the post- deregulation period, a further, definite decline in the average, nonrecession ROE has occurred. On the other hand, the airlines are producing sufficient returns to cover a higher cost of debt. Therefore, although the returns available to equity have been forced down, industry assets, overall, are producing a higher return on total investment. Air carrier ROE and ROl are shown in Figure 2-5 and Table 2-1. ROl as shown here is based on a simplified measure of investment. The nu- merator includes interest expense and net income after taxes, and the denominator includes only year-end long-term debt, obligations under capital leases, and stockholder equity.'9 The volatility of industry ROE stands out in Figure 2-5. Furthermore, in the postderegulation period, the average, nonrecession ROE was 4.4 percent. This is a significant decline from the prederegulation average of 9.3 percent, and a definite, further decline from the 1968-1978 average of 7.6 percent. Although the industry's ROE has declined in the face of the increased expense of debt, earnings from operations have been sufficient, never- theless, to yield an increase in ROI. ROl increased in the postderegulation period to 8.0 percent from 7.4 percent in the prederegulation period and 7.2 percent in the 1968-1978 period. The airline industry has been earning a better return on its investment since deregulation, but has paid a sig- nificantly higher price for the debt portion of that investment. It has been able to attract substantial amounts of debt and lease capital throughout
Changes in Airline Operations and Financial Performance 75 Percent 20i 10 0 -10 -20 -30 -40 -50 68 70 72 74 76 78 80 82 84 86 88 Year - Strong -i--- Weak FIGURE 2-6 ROE for strong and weak carriers. this period. The apparent returns to investors in airline equity have been poor. Nevertheless, this result should be embraced with caution because the deregulated airline industry has attracted new equity capital in many of the years since deregulation. As is the case for operating results, but even more so, returns on capital on an industrywide basis hide the large discrepancy between a group of relatively strong airlines and a group of weaker ones. Table 2-2 includes ROE and ROl for strong and weak carriers.20 The strong carriers have achieved, both before and since deregulation, ROEs much closer to those seen in all manufacturing industries and ones that would leave one less concerned with their future ability to attract equity investment. Average ROE excluding the recession years equaled that of the prederegulation period, whereas ROl increased. For the weak carriers, the poor operating results of the last decade are reflected in the large losses to equity. The actual figures for this group should be interpreted with caution, however, because of the high leverage of these firms in recent years. The low return on total investment is more indicative of the weak performance of these carriers since deregulation. Figures 2-6 and 2-7 show the returns for these two groups. Interestingly, the weak carriers matched the performance of the strong carriers in the period of administrative deregulation after the 1974-1975 recession, but they never recovered from the subsequent shocks of the early 1980s. The results of the strong carriers overestimate their real returns to the extent that unrecognized assets underestimate actual assets deployed
76 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION Percent 20 % ' I 1 807'27476 78018121818688 Year - Strong - Weak FIGURE 2-7 ROl for strong and weak carriers. in airline service. (Remember, however, that leased capital can be presumed to earn a competitive return to the lessor.) A general indi- cation of the extent of this kind of discrepancy can be determined through an examination of market valuations and ROE for the 1985- 1989 period. Market Valuation of Equity Because of the potential undervaluation of equity on a book basis, an alternative method of viewing returns is to look at the valuation of the airlines as indicated by their market values. Table 2-4 presents airline values based on the high and low stock prices for each year for 1985- 1989 and compares them with the value of stockholders' equity on a book basis. For half of the airlinesâAmerican, Delta, Trans World, and Unitedâ book value generally forms a lower threshold for the trading values. For Northwest and USAir, market values hover at or just below book value. For Continental and Pan American, the market values even at their low points have stayed above their eroded book value, especially in recent years. For the most part, market values have significantly, and consistently for some carriers, exceeded book values. Table 2-5 presents ROEs on a consolidated, average market value basis for the 1985-1989 period. Except for 1988, market returns, as expected, have been consistently below book returns for the strong carriers. For the
TABLE 2-4 BOOK VALUE OF STOCKHOLDERS' EQUITY AND HIGH AND LOW MARKET VALUES FOR COMMON EQUITY IN EACH YEAR 1985 1986 1987 1988 1989 American" High 3,080 3,650 3,852 3,235 6,687 Book 1,888 2,140 2,252 2,701 3,118 Low 2,030 2,306 1,573 2,735 3,242 Continental" High 1,522 1,635 2,007 673 938 Book 687 1,003 844 168 (713) Low 609 595 351 349 446 De1ta High 2,524 2,523 3,295 2,716 4,660 Book 1,546 2,170 2,057 2,369 2,902 Low 1,590 1,836 1,571 1,774 2,649 Northwest" High 1,703 1,502 2,212 1,671 3,524 Book 1,142 1,881 1,995 2,122 2,453 Low 979 991 892 1,060 1,204 Pan Americane High 1,236 1,306 892 486 757 Book 269 (168) (429) (535) (963) Low 835 550 367 324 332 Trans World1 - High 1,114 835 1,068 1,297 NA Book 660 419 511 (31) (345) Low 521 392 427 363 NA United High 2,074 2,927 5,890 2,567 6,367 Book 1,277 1,196 1,228 1,268 1,665 Low 1,386 2,091 2,907 1,593 2,279 U5Air High 1,806 2,071 2,018 1,697 2,412 Book 1,544 1,809 2,714 2,863 2,808 Low 1,276 1,415 981 1,185 1,349 NOTE: Market value equity calculated as the average number of common shares outstanding during the year multiplied by the high and low prices for common stock. Values are in millions of dollars. NA = not applicable. "Listed as AMR Corp. Includes consolidation of value of Air Cal and of commuter airlines acquired as part of American Eagle. 'Listed as Continental Holdings as of 1990. Includes Continental, Eastern, People's Express, and Frontier. Equity values reflect bankruptcy filings of Eastern and Continental. 'Includes Western. "Includes Republic. Listed as NWA Corp. before acquisition by Wings Holdings in July 1989 for $121 a share; this is the high equity value for 1989. 'Listed as Pan Am Corp. 1lncludes Ozark. Trans World was taken private by 1W Acquistion in October 1988 for a combination of $20 cash and $30 face amount (principal) of 12 percent subordinated debentures due 2008. The 1988 high value reflects a $50 per share valuation of the going private transaction. The actual value may have been less given that the debt traded at a substantial discount. Just before the acquisition, Trans World common traded at $37 per share, thus implying a 1988 market value equity high of $959.9 million. Includes Piedmont and Pacific Southwest Airlines. SOURCE: From data supplied by John S. Strong and in Air Carrier Financial Statistics, DOT.
78 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION TABLE 2-5 ROE ON AN AVERAGE MARKET VALUE BASIS Carrier 1985 1986 1987 1988 1989 Strong American 12.6 8.4 7.7 15.1 8.5 Delta 10.9 9.2 8.9 15.3 12.9 Northwest 18.6 11.0 9.1 11.9 15.0 United -5.1 -3.2 0.8 28.3 8.3 USAir 11.5 9.3 17.5 16.4 0.1 Subtotal 9.6 6.3 6.8 17.6 9.4 Book 9.7 6.9 8.9 15.7 12.4 Weak Continental 6.6 -10.1 -37.3 -127.4 -122.7 Pan American 3.9 -50.6 -43.6 -29.2 -76.2 Trans World -23.6 -17.1 14.2 30.1 NA Subtotal -2.8 -25.8 -23.8 -29.7 -102.2 Book -6.7 -19.3 -24.6 NA NA Market total 6.6 -0.2 1.7 10.6 1.9 Book total 7.1 3.9 4.9 15.7 12.4 Nom: Average market values of equity are the average of the high and low values of the common stock for the year. NA = not applicable. SOURCE: From data supplied by John S. Strong and in Air Carrier Financial Statistics, DOT. weak carriers, the returns have been about the same, and uniformly poor except for Trans World in 1987 and 1988. Market Perception of Financial Risk In addition to these financial ratio measures of industry performance and what they imply for the industry's financial health, another measure is changes in the equity market's perception of an industry's risk. Market risk can be estimated by regressing a company's stock returns on those of the market as a whole. The resulting coefficient, commonly referred to as "beta," indicates the firm's risk relative to the market, with a value greater than 1.0 indicating higher risk, and vice versa. Appendix A con- tains a discussion of the interpretation of betas and a detailed description of the technique used to estimate them. Table A-i presents the industry equity and asset betas from 1973 to 1989. Equity betas show the riskiness of holding an airline's stock (relative to a portfolio of the total market), whereas asset betas measure the underlying business risk of deploying capital in the airline industry. In the early to mid-1980s, airline asset betas more closely resembled those of the market in general, which suggests that the systematic risk of the industry became more like that of the economy as a whole. Equity
Changes in Airline Operations and Financial Performance 79 betas rose as a result of increased indebtedness, which caused the riskiness of airline stocks to rise. In 1988 and 1989, both the equity and asset betas rose sharply. The rise in the asset beta indicates an increase in the per- ceived, systematic business risk of airlines. As noted in Appendix A, the large increase in the equity beta is likely the result of the increase in the industry's leverage (partly to finance a necessary reequipment cycle) and the financial risk this entails for the airlines in covering the fixed costs it imposes on them. The results of the beta analysis support the overall view that investment in the industry carries considerable risk. Whereas in the early years of deregulation the airline industry appeared to be viewed as having somewhat lower systematic risk, that risk appears to have increased because of the high leverage of some firms, the coming reequipment cycle in the industry, and the current recession and weakened travel demand. Summary Air carrier financial performance since deregulation can be characterized as follows: Operating results have been somewhat weaker on an industrywide basis, though they were hardly robust before deregulation. On an individual firm basis, however, the industry can be divided into a group of strong performers whose results have slipped only slightly, if at all, and one whose results have brought their continued existence into question. Some increase has occurred in industry leverage, though again, this has tended to concentrate in certain firms (Continental, Trans World, Northwest, and Pan American). ROEs on an industrywide basis have declined, but can be difficult to interpret. ROEs are as bipolar as operating results, however, with the group of strong carriers achieving returns much more in line with those of other industries. Industry betas indicate that the systematic risk of the industry has increased in recent years. One may interpret the results in various ways. The weak performance of the industry as a whole could be seen as the evidence of a drastic decline that is still under way and that is leading to the collapse of the industry and the concentration of airline service into the hands of a few
80 WINDs OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION carriers. On the other hand, the weak performance can be viewed as the necessary shaking out of weak carriers, with the result that the industry will emerge with fewer, stronger carriers that are capable of competing vigorously nationwide. Indeed, deregulation was not meant to improve or aid the performance of all carriers in the industry, but to reward the best managed. Clearly, the industry is still in transition from the regulatory period. The failure of Eastern and the bankruptcies of Pan American and Conti- nental can be traced without much difficulty to the operating and structural adjustments that have been continuing since deregulation. It appears that the industry is moving toward an equilibrium of a small number of large nationwide carriers and an uncertain number of significant smaller ones. The industry's financial performance in such a world may end up on any point of the spectrum. On one hand, a loose oligopoly could form in which fares and service remain fairly stable and remunerative but at levels close to or equal to competitive returns. On the other hand, a tight oligopoly could arise, leading to higher-than-normal profits, reduced service offer- ings on routes, and efficiency losses to the economy. Examined in suc- ceeding chapters of this report are the fare and service conduct of carriers since deregulation and potential impediments to competitive behavior, particularly those that might tend to enhance the profitability of oligop- olistic behavior. From a financial standpoint, it is difficult to predict what the recent financial history portends. Thin operating margins offer little pricing room to avoid slipping below fares that are sustainable in the long run. The level of leverage in the industry may work in both directions. Highly leveraged firms will have incentives to control costs and produce effi- ciently; at the same time, the exigencies of meeting debt service may encourage initiation of or acquiescence in oligopolistic pricing behavior. Another concern often expressed is whether the industry has the financial capacity for the reequipment and expansion of its fleet during the next decade. The ability to attract capital and provide service depends on the strength of the underlying demand for an industry's output. Few doubt that airline service will continue to be in high demand in the long run. Undoubtedly, capital will be available at some price to meet that demand. Returns in the industry may have to rise to pay that price. Deregulation has provided firms with the ability to respond rapidly to changes in their costs and the fares offered by their competitors. In the end, the profitability and strength of airline firms will depend on how well they use those freedoms. Whether the emerging capital structure of the airlines will be a source of additional strength and profitability or will
Changes in Airline Operations and Financial Performance 8 add costs from inefficient structures, excessive risks, or costs of adjustment from bankruptcies and disruptions of service cannot be predicted with the available evidence. It is conceivable that some reduction in the number of carriers will reduce excess capacity in the industry and thereby improve the financial performance of the remaining carriers. The lackluster finan- cial condition of some of these carriers, however, which might be further weakened by the current recession, the increase in fuel prices, and the disruption in travel caused by the Persian Gulf war, gives reason to be concerned about the number of carriers that will survive and the effect this might have on competition. NOTES I. Wheeler (1989) pointed out that as spokes are added to a hub network the number of potential city pairs served increases dramatically. The total number is equal to N(N - 1)12, where N equals the total number of cities in the network, including the hub. Not all of these cities pairs, however, will have passenger flows between them; for example, some of the city pairs will be within easy driving distance of one another. Hub-and-spoke networks also have some disadvantages: increased congestion at hub airports; heavy use of assets at hubs, which because of the peaking required to make connection banks work, results in departure costs higher than those of linear route systems; and service disruptions caused by inclement weather at the hub (Wheeler 1989). The committee disagreed as to whether the excess wide-bodied aircraft actually hastened a shift toward hubbing. One can argue that in United's case, it had to use wide bodies in the short term until it could obtain more narrow-bodied aircraft, and found that it could use them on the most concentrated routes, especially if it shifted more rapidly into a hub-and-spoke network. Other carriers, however, such as Eastern, sought additional transcontinental or international routes (or expanded service on these routes) that have stage lengths more appropriate to the economics of wide-bodied aircraft. This discount practice, referred to as yield management, does allow carriers to hold down the cost of all seats sold, and thereby serves the interests of consumers. However, like many of the marketing innovations developed by large carriers, it has also made it harder for small carriers to compete on the basis of low fares. Recent changes require accounting for frequent flier mileage, and carriers now account for such liability. The issue of accurately reflecting the liability remains controversial. It is difficult to account for this liability because some fliers do not use their benefits at all and some convert them to first-class upgrades. For a more detailed estimate of productivity benefits and the difficulties of dis- aggregating and measuring such benefits, see works by Meyer and Oster (1981, 74-90; 1987, 83-108). For individual carriers, data were collected for the years 1968-1989. Further complicating the 198 1-1983 period is the effect of the air traffic con- trollers' strike on the industry beginning in 1981.
82 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION For example, the large price cuts Eastern made to win back traffic after its strike forced all major competitors on its routes to respond. With Eastern's cessation of service, most carriers immediately announced the end of their matching discount fares. Most discount fares had a set duration for their use, and most carriers were waiting for those periods to expire before enforcing higher fares. A study by Strong in work by Meyer and Oster (1987) showed that this may, in fact, be the case. The figure is roughly split evenly between operating and capital leases. Other estimates are for a 40/30/30 split among debt, operating, and capital leases (A via- tion Week and Space Technology I 990c; Airline Business 1989; Homer 1990). Generally, this has not been the case for the airlines, with a few exceptions. For example, Texas Air (now Continental Holdings), the holding company for Eastern and Continental, received some of Eastern Airlines' aircraft and associated debt. Airlines have enjoyed the increase in the value of their orders for new aircraft between the time of the order and the delivery of the aircraft. In some cases, they have monetized this increase by selling the aircraft to investors and leasing them back under capital leases. The point of recent buy outs has been, in fact, to capture the value of the undervalued assets of the airlines and return it, in part, to former shareholders through a premium on their equity over book value and to the new owners if they have calculated the economic value of the assets accurately. Furthermore, all airlines have been converting some of their undervalued assets in the last several years either to retire debt from a leveraged buy out, avoid the potential for a buy out, or cover operating losses. Debt incurred for both mergers and buy outs has tended to be reduced quickly thereafter, for example, USAir Group's reduction of $2 billion of acquisition debt to $1 billion, in part through $600 million of equity, and Wings Holding's re- duction of $3.8 billion in the leveraged buy out of Northwest to an estimated $2.1 billion at the end of 1990 and predicted $1.7 billion by early 1991. The leveraged buy out of Northwest in 1989 was essentially an all-debt transaction that left no traded shares. The $3.8 billion of debt incurred certainly represented a heavy fixed commitment by the airline, though this debt has been reduced according to the carrier through asset sales and leasebacks. Trans World and Northwest were taken private in late 1988 and mid-1989, re- spectively. Book returns on equity are calculated for Northwest on the basis of equity before the leveraged buy out. Trans World had negative book equity in 1988. Interestingly, 1978 was the peak year of administrative deregulation. The strong returns to equity in that year perhaps reflect increased competition for traffic and more intensive use of assets as the incumbent airlines used their increased service and pricing freedoms. Subsequent entry by new, low-cost carriers, the run-up in fuel prices in 1979, the recessions of the 1980s, and the controllers' strike com- bined to severely damage returns to equity until 1984. This measure differs from historical CAB calculations of total investment and regulatory investment. The former is based on five-quarter moving averages and includes numerous smaller accounts in the investment base. The latter was cal- culated for a number of years in the late 1960s and early 1970s and excluded from returns and the investment base earnings and assets that were not directly related to air carrier certificated service. The definition of investment used in Table 2-2 differs slightly from that used in Table 2-I because additional data were available, though the results are virtually
Changes in Airline Operations and Financial Performance 83 identical for the totals in Table 2-2 and the industry averages in Table 2-1. Investment here includes both current and noncurrent long-term debt and capital leases and net stockholders' equity. Furthermore, the average investment for the year is used instead of year-end figures. REFERENCES ABBREVIATIONS DOT U.S. Department of Transportation FAA Federal Aviation Administration Airline Business. 1989. The Super Marketers. July 1989, p. 34. Aviation Week and Space Technology. I 990a. Southwest Airlines Gains Major Carrier Status by Using Go-It-Alone Strategy. March 5, pp. 82-84. Aviation Week and Space Technology. 1990b. America West Growth Program Forges Ahead with Airbus Order. Oct. 8, p. 30. Aviation Week and Space Technology. 1990c. New Regulations, Surplus Will Force Narrow-Body Transports Out of Service. July 2, 1990. p. 77. Bailey, E., D. Graham, and D. Kaplan. 1985. Deregulating the Airlines. MIT Press, Cambridge, Mass. Blyskal, J., and M. Hodge. 1989. Your Airline Survival Guide. New York Magazine, Aug. 28. Brenner, M. 1988. Airline DeregulationâA Case Study in Public Policy Failure. Transportation Law Journal, Vol. 16, pp. 179-227. Brenner, M. 1990. Analysis of Airline Concentration Issue. Melvin Brenner and Associates, Rowayton, Conn., July. Clippinger, M., and J. Strong. 1987. Changes in Distribution Channels and the Travel Agent Business. In Deregulation and the Future of Intercity Passenger Travel, MIT Press, Cambridge, Mass. DOT. 1990. Secretary's Task Force on Competition in the U.S. Domestic Airline Industry. Airline Marketing Practices: Travel Agencies, Frequent-Flyer Programs, and Computer Reservation Systems. FAA. 1989. FAA Aviation Forecasts Fiscal Years 1989-2000. Report FAA-APO-89- 1. U.S. Department of Transportation. Homer. 1990. A Broker's View of the Aircraft Finance Market. Airfinance Journal, Jan. 1990, p. 42. Kahn, A. 1988. Airline DeregulationâA Mixed Bag, But a Clear Success Never- theless. Transportation Law Journal, Vol. 16, pp. 229-251. Levine, M. E. 1987. Airline Competition in Deregulated Markets: Theory, Firm Strategy and Public Policy. Yale Journal on Regulation, Vol. 4, Spring. Meyer, J., and C. Oster, Jr. 1981. Deregulating the Airlines: The Early Experience. Auburn House Publishing Co., Dover, Mass. Meyer, J. and C. Oster, Jr. 1984. Deregulation and the New Airline Entrepreneurs. MIT Press, Cambridge, Mass. Meyer, J., and C. Oster, Jr. 1987. Deregulation and the Future of Intercity Passenger Travel. MIT Press, Cambridge, Mass.
84 WINDs OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION Morgan, I. 1981. Government and Industry's Early Development, In Airline Dereg- ulation: The Early Experience, Auburn House Publishing Co., Dover, Mass. Oster, C., Jr. 1981. The Impact of Rising Fuel Prices. In Deregulating the Airlines: The Early Experience, Auburn House Publishing Co., Dover, Mass. Wheeler, C. 1989. Strategies for Maximizing the Profitability of Airline Hub-and- Spoke Networks. In TransportationResecirchRecord 1214, TRB, National Research Council, Washington, D.C., pp. 1-9. Yom, Sue-Sen. 1990. Fliers Feel the Pinch in Tighter Seats. Wall Street Journal, June 27, p. BI.
Passenger Fares and Airline Service The economic deregulation that swept through the transportation sector in the mid-1970s to early 1980s stemmed in large part from growing evidence that economic regulation had raised prices and limited the goods and services available to consumers. In the case of the airline industry, the removal of economic (fare and route) regulations tentatively began in 1975-1976, gathered momentum through additional administrative mea- sures taken in 1977-1978, and culminated in the implementation of the Airline Deregulation Act in 1978. The primary goal of airline deregulation was to provide air travelers with a wider selection of fares and services by stimulating competition and forcing airlines to market their services and manage their costs more aggressively. Since deregulation, air travelers have no doubt noticed major changes in the type and quality of services available to them and the prices and pro- motional practices of the airlines they fly. For many travelers, the removal of price and route regulations has proved favorable. The key points made in this chapter are that average passenger fares in most markets are lower when adjusted for inflation, and flight frequencies have increased sharply through the expanded use of hub-and-spoke systems. Clearly, however, fare and service options have not changed equally in every market, and substantial trade-offs between the two have affected some travelers more than others. Not long after passage of the Airline Deregulation Act in 1978, several attempts were made to evaluate deregulation's impact on fares and service in more detail, across different kinds of airline markets, and for different kinds of travelers (CAB 1984; GAO 1985; Morrison and Winston 1986; Meyer and Oster 1981, 1984, and 1987). As discussed in Chapter 1, however, the advent of deregulation was accompanied by several external events, including the 1979 oil supply shock, which sharply raised jet fuel prices; the nationwide economic recession of the early 1980s, which dampened travel RR
86 WINDS OP CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION demand; and the air traffic controllers' strike in 1981. The impact of these events, coinciding with an unsettled transition toward deregulation (charac- terized by new entry, below-cost pricing, and numerous airline failures), made early evaluations difficult and clouded the effect of many evolving industry phenomenaâsuch as the trend toward vertical and horizontal in- tegration of airlines, the proliferation of hub-and-spoke networks, and the emergence of more complex fare structuresâthat would come to characterize a more mature, postderegulation airline industry. Accordingly, several at- tempts have been made in recent years to reexamine trends in airline fares and service in order to account for the effect of these industry phenomena on consumers [e.g., studies by the U.S. Department of Transportation (DOT) (1990a) and the General Accounting Office (GAO) (1990)]. CHAPTER ORGANIZATION Some of the analyses found in previous studies are retraced in this chapter, and different approaches are applied to estimate the effect of industry changes since deregulation on passenger fares and service. The first section begins with a general discussion of fare and service patterns nationwide since de- regulation, including trends in passenger traffic, average fares, and the variety of fares being offered. This discussion is followed by a more detailed ex- amination of fare and service changes by market distance and density. The focus of the chapter then turns to more recent concerns about airline competition. In particular, as the share of traffic handled by major carriers has grown, and as these carriers have become dominant at their hubs, concern has risen over the level of fares and service in some concentrated markets. Accordingly, analyses in this section are devoted to the exam- ination of trends in fares and service disaggregated by the number of competitors and by market type (e.g., concentrated hubs versus nonhubs). Finally, because a point of controversy throughout the debate over deregulation has been its impact on small communities, a final section contains a brief review of fares and service trends for small cities, including those that received subsidized service before deregulation. DATA SOURCES Several data sources are employed in this chapter. The source used most frequently is the DOT origin and destination (O&D) survey of tickets from passengers traveling on domestic carriers. The information provided in this survey of every tenth ticket includes each passenger's origin and destination,
Passenger Fares and Airline Service87 number of flight changes en route, airline flown, and fare paid. The O&D survey is particularly useful for estimating the number of passengers flying in individual markets, the market share of each carrier, and the average yield (fare per passenger mile) paid by passengers. Unfortunately, historical O&D data on fares only date back to 1979, and therefore comparable prederegu- lation information is not available. Also, the data have serious limitations when applied to the smallest communities receiving air service, both because of sampling problems (related to the 1-in-1 0 ticket sample) and the lack of participation in the survey by many smaller commuter carriers that are exempt from federal O&D reporting requirements. The O&D survey is the primary source of fare and passenger trip information, but several other data sources also are employed in the chap- ter. The two main sources for passenger enplanements are the DOT's traffic reports [T-3 and 1-9 from Research and Special Programs Admin- istration (RSPA) Form 41] for large carriers (majors and nationals) and similar reports for commuters (RSPA Form 298-C). Departures also are reported by the large carriers in the T-3 and T-9 reports, but are not reported by commuters; accordingly, departure data for several analyses are also derived from the scheduled flight listings published in the Official Airline Guide (OAG). Table 3-1 is a summary of the various data sources used in this chapter. As noted, shortcomings in the data make it impossible to examine all trends for consistent time periods and for all kinds of airline markets. In general, however, the information base for the airline industry is far superior to that of most other transportation industries, making possible a wide range of informative, public policyârelated studies. GENERAL TRENDS Examined in this section are general trends since deregulation in passenger traffic, air travel prices in real terms (adjusted for inflation), variety of fare offerings, and fare patterns by market distance and density. Passenger Traffic If one judges solely from the number of people flying today compared with that of a dozen years ago, it would appear that deregulation has provided substantial benefits to air travelers. Many more travelers are flying now than ever before, as measured by both passenger enplanements and passenger trips, which exclude connecting enplanements.
TABLE 3-1 SUMMARY OF AIRLINE DATA SOURCES FOR FARE AND FLIGHT INFORMATION Source Years Available Carriers Reporting Relevant Data T-3 (RSPA Form 41) 1972 to present Major and national By quarter, carrier, and city domestic Departures Passengers enplaned T-9 (RSPA Form 41) 1984 to present Major and national By quarter, carrier, and city pair domestic Departures Passengers enplaned Available seat miles Revenue passenger miles Average load factor RSPA From 298-C 1972 to present Commuter Annually, by carrier and city pair Passengers enplaned O&D survey (DATABANK I-A) 1979 to present Major and national By quarter, carrier, and city pair domestic O&D passenger trips Average fare Number of coupons (connections) Revenue statistics Official Airline Guide (OAG) North Automated version, All scheduled Flight frequency American 1980 to present Flight times Hard copy before 1980
Passenger Fares and Airline Service 89 Millions 500 400 300 200 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 8384 85 86 87 88 89 Year -*-- Enplanernents - Passenger Trip8 FIGURE 3-1 National trends in passenger traffic, 1968-1989 (FAA fore- casts, DOT O&D survey). In some respects, this growth in passenger traffic has continued a long- term trend upward in air travel that followed the widespread introduction of jet aircraft in the 1960s, which greatly increased air travel demand and convenience while reducing its cost (Bailey et al. 1985, 17). During the decade before deregulation (from 1968 to 1977), both passenger enplane- ments and passenger trips (which exclude connecting enplanements) grew by nearly one-half (Figure 3-1). Since deregulation, however, the number of enplanements and trips has grown at a much faster pace, nearly doubling during the 12-year period from 1977 to 1989 (Figure 3-1). Consumer surveys by the Gallup Organization show that this upward trend in passenger trips has coincided with considerable increases in the proportion of the population that relies on air transportation for pleasure travel. In 1977, for example, 19 percent of the population used air service for pleasure trips; by 1988, this proportion had increased to 25 percent (ATA 1988). This increased proportion, combined with the growth in population, resulted in a 51 percent increase in total pleasure trips by air during this period (ATA 1988). Average Fares Growth in passenger traffic has been accompanied by, and indeed largely stimulated by, declines in the price of air travel relative to most other
90 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION Cents per Mile 20 19 18 17 16 15 14 13 12 11 in 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 Year - Industry Yield - Average Trip Yield FIGURE 3-2 National trends in average real yield, 1968 to 1989 (yields inflated to $1989 using the GNP implicit price deflator) (DOT O&D survey; DOT 1990b). consumer goods and services. Traditionally, average yield, which is a mileage-adjusted measure of average fare, was calculated by dividing total industry ticket revenue by total revenue passenger miles. In the decade before the Airline Deregulation Act (1968 to 1977), average yield defined in this way fell by about 16 percent when adjusted for inflation using the gross national product (GNP) implicit price deflator (Figure 3-2). During this period, air travelers continued to benefit from technological improve- ments in jet aircraft, which lowered average costs per passenger mileâ especially in longer-haul, high-density marketsâand the beginning of administrative deregulation in the mid-1970s, which resulted in airlines offering lower discount and specialty fares in many of the country's heavily traveled markets (e.g., Peanut and Supersaver fares) (Bailey et al. 1985, 17, 18, 40-45). The trend in average industry yield since passage of the Airline De- regulation Act in 1978 is also shown in Figure 3-2. From 1977 to 1989 industry average yield declined in real terms by about 20 percent. Today, however, average yield also can be calculated by simply averaging the yields paid on tickets sampled in the O&D survey since 1979. Instead of total ticket revenue divided by total passenger miles, this alternative mea- sure of yield is calculated by averaging actual yields paid by passengers. When adjusted for inflation, average yield derived in this way has fallen by 16 percent since 1979 (Figure 3-2).
Passenger Fares and Airline Service 91 - 1977.100 I 1 1 1 VV 80 70 60 20 to- 00 77 78 79 80 81 82 83 84 85 86 87 88 89 Year -i-- SIFL -b-- Average Yield -- GNP Price Deflator FIGURE 3-3 Nominal average yield versus SIFL index and GNP price deflator, 1977 to 1989 (DOT 1990; DOT O&D survey). During the early stages of deregulation, this downward trend in real fares was interrupted by the 1979 oil supply shock, which caused a rise in jet fuel prices, and the 1981 air traffic controllers' strike, which severely limited the ability of carriers to adjust supply. Hence, during this period, airline costs, and subsequently fares, tended to increase much faster than the general rate of inflation. Figure 3-3 shows the trend in average yield in nominal terms (unadjusted for inflation) since deregulation compared with trends in the GNP price deflator and industry costs, measured by the Standard Industry Fare Level (SIFL) index [developed shortly after de- regulation by the Civil Aeronautics Board (CAB) to analyze fare trends]. In many respects, the SIFL index is a better deflator of yields than the GNP index because it focuses specifically on airline input costs (e.g., jet fuel, maintenance, and terminal charges).' After the supply shock events of 1979 and 1981, the GNP price deflator increased by about 20 percent (1979 to 1981), whereas both the SIFL and nominal yield increased nearly twice as fast. Since these early increases, however, average yield has dropped dra- maticallyâby about 25 percent since 1981 when adjusted for inflation with the GNP deflator (Figure 3-2). The mid- 1980s brought lower jet fuel prices and continued efforts by deregulated airlines to control costs, es- pecially by revision of labor agreements and improvement in worker pro- ductivity. From 1982 to 1987, average costs per seat mile, as measured
92 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION by the SIFL index, declined by about 10 percent, which stimulated further reductions in fares (Figure 3-3). Moreover, this decline took place when the economy was recovering from recession (in 1980 and 1982) and when many new-entrant airlines and holdover carriers were trying to expand their market share (Bailey et al. 1985, 61-65). This period of economic recession and the resulting downturn in travel demand forced carriers to compete for increasingly fewer travelers by further slashing fares, even while many new, low-cost carriers were entering the marketplace (Bailey et al. 1985, 61-65). More recently, the trend in passenger fares has not been nearly so dramatic; indeed, for the nation as whole, average yield actually increased by 5 percent in real terms between 1987 and 1989 (Figure 3-2). During this period, airline costs, as illustrated by the SIFL, increased at a faster pace than inflation, thereby contributing to this slight upward trend (Figure 3-3). Also, by this time, most new entrants had either failed or merged with the surviving incumbent carriers (as discussed in Chapter 1). Whether this greater industry concentration played a role in the upward trend in fares is unclear. During the past few years, several studies have found noticeably higher fares in markets with few competing carriers (GAO 1988, 1990; DOT 1990b). Accordingly, in subsequent sections of this chapter, fare and service trends are examined more closely by level of market competition. Fare Variety In addition to lower average fares, another anticipated benefit of de- regulation was the proliferation of a wider variety of fare and service combinations. During the debate over deregulation, regulated airlines were depicted as being far more service intensive and far less cost conscious than they would be in an unregulated, competitive environment. For example, be- cause airlines were prohibited from making frequent fare adjustments, carriers were effectively encouraged to compete for passengers by using methods of competition not related to price, such as offering larger aircraft with more spacious seating arrangements (Douglas and Miller 1974). Such service competition was especially prevalent in high-density and long- haul markets, in which carriers had benefited from the introduction of cost-saving aircraft and regulated fares that were often intentionally held above prevailing costs (Bailey et al. 1985, 20). Seldom did this compe- tition spill over into lower fares or experimentation with new combinations of fares and service (Meyer and Oster 1984, 3).
Passenger Fares and Airline Service 93 Percent of Trips 18 16 J 14 50% 150% I2d II 10 -1 R !I - in in in in in in P!. -i' 0 1 30-49 1 70-89 1110-1291150-169 I 190-209 1 230-249 1 270-2891 310-329 1 350+ 1-29 50-69 90-109 130-149 170-189 210-229 250-269 290-309 330-349 Percent of Industry Average Yield 1989 = 1979 FIGURE 3-4 Dispersion from average yield, 1979 versus 1989 (zero fares not included in calculation of average yield) [DOT O&D survey (third quarters of 1979 and 1989)]. The removal of fare and route restrictions was expected to produce a much wider range of fare and service combinations, and especially more lower-fare and lower-service options for leisure travelers, to whom price is generally more important than service. Because of this tendency toward service methods of competition, average service quality was probably much higher than many passengers would have chosen had lower-quality service been available at lower prices (Meyer and Oster 1984, 3). Once carriers were given greater freedom to adjust prices and services rapidly according to market demand and cost, it was anticipated that many pas- sengers would switch to lower-priced and lower-service air travel. More specifically, it was predicted that vacation travelers would accept reduced service quality, such as off-peak departures and more connections, in exchange for lower, discounted fares, whereas more high-service options, such as more frequent and conveniently scheduled flights, would be de- manded by business travelers to whom service tends to be more important than price. As expected, this change has led to a much wider variation in fares. Figure 3-4 shows the proportion of passengers who paid various per- centages of the industry average yield in the third quarters of 1979 and 1989. The data show that in 1989, 49 percent of passengers paid yields between half and 1.5 times the industrywide average yield. By comparison,
94 Wir.ios OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION in 1979, the first year after passage of the deregulation act (but before the widespread proliferation of low-cost airlines and discount fares) 63 percent of all travelers paid yields within this range, which is indicative of the more uniform fare levels found under the old regulatory system that was being phased out at the time. Likewise, today more passengers are paying much higher-than-average and much lower-than-average yields than they were 10 years ago. For example, 25 percent of passengers paid more than 2.5 times the industry average yield in 1989, compared with only 19 percent 10 years earlier. Conversely, whereas only 5 percent of passengers paid less than half the industry average yield in 1979, 20 percent are doing so now. These data include the 5 percent of passengers who paid zero fares in 1989 (compared with less than 1 percent in 1979). The recent proliferation of zero fares is due largely to the advent of frequent flier bonus programs. This wider variety of fare and service options, however, did not take the form that was widely anticipated before deregulation, whereby par- ticular airlines or flights would specialize in serving one kind of traveler, such as no-frill, low-cost carriers serving primarily leisure travelers. Peo- ple Express, World, and Capital airlines are examples of the type of single- class carriers that were predicted but that failed to compete effectively with carriers that offered a variety of fare options, usually on the same flight. In contrast to most predictions before deregulation, the main service difference between high- and low-fare passengers is the degree of schedule flexibility provided, and not in-flight amenities. Hence, instead of spe- cializing in single-class service, incumbent carriers introduced a wide array of ticket restrictions that allow them to offer less-restricted, higher-priced tickets to travelers who value frequent flights and greater flexibility in scheduling, while offering more-restricted, discount fares to travelers who are willing to fly during off-peak hours and book flights weeks in advance. These restrictions, which include advance-purchase requirements, refund and exchange penalties, and time-of-day and day-of-week provisions, often result in widely different fares paid by passengers seated side by side and headed for the same destination (Schwieterman 1985). However, even though the in-flight services enjoyed by two passengers may be similar, the total service that is provided may be quite different depending on the degree of flexibility in scheduling. Schedule flexibility is especially valuable to business travelers, who are often unable to make travel arrangements weeks in advance and are therefore willing to pay higher, full fares in return for greater flight frequencies and last-minute bookings and flight changes. For airlines, however, this type of service is more expensive to provide because of last-minute cancellations and the
Passenger Fares and Airline Service 95 larger number of flights that must be scheduled. Providing this type of schedule flexibility often results in more empty seats that must be paid for by charging higher fares to passengers who demand this type of service. In contrast, leisure travelers, who are often willing to purchase tickets weeks in advance and accept limited schedule flexibility in return for lower fares, are less costly to airlines and therefore can be offered less- expensive, discounted fares. In retrospect, many of these ticket restrictions have resulted in the expansion of fare and service options that was an intended consequence of deregulation. Yet it is not clear that all of these restrictions have had this effect. For instance, fare discounts with minimum-stay provisions, such as Saturday-night stay overs, that do not require off-peak travel are not based on any apparent differences in the cost of providing service to complying and noncomplying passengers; instead, they are designed to separate price-sensitive travelers (primarily leisure travelers) from time- sensitive travelers (primarily business travelers), for whom weekend stay overs are seldom a practical option. Pricing practices such as these are often labeled as "price discriminatory" by economists because they are designed to segment travelers by their individual demand for air travel and charge each according to the amount he or she (or the actual purchaser) is willing to pay and not the actual cost of the service. Unfortunately, postderegulation trends in the distribution of fares paid by different types of travelers (e.g., price-sensitive versus time-sensitive) on the same flight cannot be determined using the available fare data. One possible substitute for this analysis, however, is to compare yields paid by travelers in the same kinds of markets on the same kinds of flights. For example, Figures 3-5 and 3-6 show average yield distributions in samples of long-haul and short-haul city-pair markets on direct flights in 1979 versus 1989. Presumably, individual travelers in their respective markets enjoy essentially the same kinds of in-flight service (direct flights headed to the same destination), although they may be paying widely different fares, depending on their individual ticket restrictions. For this analysis, the share of travelers paying different percentages of the average yield in their market was calculated for each of the city pairs, and then averaged for the entire city-pair group. As might be expected, yield dis- tributions were much wider in 1989 than 10 years earlier. For example, in 1979, 80 percent of passengers on long-haul direct flights (more than 2,000 miles) and 93 percent of passengers on short-haul direct flights (fewer than 500 miles) paid between half and 1.5 times the market average yield; by comparison, only 68 and 78 percent, respectively, did so in 1989.
Percent of Trips 50 40 30 20 50% 150% 10 0 0 1-29 30-49 50-69 70-89 90-109 110-29 130-49 150-69 170+ Percent of Market Average Yield 1979 = 1989 FIGURE 3-5 Yield dispersion for short-haul direct flights, 1979 versus 1989 (zero fares not included in calculation of average yield) [DOT O&D survey (third quarters of 1979 and 1989)]. Percent of Trips 40 35 30 25 20 15 10 5 0 50% 150% 0 1-29 30-49 50-69 70-89 90-109 110-29 130-49 150-69 170+ Percent of Market Average Yield 1979 = 1989 FIGURE 3-6 Yield dispersion for long-haul direct flights, 1979 versus 1989 (zero fares not included in calculation of average yield) [DOT O&D survey (third quarters of 1979 and 1989)].
Passenger Fares and Airline Service 97 Average Yield (cent8/mile) ($1989) 50 45 40 35 30 25 20 15 10 1 2 3 4 5 6 7 8 910 14 18 22 26 30 34 38 42 46 50 Length of Trip (hundred8 of miles) - 1979 - 1989 FIGURE 3-7 Yield taper, 1979 versus 1989 [DOT O&D survey (third quarters of 1979 and 1989)]. Fares by Market Distance and Density Perhaps the most frequently predicted consequence of deregulation was a reduction in fares and resultant increase in demand for travel in longer- distance and higher-density markets. Under the old regulatory framework, fares in many of these markets were intentionally held above prevailing costs, because carriers were expected to use the resulting profits to cross- subsidize required service on shorter-haul, lower-density routes (on which fares were often held below prevailing costs). Hence, once given the freedom to adjust their prices and realign their networks, carriers were expected to cut fares and expand service in many longer-haul, heavily traveled markets while raising fares and reducing service in many shorter- haul, lightly traveled routes. Indeed, shortly after deregulation, several studies confirmed these pre- dictions. For example, in analyses of yields on nonstop routes in 1977 and 1983, both the GAO and the CAB found pronounced real yield declines on most long-haul, high-density routes compared with slight yield in- creases in most short-haul, low-density markets (CAB 1984; GAO 1985). Likewise, in a sample of more than 300 city-pair markets (for 19.77 and 1984), Meyer and Oster (1987) found that real yields fell in more than two-thirds of the longer-haul markets compared with only one-third of the short-haul markets. To update and expand these previous analyses, Figure 3-7 shows average yields (in 1989 dollars) for all O&D surveyed passengers for the third
98 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION Percent Change 20 15 10 -5 -10 -15 -20 -25 -30 -35 -40 1 2 3 4 5 6 7 8 910 14 18 22 28 30 34 38 42 48 50 Length of Trip (hundred8 of miles) FIGURE 3-8 Percent change in average real yield by trip distance, 1979 to 1989 [DOT O&D survey (third quarters of 1979 and 1989)]. quarters of 1979 and 1989, broken down into 100- and 200-mile blocks.2 Again, as in previous studies, these data indicate that real fares have fallen markedly for long-distance travel, but have increased for many short trips. For example, as illustrated in Figure 3-8, average yields for most market distances exceeding 1,000 miles declined in real terms by 10 to 30 percent, whereas yields in shorter-distance, 500- to 800-miles markets increased by approximately one-tenth. Accordingly, these fare changes have been accompanied by changes in national travel patterns. For example, between 1979 and 1989, the total number of passenger trips taken nationally grew by roughly 50 percent, yet travel in longer-distance markets (more than 1,500 miles) jumped by approximately two-thirds (Figure 3-9). Hence, approximately one-fourth of domestic air travel now occurs in markets of 1,500 miles or more (Figure 3-10). Conversely, Figure 3-11 shows the change (between 1979 and 1989) in travel by market density. These data indicate that the largest increases in passenger trips occurred in the highest-density markets of 800 or more passengers per day. Between 1979 and 1989, the number of passengers in these markets grew by more than 60 percent. Although part of this growth was undoubtedly due to more passengers in the natioti's tradi- tionally busiest air markets (such as New YorkâLos Angeles), a large part also can be attributed to increased travel in somewhat smaller markets (such as Minneapolis-Phoenix), which also have experienced sharp in-
Percent Change 70 60 50 40 30 20 10 0 1-499 500-749 750-999 1000-1499 1500-1999 2000. Market Dt8tance (mile8) FIGURE 3-9 Percent change in passenger trips by market distance, 1979 to 1989 [DOT O&D survey (third quarters of 1979 and 1989)]. Percent Share 35 I 30 25 20 15 10 5 0 1-499 500-749 750-999 1000-1499 1500-1999 2000. Market Distance (miles) FIGURE 3-10 Share of passenger trips by market distance, 1989 [DOT O&D survey (third quarter of 1989)].
100 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION 800. 400-799 200-399 100-199 50-99 1-49 Pa88enger8 per Day (Round-Trip Market8) FIGURE 3-11 Change in passenger trips by market density, 1979 to 1989 [DOT O&D survey (third quarters of 1979 and 1989)]. creases in air travel demand. Overall, the nation's busiest markets now account for about three-fourths of all passenger trips compared with two- thirds 10 years earlier (Figure 3-12). Summarized in Table 3-2 are real yield changes by market distance and density combined. Again, the results from this cross tabulation consistently show that travelers in longer-distance markets have enjoyed 10 41 \It- i& FIGURE 3-12 Share of passenger trips by market density, 1979 versus 1989 [DOT O&D survey (third quarters of 1979 and 1989)].
Passenger Fares and Airline Service 101 TABLE 3-2 CHANGE IN AVERAGE REAL YIELD BY MARKET DISTANCE AND DENSITY, 1979 TO 1989 Change in Real Yield (%) by Market Distance (miles) Passenger Trips I- 500- 750- 1,000- 1,500- per Day° 499 749 999 1,499 1,999 2,000+ 800+ -3.1 + 10.4 -0.2 -10.7 -17.2 -18.2 (28.3) (10.2) (9.9) (11.3) (6.6) (10.6) 400-799 +31.6 + 14.0 -3.4 -16.2 -26.6 -20.1 (1.5) (1.0) (1.3) (1.4) (0.9) (1.4) 200-399 +21.9 + 16.0 +1.5 -15.0 -25.0 -19.6 (0.9) (0.9) (0.9) (1.2) (0.8) (1.0) 100-199 +23.4 + 18.5 +0.3 -17.3 -23.4 -21.4 (0.6) (0.7) (0.6) (0.6) (0.6) (0.6) 50-99 +24.8 +17.1 +1.4 -17.2 -23.5 -20.2 (0.4) (0.5) (0.5) (0.6) (0.4) (0.4) 1-49 +24.2 + 16.3 ±0.1 -17.3 -26.4 -22.8 (0.6) (0.7) (0.6) (0.9) (0.5) (0.7) NOTE: Figures in parentheses indicate percentage of total trips in cell during 1989 Total passenger trips in both directions of a city-pair market. SOURCE: O&D survey (third quarters of 1979 and 1989). the largest yield decreases since deregulation; real yields have fallen by 10 to 25 percent. Yet changes in real yield have varied relatively little across market densities. This is especially true in longer-haul markets, which have benefited from the traffic efficiencies created by hub-and-spoke systems. As discussed in Chapter 2, these systems allow airlines to funnel traffic from many lower-density markets, facilitating higher load factors, which generates lower fares in both higher- and lower-density markets alike. Because of the ability of hub-and-spoke systems to consolidate traffic from many smaller markets, market size has less influence on the price of air travel than it would if direct, point- to-point service was more prevalent. Accordingly, benefiting least from these hub-and-spoke systems are travelers in short-haul markets, which are usually linked by point-to- point service because of the short distances involved. Since deregula- tion, real yields have increased by 10 to 25 percent in many short-haul markets, as unregulated fares have been adjusted upward to reflect true industry costs (instead of cross subsidies). One important exception to this rule, however, has occurred in high-density short-haul markets, which ordinarily do not need hub-and-spoke systems to provide the higher load factors that are necessary to generate lower fares. Accord-
102 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION TABLE 3-3 CHANGE IN AVERAGE REAL FARES BY MARKET DISTANCE AND DENSITY, 1979 TO 1989 Change in Average Real Fare (in dollars) by Market Distance (miles) Passenger Trips 1- 500- 750- 1,000- 1,500- per Daya 499 749 999 1,499 1,999 2,000+ 800+ -2.33 +14.25 -0.35 -21.0 -42.0 -46.58 400-799 +18.33 +20.13 -5.69 -34.13 -73.33 -55.80 200-399 + 19.75 +21.43 +2.42 -32.25 -66.85 -52.88 100-199 + 19.82 +25.56 +0.61 -38.62 -62.83 -59.40 50-99 + 19.90 +24.56 +2.45 -39.13 -63.88 -56.93 1-49 +20.35 +24.81 +0.18 -42.88 -78.40 -69.98 Total passenger trips in both directions of a city-pair market SOURCE: O&D survey (third quarters of 1979 and 1989). ingly, real yields, which have generally increased in most short-haul markets, have declined by about 3 percent in these much denser short- haul routes. Overall, the findings of this analysis imply that most, but not all, passengers have enjoyed real fare reductions since deregulation. Table 3-2 identifies the share of passengers in each market distance and density category in the third quarter of 1989, as well as changes in average real yield since the third quarter of 1979. Altogether, 79 percent of travel occurred in markets that have experienced real yield declines since de- regulation. Clearly, the principal beneficiaries have been travelers in medium-to-long-distance markets (more than 750 miles), which alone account for 54 percent of all trips. There is no simple way to measure the total dollar savings achieved by passengers in these markets; however, as shown in Table 3-3, because fares in absolute terms are highest for long- distance travel, even small percentage declines in yield can add up to substantial fare savings. EFFECTS OF COMPETITION ON FARES AND SERVICE In theory, the removal of route restrictions after deregulation was sup- posed to stimulate the entry of new firms into the airline industry and cause existing airlines to expand or shift their operations into other, more profitable markets, thereby forcing fares down and expanding service options in markets in which carriers had previously enjoyed little competition. Indeed, as predicted by theory, most established
Passenger Fares and Airline Service 103 carriers greatly enlarged their networks shortly after deregulation, and many new firms entered the marketplace, creating a competitive en- vironment that produced much lower fares during the mid-1980s. Since 1987, however, virtually all these new entrants have either failed or merged with the larger incumbent carriers, whereas passenger fares have coincidentally started to rise. Hence, the discussion in this section deals specifically with trends in passenger fares and service by the level of competition. The section begins with analyses to determine whether there might be some correlation between competition and fares in general. A sample of 795 city pairs was analyzed using simple cross-tabulations. Each of these city pairs was grouped by market distance and density and then characterized by the number of competing carriers. The purpose of this analysis was to determine how fare levels are likely to vary by the number of competitors in a market and how levels of market competition have been changing over time. The discussion then turns to more specific concerns about fare and service levels in hub markets, which are widely believed to be less competitive than nonhub markets because of the strong position of many hubbing carriers. Accordingly, hub fares, services, and levels of competition are compared with those of nonhubs to determine if significant differences exist that might be explained by market dominance. In many ways these comparisons com- plement those of previous studies that have attempted to estimate the impact of hub concentration on fares and service. Hence, findings from this earlier work are discussed in a prelude to this analysis. Systemwide Benefits In order to determine general levels of competition within individual mar- kets, the 795-city-pair sample was disaggregated by the number of effec- tive competitors, while controlling for market distance and density. In 1988-1989, 44 percent (356) of the sampled city pairs was served by three or more competing carriers, 34 percent (267) was served two carriers, and 22 percent (172) was dominated by a single carrier (Table 3-4). Overall, market distance was the single most important determinant of competition, particularly because longer distances allow for increased rivalry between competing hub-and-spoke systems. As a group, 74 percent (85 of 115) of long-distance (more than 2,000 miles) city pairs was served by three or more carriers, whereas only 9 percent (10 of 115) was served by a single carrier. Conversely, only 18 percent (38 of 205) of short-haul (more than 500 miles) markets were multicarrier markets (served by 3 or
TABLE 3-4 AVERAGE YIELD BY MARKET DISTANCE, DENSITY, AND COMPETITION FOR CITY-PAIR SAMPLE, 1988- 1989 Market Distance (miles) 100-499 500-999 1,000-1,499 1,500-1,999 2,000 + Passengers Number of Yield (cents Sample Yield (cents Sample Yield (cents Sample Yield (cents Sample Yield (cents Sample Per Daya Competitors per mile) size per mile) size per mile) size per mile) size per mile) size 800+ 1 20.0 1 23.5 2 15.0 I 11.0 3 12.0 2 26.0 12 23.0 4 13.7 10 12.0 9 10.0 2 3 24.3 17 21.3 19 12.6 9 11.0 3 9.2 12 400-799 1 34.9 7 25.0 4 15.3 4 11.7 3 10.5 2 2 35.6 II 19.6 14 13.0 5 10.9 9 9.3 3 3 29.2 7 14.9 7 11.2 6 10.9 8 9.5 10 200-399 1 37.4 19 23.4 10 10.0 3 12.0 2 11.3 3 2 37.8 5 20.3 16 13.4 8 9.4 7 7.0 1 3 43.4 1 20.2 4 12.1 9 10.6 II 9.3 Il 100-199 I 39.3 22 22.1 10 15.0 2 12.0 3 10.5 2 2 38.1 9 21.4 10 15.8 5 9.3 3 10.0 1 3 25.0 4 20.4 10 14.8 8 10.7 14 10.4 12 50-99 1 39.0 24 22.2 9 NA NA 12.0 I NA NA 2 45.6 8 21.3 14 14.0 3 11.0 6 10.2 6 3 28.7 3 19.6 17 12.9 22 11.8 18 9.5 19 25-49 I 39.8 19 21.9 9 12.7 3 14.0 1 11.5 2 2 39.2 3 24.9 28 12.4 10 11.5 11 9.9 7 3 32.3 6 20.0 18 12.0 27 10.3 23 9.9 21 All I 38.4 92 22.7 44 13.4 13 11.8 13 11.1 10 densities 2 37.0 75 20.5 86 13.5 41 10.9 45 9.8 20 3 27.4 38 19.8 75 12.5 81 10.9 77 9.7 85 NOTE: Data are for the last two quarters of 1988 and the first two quarters of 1989. NA = None available in sample 'Total passenger trips in both directions of a city-pair market. SouRce: DOT O&D survey.
Passenger Fares and Airline Service 105 more carriers), whereas 45 percent (92 of 205) was dominated by one carrier. In addition, market density was also an important determinant of competition, as 57 percent (60 of 105) of the highest density markets (800 or more passengers a day) were served by three or more carriers, compared with only 19 percent (40 of 215) of the lowest density markets (25 to 49 passengers a day). However, as market distance increased, competition tended to increase irrespective of market density; for in- stance, 70 percent (21 of 30) of very low density (25 to 49 passengers a day), long-haul (2,000 or more miles) markets was served by three or more carriers, which is similar to the levels of competition in much larger markets. Not surprisingly, the data indicate that the least competitive markets typically have the highest yields. Altogether, single-carrier markets had the highest average yields in 18 of the 30 distance and density market categories examined (Table 3-4). Among the highest-density markets (800 or more passengers a day), single-carrier markets had the highest average yield in 3 of 5 distance groupings. Similarly, among the lowest-density markets (25 to 49 passengers a day), single-carrier markets had the highest average yield in 4 distance categories. In general, average yields in two- carrier markets were not significantly different from those in the single- carrier markets, usually only 2 or 3 percentage points lower. Typically, the multicarrier markets had the lowest average yieldâoften 10 to 35 percent lower than those in one- and two-carrier markets. Overall, the multicarrier markets had the lowest (or tied for the lowest) average yield in 21 of the 30 market categories examined. In recent years, as the airline industry has moved toward greater con- centration at the national level, concern has risen about the potential for increasingly more concentration at the market level, resulting in more one- and two-carrier city pairs. Accordingly, to determine how levels of com- petition in city pairs have changed over time, the 795-city-pair sample was also analyzed for three different years-1979, 1984, and 1988 to 1989. The number of markets was compared with three or more com- petitors during each time period. Table 3-5 presents the results of this analysis for the higher-density markets sampled, which together account for more than three-quarters of all passenger trips. Once again, the data indicate that travelers in the longer-distance markets enjoyed the greatest competition in all 3 years; however, the most striking finding is the wide variation in competition from year to year. In the most recent year, 1988 to 1989, 40 percent of the city pairs had three or more effective compet- itors. This figure compares favorably with 20 percent in 1979, but some- what less favorably with 53 percent in 1984.
TABLE 3-5 SHARE OF HIGHER-DENSITY CITY PAIRS SAMPLED WITH THREE OR MORE EFFECTIVE COMPETITORS, 1979, 1984, AND 1988-1989 Passengers per Day° Year Percentage of Markets by Distance (miles) 100-499 500-999 1,000-1,499 1,500âI ,999 2,000 + Total 800+ 1979 27 (8/30) 20 (5/25) 15 (3/20) 20 (3/15) 33 (5/15) 23 (24/105) 1984 67 (20/30) 56 (14/25) 80 (16/20) 60 (9/IS) 73 (11/15) 67 (70/105) 1988-1989 57 (17/30) 44 (11/25) 45 (9/20) 20 (3/15) 80 (12/IS) 49 (52/105) 400-799 1979 24 (6/25) 16 (4/25) 27 (4/15) 15 (3/20) 30 (5/15) 22 (22/100) 1984 32 (8/25) 28 (7/25) 40 (6/15) 85 (17/20) 87 (13/15) 51(51/100) 1988-1989 28 (7/25) 28 (7/25) 40 (6/IS) 40 (8/20) 67 (10/15) 38 (38/100) 200-399 1979 4 (1/25) 13 (4/30) 15 (3/20) 20 (4/20) 27 (4/15) 15 (16/110) 1984 8 (2/25) 13 (4/30) 45 (9/20) 80 (16/20) 73 (11/15) 41(45/110) 1988-1989 4 (1/25) 13 (4/30) 45 (9/20) 55 (11/20) 73 (11/15) 33 (36/110) Total 1979 19 (15/80) 16 (13/80) 18 (10/55) 18 (10/55) 31(14/45) 20 (62/315) 1984 38 (30/80) 35 (28/80) 56 (31/55) 76 (42/55) 78 (35/45) 53 (166/315) 1988-1989 31(25/80) 28 (22/80) 44 (24/55) 40 (22/55) 73 (33/45) 40 (126/3 IS) NOTE: Data are for the last two quarters of 1988 and the first two quarters of 1989. The numbers in parentheses represent the number of city pairs in each category. 'Total passenger trips in both directions of a city-pair market. SOURCE: DOT O&D survey.
Passenger Fares and Airline Service 107 In retrospect, these findings are not surprising. In 1979âthe year fol- lowing deregulationâonly a handful of new firms had fully entered the airline business, and most incumbent carriers were just starting to expand their network of routes. Hence, many markets were still being served by the same carriers that had dominated under the traditional route-franchise system administered by the CAB. Yet by 1984, numerous airlines had already entered the industry, and most incumbent carriers had expanded their networks. However, the intense levels of competition in the mid- 1980s were probably unsustainable in the long run and therefore offer an unrealistic point of comparison for today's industry. By 1989, well after most new entrants had failed, the remaining airlines were essentially the same ones that had been operating 10 years earlier, although they had greatly expanded their networks and were competing more with each other than at any time before 1979. Most importantly, the vast majority of all passengers are still flying in competitive markets, because most of the country's highest-density markets continue to be served by three or more carriers. According to esti- mates by the Air Transport Association of America (ATA), about two- thirds of all passengers traveled in multicarrier markets in 1988 (ATA 1989, Table 2). Consequences of Hub Dominance As discussed in Chapter 1, a prominent feature of the airline industry since deregulation has been the proliferation of hub-and-spoke systems. Although these systems are not a new concept, the removal of route restrictions after deregulation greatly stimulated their development when carriers quickly recognized the efficiencies that could be achieved by consolidating traffic flows from many spoke airports through a central hub airport. Another consequence of hubbing, however, has been the growing dom- inance of major carriers at their hub airports. Examples of highly domi- nated hubs include Charlotte, North Carolina, where the hubbing carrier, USAir, accounts for more than three-fourths of local passengers; Cincin- nati, Ohio, where Delta carries nearly 70 percent of local traffic; and Minneapolis, Minnesota, where Northwest handles nearly two-thirds. As these and other airlines have strengthened their positions in hub markets, increasingly more attention has been focused on their ability to build local monopolies and earn revenues, or rents, in excess of what might be possible under more competitive market conditions.
108 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION Previous Studies of Hub Pricing The issues associated with hub dominance have been examined in several studies. For example, Levine (1987) and Bailey and Williams (1988) described the role of airport dominance, along with economies of scope and scale, in generating rents to air carriers. They contend that in recent years the tendency toward hub dominance has blocked further competition in local hub markets, for example, by locking up available gate space in long-term lease arrangements and distorting local travel agency incentives. The impact of hub dominance on the actual fares paid by travelers in hub markets has been estimated in other recent studies. For example, Borenstein (1989) examined the effect of hub concentration on fares using data for a sample of city-pair markets that included the 200 largest airports. Borenstein's model attempted to explain yields as a function of the pro- portion of travelers in a city-pair market carried by a single carrier, the total share of travelers arriving at the destination airport carried by a single carrier, trip distance, and several other variables related to service quality. The variables measuring market concentration were found to be statisti- cally significant, and a 10-percent increase in local market share was found to correlate with a 4.6-percent increase in yield. Borenstein sug- gested that several other airline marketing features, such as frequent flier programs, computer reservation systems (CRS5), and travel agent incen- tives, may also partly explain the rents earned by dominant hubbing car- riers. In addition, Borenstein noted that the benefits provided by hub operations to local travelers, such as more nonstop flights and greater flight frequency, could justify the higher fares at hubs, although no attempt was made to quantify this trade-off. In a simpler empirical analysis, the GAO compared average yields earned at 15 concentrated airports with those in 38 unconcentrated airports (GAO 1990). Concentrated airports were defined as those at which 60 percent of passengers was handled by a single carrier or 85 percent was handled by two carriers. For each airport, the average yield was calculated for all originating passengers flying to all domestic points. The average yields in each of the 38 unconcentrated airports were then averaged for the entire group. In 1988 this figure was 14.5 cents per mile. In com- parison, yields at the 15 concentrated airports averaged 18.5 cents per mile, or about 27 percent higher. In response to these findings, the ATA noted that part of this yield differential could be attributable to shorter trip distances from the hub cities sampled (GAO 1990). By itself, trip length has an important influ- ence on yield because travelers from hub cities tend to fly shorter distances than travelers from nonhub cities (partly because of the central location
Passenger Fares and Airline Service 109 of most hubs), and short-distance travel is usually more costly per pas- senger mile to provide. Accordingly, to control for this influence the GAO excluded markets with average trip lengths of 900 miles or more, which reduced the yield differential from 27 to 20 percent. Shortly after the GAO study, the DOT completed a widescale study of the postderegulation airline industry, in which the issue of pricing at concentrated hubs also was analyzed (DOT 1990b). In this study, eight single-carrier and eight two-carrier hubs were identified. Average yields in these 16 concentrated hubs were then compared with the industrywide average yield. After adjustment for market density and distance factors, average yields were found to be 18.7 percent higher in the single-carrier hubs and 8.9 percent higher in the two-carrier hubs. These estimates of higher yields at hubs are lower than those estimated by the GAO. In a recent working paper, Dresner and Windle (1991) implied that on com- parable routes the hub premium may be even smaller; their estimate of the hub premimum was about 3 percent. In its conclusions, the DOT did not directly attribute higher yields it found at hubs to market concentration, although it expressed concern about the lack of new entry in many large city-pair markets that have concentrated hubs and the adverse effect that this might have on market pricing (DOT 1990b, 151). Alternative Analyses Two approaches were taken in this study to examine the effect of hub concentration on passenger fares and service. First, each of the 795 city pairs sampled earlier was grouped into the categories of concentrated hub, unconcentrated hub, and nonhub. After market distance and density were controlled for, the number of effective competitors and average yield differentials were determined for each of the three market groups. In the second approach, passenger trip data were collected for each of the 60 largest U.S. cities (instead of city-pair markets). These cities also were categorized as concentrated hub, unconcentrated hub, and nonhub cities. For each of the three groups, average yield and measures of service were calculated and compared. For both of these analyses, when markets were defined, data from all airports were combined for cities located within 50 miles of each other, such as Miami and Fort Lauderdale, and for cities that have more than one airport, such as Chicago, Houston, and New York. Presumably, travelers in these cities are less likely to be adversely affected by the strong position of a carrier at its hub because competition is available elsewhere
110 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION in the metropolitan area. This distinction provides an alternative approach to the DOT study, which emphasized carrier dominance at a single airport market instead of the entire metropolitan market, and the GAO study, which excluded several major cities with more than one airport from its sample of markets. In addition, the definition of market concentration used in this study differs from those used by the GAO and the DOT. In both of those studies total enplanement share of a hubbing carrier was used as the criterion for defining market concentration. By their very nature, hub-and-spoke sys- tems increase a hubbing carrier's share of total enplanements, simply because more of its passengers transfer to spoke cities by way of the hub. Accordingly, in this study, market concentration is defined on the basis of local enplanement share, whereby concentrated markets are defined as cities in which at least half of local enplanements is handled by one carrier or a total of 75 percent is handled by two carriers.3 Distinctions between hubs and nonhubs are based on the proportion of connecting enplanements in the city. Cities in which connecting passengers account for more than 40 percent of total enplanements are defined as hubs. Altogether, 15 cities were classified as hubs. Ten of these hub citiesâ Atlanta, Georgia; Charlotte, North Carolina; Cincinnati and Dayton, Ohio; Denver, Colorado; Memphis, Tennessee; Minneapolis, Minne- sota; Pittsburgh, Pennsylvania; Salt Lake City, Utah; and St. Louis, Missouriâwere defined as concentrated, and the remaining fiveâChi- cago, Illinois; Dallas, Texas; Detroit, Michigan; Nashville, Tennessee; and Raleigh-Durham, North Carolinaâwere defined as unconcen- trated. These groupings were based on data from 1988 and 1989, which correspond to the time periods used in the GAO and the DOT studies, although some analyses were later updated through the first two quarters of 1990. City-Pair Analysis Altogether, 226 of the 795 city pairs sampled were characterized as concentrated-hub markets because they involved at least 1 of the 10 con- centrated hubs. Of the remaining 569 city pairs, 99 were characterized as unconcentrated-hub markets because they involved at least 1 of the 5 unconcentrated hubs, and 470 were characterized as nonhub markets, involving nonhub cities only. The first analysis of the sample involved a comparison of the proportion of city pairs dominated by one carrier in each of the three types of markets.
Passenger Fares and Airline Service TABLE 3-6 SHARE OF CITY PAIRS SAMPLED WITH ONLY ONE EFFECTIVE COMPETITOR BY MARKET DISTANCE, HUBS VERSUS NONHUBS, 1988-1989 Distance (miles) Percentage of City Pairs by Market Type Concentrated Hub Unconcentrated Hub Nonhub Total 100-499 65 (4 1/63) 36 (10/28) 36 (41/114) 45 (92/205) 500-999 40 (27/68) 30 (8/27) 8 (9/110) 21 (44/205) 1,000-1,499 23 (7/31) 10 (2/20) 6(4/84) 10(13/135) 1,500-1,999 18 (9/49) 5 (1/20) 5 (3/66) 10 (13/135) 2,000+ 53(8/15) 0(0/4) 2(2/96) 9(10/115) Total 41(92/226) 21(21/99) 13 (59/470) 22 (172/795) NOTE: Data are for the last two quarters of 1988 and the first two quarters of 1989. The numbers in parentheses represent the number of city pairs in each category. SOURCE: DOT O&D survey. Such city pairs were defined as those in which no other competing carrier handled more than 10 percent of local passenger traffic (a standard def- inition in studies of this type). As shown in Table 3-6, the share of city pairs in the three market categories differed greatly. As might be expected, the most competitive markets tended to be on longer-distance routes, on which travelers can usually choose among several competing hub-and- spoke systems. The most notable finding, however, is that in all cases the proportion of city pairs dominated by one carrier was much higher among concentrated hubs than in the other market groups. Altogether, 41 percent of the concentrated-hub city pairs were dominated by single car- riers compared with 21 and 13 percent of the unconcentrated-hub and nonhub pairs, respectively.4 Next, in order to determine if these concentrated-hub markets have higher yields, the difference in average yield was calculated for the concentrated-hub pairs versus all other pairs combined (nonhubs and un- concentrated hubs). As shown in Table 3-7, the results from this analysis show a pattern of consistently higher yields in the concentrated-hub city pairs than in the comparison pairs, regardless of market distance or density. Altogether, yields were higher in the concentrated-hub city pairs in 26 of the 30 market categories, strongly indicating that concentrated-hub cities have higher-than-average fares. Besides market concentration, other differences between hubs and non- hubs are likely to play an important role in these yield differentials. For example, because hubs are seldom located in tourist areas (such as Cal- ifornia, Nevada, and Florida), the nonhub city-pair sample may contain more tourist markets than the hub sample. As discussed earlier, pleasure travelers are often willing to fly during off-peak hours and make additional
112 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION TABLE 3-7 DIFFERENCE IN AVERAGE YIELD IN CONCENTRATED-HUB CITY PAIRS VERSUS ALL OTHERS, 1988-1989 Difference in Yield (%) by Distance (miles) Passengers Sample 100â 500â 1,000â 1,500- per Day" Size 499 999 1,499 1,999 2,000 + 800+ 105 +27 +20 +2 +9 +13 400-799 100 +20 +42 â5 +10 +7 200-399 110 +6 +5 +1 +3 +24 100-199 115 +11 0 â5 +13 +10 50-99 150 +9 +4 +30 â2 +20 25-49 215 +9 +20 +15 +16 +15 NOTE: Cities in which more than half of local passenger traffic is handled by one carrier or more than 75 percent is handled by two carriers are defined as concentrated cities. Cities in which connecting traffic accounts for more than 40 percent of passenger enpianements are defined as hubs. (Cities in Alaska and Hawaii are excluded for this analysis because of their unusual geographic characteristics.) Data are for last two quarters of 1988 and first two quarters of 1989. "Total passenger trips in both directions of a city-pair market. SOURCE: DOT O&D survey. connections in exchange for lower fares, which may contribute to lower average yields in the nonhub sample. Also, nonhub travelers are offered more flight options (greater schedule frequency) and tend to fly on fewer nonstop flights than travelers in hub markets, partly because less nonstop service is available. Presumably, this difference in service also contributes to higher average yields in hub markets. Hence, in order to control for these influences, a follow-up analysis was conducted in which average yield differentials in only the concentrated and unconcentrated-hub city pairs were compared. Presumably, travelers in un- concentrated hubs enjoy service levels somewhat similar to those in con- centrated hubs (e.g., more nonstop service), and therefore yields in the two kinds of markets should be more comparable. However, although compar- isons could not be made across all 30 market categories (because some cells did not contain unconcentrated-hub markets), yields were still higher in the concentrated hubs in 14 of 25 cells tested (Table 3-8). City Analysis In the second analysis, the 60 U.S. cities that have the most originating traffic were disaggregated by market type and characterized as concen- trated hubs, unconcentrated hubs, or nonhubs. The same 10 hub citiesâ Atlanta, Charlotte, Cincinnati, Dayton, Denver, Memphis, Minneapolis, Pittsburgh, Salt Lake City, and St. Louisâwere defined as concentrated,
Passenger Fares and Airline Service 113 TABLE 3-8 DIFFERENCE IN AVERAGE YIELD IN CONCENTRATED-HUB CITY PAIRS VERSUS UNCONCENTRATED-HUB CITY PAIRS, 1988-1989 Difference in Yield (%) by Distance (miles) Passengers Sample 100â 500â 1,000â 1,500â per Day° Size 499 999 1,499 1,999 2,000+ 800+ 55 +29 +7 +7 +5 NA 400-799 54 0 +65 â20 +13 +17 200-399 55 +8 â8 +18 â13 NA 100-199 46 +2 â15 â21 NA NA 50-99 66 +20 â15 â9 0 +10 25-49 47 +11 +46 NA +7 0 Nom: Cities in which more than half of local passenger traffic is handled by one carrier or more than 70 percent is handled by two carriers are defined as concentrated cities. Cities in which connecting traffic accounts for more than 40 percent of passenger enplanements are defined as hubs. (Cities in Alaska and Hawaii are excluded for this analysis because of their unusual geographic characteristics.) Data are for last two quarters of 1988 and first two quarters of 1989. NA = not applicable. Total passenger trips in both directions of a city-pair market. SouRcE: DOT O&D survey. and the same 5 citiesâChicago, Dallas, Detroit, Nashville, and Raleigh- Durhamâwere defined as unconcentrated. The remaining 45 cities were categorized as nonhubs. Together, these 60 U.S. cities accounted for 90 percent of domestic O&D traffic in 1988: the 10 concentrated hubs ac- counted for 12 percent, the 5 unconcentrated hubs accounted for 11 per- cent, and the 45 nonhubs accounted for the remaining 67 percent. For each of these three market categories, average yield was determined by calculating the simple mean of the average yields in all 60 cities. Table 3-9 presents the results of this analysis for 1988, in which the concentrated hubs had an average yield 27 percent higher than that of the nonhubs (17.8 cents per mile versus 14 cents per mile). As shown in Table 3-10, TABLE 3-9 AVERAGE YIELD AND PASSENGER TRIP LENGTH BY MARKET TYPE, 1988 Average Passenger Trip Length Average Yield (One Way) Market Type Sample Size (cents/mile) (miles) Hub Concentrated 10 17.8 832 Unconcentrated 5 16.3 839 Total 15 17.3 835 Nonhub 45 14.0 915 SOURCE: DOT O&D survey
114 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION TABLE 3-to AVERAGE YIELD BY MARKET TYPE AND PASSENGER TRIP LENGTH, 1988 Average Passenger Trip Length (One Way) (miles) Less than 800 800 to 1,000 More than 1,000 Yield Sample Yield Sample Yield Sample Market Type (v/mi) Size (tt/mi) Size (/mi) Size Hub Concentrated 18.7 18.5 14.6 Unconcentrated 18.1 15.2 NA Total 18.5 6 17.1 7 14.6 2 Nonhub 16.0 14 13.6 17 12.3 14 Nom: NA = none applicable SOURCE: DOT 0&D survey. even when average yields were compared across mileage blocks (to control for the influences of average trip length on yield), concentrated hubs still had significantly higher yields than the comparison group of nonhubs. To make this analysis more meaningful, however, an additional step was taken to exclude from the nonhub sample all cities in New England, Florida, Nevada, Washington State, and California, as well as the cities of Philadelphia, Tucson, and New York. Arguably, these cities are not comparable with the hubs because they are either tourist markets (e.g., Las Vegas and Orlando) or coastal cities that have unusually high average trip lengths (more than 1,000 miles) (e.g., Philadelphia and Los Angeles). Both factors cause lower-than-average yields. Indeed, after this additional step was taken, the average yield in the remaining 27 nonhub cities in- creased noticeably from 14 to 15.1 cents per mile, and accordingly, the yield differential between the concentrated-hub and nonhub cities declined. As shown in Table 3-11, this second analysis resulted in an average yield differential of 18 percent and not 27 percent as found previously. This single-year (1988) analysis was updated and expanded using quarterly data from mid-1987 through mid-1990 (Figures 3-13 and 3-14). During this 3-year period, the yield differential between the nonhubs and the concentrated hubs rose from approximately 20 to 25 percent; meanwhile the yield differ- ential betwen the unconcentrated hubs and nonhubs also widened from 11 to 17 percent (Figure 3-13). At the same time, most of the hubs grew increasingly more concentrated, as average local-market shares of the largest hubbing carriers increased from 59 to 65 percent (Figure 3-14). For the entire 12-quarter period, the average yield differential between the concentrated hubs and nonhubs was 21 percent compared with a 14 percent difference between the unconcentrated hub and nonhubs.
Passenger Fares and Airline Service 115 TABLE 3-11 AVERAGE YIELD AND PASSENGER TRIP LENGTH BY MARKET TYPE, 1988 Average Average Passenger Yield Trip Length Sample (cents per (One Way) Market Type Size mile) (miles) Hub Concentrated 10 17.8 832 Unconcentrated 5 16.3 839 Total 15 17.3 835 Nonhub' 27 15.1 887 "All cities in New England, Florida, Nevada, Washington State, and Cal- ifornia, as well as Philadelphia, Tucson, and New York City, were ex- cluded from the analysis because they are either tourist markets or have unusually high average trip lengths (more than 1,000 miles). SOURCE: DOT O&D survey. Part of this yield differential might stem from the fact that hub travelers are provided different, and often superior, levels of service than travelers in most nonhubs. For example, residents of hub cities can choose from about twice as many flights as residents of nonhubs of comparable size (Table 3-12). This is an important benefit for business travelers, who are often willing to pay higher fares in return for more flexible scheduling. Moreover, travelers in hubs enjoy less circuitous routing and more direct Nonhub Sample M Unconcentrated Hubs M Concentrated Hubs FIGURE 3-13 Indexed average yield, hubs versus nonhubs (DOT O&D survey).
Share of Local Traffic by Hub Carrier 80% 70% 60% 50% 40% Yield Differential 30% 25% 20% 15% 10% 30% I1 1 1 1 1 I 15% 87/3 87/4 88/1 88/2 88/3 88/4 89/1 89/2 89/3 89/4 90/1 90/2 Year/Quarter -*-- Concentrated-Hub Market Share -- Unconcentrated-Hub Market Share Concentrated-Hub Yield Differential Unconcentrated-Hub Yield Differential FIGURE 3-14 Market concentration and yield differential at hubs (compared with nonhub sample) (DOT O&D survey). TABLE 3-12 AVERAGE ENPLANEMENTS, TRIP CIRCUITY, AND DEPARTURES PER PASSENGER TRIP BY MARKET TYPE, 1988 Average Departures Enplane- per Share of ments Thousand Local per Average Local Passengers Sample Round Trip Enplane- Flying Market Typea Size Trip Circuity ments Direct (%) Hub Concentrated 10 2.5 1.04 40.2 75 Unconcentrated 5 2.6 1.05 30.8 69 Total 15 2.5 1.04 37.1 73 Nonhub Concentrated 8 3.0 1.05 23.3 52 Unconcentrated 37 2.8 1.05 19.8 62 Total 45 2.8 1.05 20.4 60 'Cities in which more than half of local passenger traffic is handled by one carrier or more than 75 percent is handled by two carriers are defined as concentrated cities. Cities in which connecting traffic accounts for more than 40 percent of passenger enpianements are defined as hubs. (Cities in Alaska and Hawaii are excluded from this analysis because of their unusual geographic characteristics.) SouRcE: DOT O&D survey and RSPA Form 41.
Passenger Fares and A jr/inc Service 117 service than travelers in most nonhubs; for instance, travelers in hub cities fly direct on nearly 75 percent of their trips, whereas passengers in nonhubs do so only 60 percent of the time (Table 3-12). For some travelers, such as leisure travelers, this service differential might not be important, and the higher price (higher fares) may be more than what they would otherwise demand if less service was available. However, for many other travelers, such as business travelers, this service is likely to be valuable.5 A rough method for determining whether higher yields in concentrated- hub cities are in fact compensation for superior hub service is to compare the yield differentials of nonhubs and unconcentrated hubs. Indeed, the data suggest that these service differences might account for some of the 21 percent yield differential, although perhaps not all of it. As shown in Figure 3-13, the difference in average yield between unconcentrated hubs and nonhubs was 14 percent (during the 3-year period from 1987 to 1990), compared with a 21 percent differential between concentrated hubs and nonhubs. Accordingly, for the period examined, yields at concentrated hubs were 5 to 10 percentage points greater than yields at unconcentrated hubs 6 In summary, the bulk of the evidence, both from this study and the empirical work already mentioned, consistently indicates higher yields at concentrated hubs. However, whether all or part of this yield differential stems from monopoly profits earned by carriers exercising market power is unclear. Findings presented earlier in this chapter, which show 10 to 35 percent higher yields in one- and two-carrier city-pair markets and a greater prevalence of these markets in concentrated hubs, suggest that concentration could be a factor. Proving this empirically, however, would require a detailed comparison of how hub prices correspond to hub op- erating costs while controlling for a variety of explanatory variables. The many difficulties associated with allocating costs to specific services in a multiservice industry, the proprietary nature of cost data, and the data collection required for multivariate analyses of the kind required preclude such a detailed analysis in this study. A closer look at the cities defined as hubs in this study reveals that the highest yields tend to be in the smallest hubs. For example, through the first two quarters of 1990, the seven hubs that have the least amount of local passenger traffic (Charlotte, Cincinnati, Dayton, Memphis, Nashville, Raleigh-Durham, and Salt Lake City) had an average yield that was roughly 20 percent higher than that of the eight hubs that have the most amount of local passenger traffic (Atlanta, Chicago, Dallas, Denver, Detroit, Minne- apolis, Pittsburgh, and St. Louis) (Table 3-13). Part of this difference is explained by the lower traffic volumes in the smaller hubs, as well as the tendency of travelers in those markets to fly shorter distances (Table 3-13).
118 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION TABLE 3-13 MARKET DATA FOR LARGE VERSUS MEDIUM HUBS, 1988 Change in Sample Average Yield Average Trip Departures (%) Hub Size Size (cents per mile) Length (miles) (1979-1988) Large 8 15.9 890 +31 Medium 7 18.9 804 +131 SOURCE: DOT O&D survey and RSPA Form 41 However, the medium-sized hubs also tend to be the most highly concentrated and readily dominated by a single carrier. For example, of the five largest hubs examined (Atlanta, Chicago, Dallas, Denver, and Detroit) only Atlanta has a hubbing carrier that carries more than 50 percent of local traffic. By comparison, in all of the seven smallest hubs except one (Nashville), the hubbing carriers now carry more than 50 percent of local traffic. Altogether, the 5 largest hubs account for about 17 percent of the country's O&D traffic, compared with 8 percent in the 7 smallest hubs. For many travelers in medium-sized hubs, the additional service that hubbing airlines provide may be worth the added price. For example, typically in most medium-sized nonhub cities, only half of all passenger trips are on direct flights compared with nearly three-fourths in the medium- sized hubs (Table 3-14). This is a level of service that is comparable with that of large cities and is far more intensive than the local traffic alone would justify in these medium-sized hubs. Given that medium-sized hubs are increasingly being used by airlines in their regional hub-and-spoke networks [e.g., the number of departures grew by 131 percent at medium-sized hubs between 1979 and 1988 (Table 3-13)], single-carrier dominance and above-average fares will probably continue to characterize most airline hubs in the future. RURAL AND SMALL COMMUNITIES During the development of the aviation system, it was widely believed that subsidies were necessary for air service to be extended to small communities that might not be able to generate adequate traffic to support regular air service. Accordingly, in the 1950s the CAB certificated several local-service carriers (e.g., Piedritotit, Ozark, Frontier), whose primary function was to provide local service among small communities and feeder service to the larger airports served by the trunks. These carriers were compensated both directly from the CAB (on the basis of a cost formula)
TABLE 3-14 MARKET DATA FOR LARGE VERSUS MEDIUM CITIES BY HUB TYPES, 1988 Average City Size Average Oie-Way Enplanements Share of Local Departures and Market Sample Average Yield Trip Length per Passenger Passengers per Thousand Type Size (cents per mile) (miles) per Round Trip Flying Direct (%) Local Passengers Large Hub 8 15.9 890 2.3 80 30.8 Nonhub 17 12.0 950 2.4 77 16.4 Medium Hub 7 18.9 804 2.7 72 44.0 Nonhub 28 16.1 892 3.1 48 23.1 SOURCE: DOT O&D survey and RSPA Form 41
120 WINOs OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION and indirectly through cross subsidies from profits earned on moderate- density, monopoly routes. In many ways, the local-service airlines came to resemble the larger trunk airlines they were intended to supplement. Over the years, incentives in the subsidy system stimulated the locals to acquire modem jet aircraft similar to those used by the trunks (Meyer and Oster 1984, 17, 18). Moreover, local carriers were often allowed to compete with trunks on many medium-haul and moderate-density routes (thereby cross subsidizing service on low-density routes), as the CAB tried to reduce local carrier dependence on subsidies. In some regions of the country, such as New England, smaller commuter carriers, which were not regulated or subsidized (by the CAB), also served small communities before deregulation; however, seldom were those carriers able to compete with the larger local-service airlines. On the eve of deregulation, about 150 communities were receiving local, subsidized service. These communities ranged in size from several that averaged only a handful of passengers a day (e.g., Devils Lake, North Dakota, averaged fewer than 10 enplanements a day) to others that handled more than 200 passengers a day (e.g., Kalamazoo, Michigan). Although considerable evidence indicated that air service to these 150 small com- munities had been gradually declining for several years (Eads 1972; MacAvoy and Snow 1977), the prospect of deregulation created a great deal of uneasiness and debate. In particular, many subsidized communities were concerned that deregulation might mark the end of small-community ser- vice altogether, as the local-service jet operators withdrew to more prof- itable large and medium-sized markets. Indeed, once given the freedom to adjust fares and routes, most local- service airlines did shift much of their operations to higher-density markets more suitable for large jet aircraft. Quickly replacing the resulting vacan- cies, however, were many small commuter carriers, flying smaller, less costly aircraft (typically driven by turboprop instead of jet engines), which could provide direct feeder service to larger airports and the new major-carrier hubs. In many respects, this rapid emergence of commuter airlines was an industry surprise, growing from only 10 percent of rural small-community enplanements in 1977 to more than 50 percent in 1988 (Table 3-15). Early growth in commuter airlines may have been fostered in part by the Essential Air Service (EAS) program, which was established in the wake of deregulation. As discussed previously, many members of Con- gress from rural areas were concerned that deregulation would significantly reduce levels of air service in many small communities as local-service carriers shifted their operations to much larger city-pair routes. Hence to ensure that all small communities that received local service before de-
Passenger Fares and Airline Service 121 TABLE 3-15 SHARE OF TOTAL PASSENGER ENPLANEMENTS ON COMMUTER CARRIERS BY CITY SIZE, 1977 AND 1988 Percent Share of Enplanements City Size 1977 1988 Large 1.9 2.8 Medium 1.2 3.6 Small 2.0 10.0 Rural/small 8.0 51.1 NOTE: Large = cities with 2,000,000 or more enplanements in 1977; medium = cities with 500,000 to 2,000,000 enplanements in 1977; small = cities with 100,000 to 500,000 enplanements in 1977; rural/small = cities with fewer than 100,000 enplanements in 1977. SOURCE: DOT RSPA Forms 41 and 298-C. regulation would receive scheduled service afterward, Congress estab- lished the EAS. This program, which was recently reauthorized for an additional 10 years, allows the DOT to provide direct subsidies to com- muter carriers for providing scheduled minimum air service to small com- munities that might otherwise lose scheduled air service altogether. EAS- subsidized carriers must provide at least 2 round-trip flights, 5 days a week, from the small community to a larger city (typically a hub), although the use of jet aircraft is not required. In 1987, 110 communities received EAS-subsidized service in the contiguous United States, and EAS pay- ments totaled about $25 million (Stommes and Beningo 1989, 18). Changes in Service Two types of data are analyzed to determine trends in small-community service: passenger enplanements and daily departures (as a measure of flight frequency). Enplanement data are derived from the DOT's traffic reports for major carriers (T-3 and T-9 from RSPA Form 41) and from similar reports for commuters (RSPA Form 298-C). However, there is some question about the consistency over time of this data, especially because commuter car- riers do not comply with the same federal reporting requirements as the prederegulation local-service carriers. (Commuters file RSPA Form 298- C instead of Form 41, which was formerly filed by local carriers.) Ac-
122 WINOs OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION TABLE 3-16 SHARE OF TOTAL PASSENGER ENPLANEMENTS BY CITY SIZE, 1977 AND 1988 Percent Share of Enplanements City Size 1977 1988 Large 67 69 Medium 16 20 Small 10 7 Rural/small 9 4 Nom: Large = cities with 2,000,000 or more enplanements in 1977; medium = cities with 500,000 to 2,000,000 enplanements in 1977; small = cities with 100,000 to 500,000 enplanements in 1977; rural/small = cities with fewer than 100,000 enplanements in 1977. Columns may not sum to 100 because of rounding errors. SOURCE: DOT RSPA Forms 41 and 298-C. cordingly, as an alternative measure of service, departure data, derived from the OAG, also were analyzed for a sample of small communities. Passenger Enpianements To determine trends in enplanements, 200 cities were sampled and grouped as large, medium, small, or rural/small on the basis of 1977 enplanement data. Of these 200 cities, 25 were defined as large (2,000,000 or more enplanements); 35 were defined medium (500,000 to 2,000,000 enplane- ments); 65 were defined as small (100,000 to 500,000 enplanements); and 75 were defined as rural/small (fewer than 100,000 enplanements). In 1977, large cities accounted for about two-thirds of all enplanements nationally, whereas medium cities accounted for slightly more than 15 percent. By comparison, small cities accounted for 10 percent, whereas rural/small com- munities accounted for slightly less (9 percent) (Table 3-16). By 1988 the share of enplanements in the large and medium-sized cities increased to 69 and 20 percent, respectively, whereas the portion in small and rural/small communities dropped to 7 and 4 percent, respectively. This loss in traffic share in smaller communities can be attributed to a substantial increase in enplanements in large and medium-sized cities (partly as a result of increased hubbing), as well as to a slight decline in enplanements in many small and rural airports. As shown in Table 3-17, total enplanements in large
Passenger Fares and Airline Service 123 TABLE 3-17 CHANGE IN PASSENGER ENPLANEMENTS BY CARRIER TYPE AND CITY SIZE, 1977 TO 1988 Percent Change in Enplanements City Size Large Carriers Commuters Total Large +93 +181 . +94 Medium +131 +574 +132 Small +25 +569 +37 Rural/small â59 +490 â15 Nore: Large = cities with 2,000,000 or more enplanements in 1977; medium = cities with 500,000 to 2,000,000 enplanements in 1977; small = cities with 100,000 to 500,000 enplanements in 1977; rural/small = cities with fewer than 100,000 enplanements in 1977. SOURCE: DOT RSPA Forms 41 and 298-C. and medium cities doubled between 1977 and 1988, whereas enplanements in rural/small communities declined by 15 percent. These results correspond with findings from a more detailed analysis of enpianement trends (from 1977 to 1988) in a sample of 170 small communities, which also suggested that most small communities have lost passenger traffic since deregulation. Altogether, the median change in enplanements in 100 formerly subsidized communities was â26 percent (with 62 percent of sampled communities losing enplanements), whereas the median change in 70 unsubsidized communities was - 32 percent (with 65 percent of communities losing enplanements) (Figure 3-15). This decline in passenger traffic, however, may not be indicative of an erosion in service. An even closer look at the data suggests that many small- community travelersâespecially price-sensitive pleasure travelersâmay be taking advantage of service offerings at nearby larger airports. Of the 100 formerly subsidized communities sampled, 85 are within 100 miles of larger airports, which are likely to offer lower fares, more nonstop service, and a wider selection of competing carriers than the small- community airports. Indeed, if this analysis is expanded to include all enplanements within a 100-mile radius of the formerly subsidized com- munities, then only 30 percent of small communities lost enplanements overall (Figure 3-16). Perhaps because of this consolidation of traffic at larger airports (char- actenzed by one airport serving several small communities), few rural states have actually lost passenger traffic since deregulation. As shown in Table 3-18, only two states (South Dakota and West Virginia) had
124 WINDS OP CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION Median change in enplanements in sample Formerly Subsidized Unsubsidized (62/100 communities (46170 communities lost enplanements) lost enplanements) FIGURE 3-15 Median change in passenger enplanements for small commu- nities sampled, 1977 to 1988 (formerly subsidized communities are those that received local-class subsidized service in 1977). fewer enplanements in 1988 than in 1977; both of these states lost pop- ulation during the 1980s. Flight Frequency (Departures) Before deregulation, subsidy payments to local-service carriers were based partly on the number of departures from subsidized communities. However, because these payments did not vary by time of dayâand because local-service carriers tended to use their jet aircraft to serve more profitable, larger markets during peak periodsâflights to sub- sidized small communities were often infrequent and scheduled at in- convenient times of the day (e.g., early morning or late evening) (Meyer and Oster 1984, 17). Since deregulation, flight frequencies in most small communities have increased sharply. For example, daily departures derived from the OAG (which includes both commuter- and large-carrier flight listings) were compared for 1977 and 1988 using the sample of 100 formerly 0% -5% -10% -15% -20% -25% -30% -35%
Passenger Fares and Airline Service 125 Median change in enpianements in sample 25% 20% 15% 10% 5% 0% -5% -10% Within 50-mi radius Within 100-mi radius (55/100 communities (70/100 communities lost enpianements) gained enpianements) FIGURE 3-16 Median change in regional passenger enpianements for formerly- subsidized communities sampled, 1977 to 1988 (formerly subsidized commu- nities are those that received local-class subsidized service in 1977). subsidized small cities. As shown in Figure 3-17, the majority of these cities have experienced an increased number of scheduled departures since deregulation, with a median change of + 29 percent. Of the 100 communities sampled, 61 experienced an increase in departures during the il -year period. The increase is due largely to the proliferation of commuter carriers. Because they fly smaller aircraft, commuter carriers can provide more frequent and timely flights in smaller markets (often to and from hubs) than larger jet operators, even though they carry fewer passengers overall. This level of service is especially important to business travelers, who place a high value on convenient and flexible scheduling. The findings of improved service in small communities in the sample used in this study are corroborated by other analyses using different samples (Butler and Huston 1990). Changes in Fares The elimination of direct CAB subsidies and indirect cross subsidies from travelers in larger markets was expected to have an important
TABLE 3-18 CHANGE IN TOTAL ENPLANEMENTS IN STATES THAT LOST LARGE-CARRIER SERVICE, 1977 TO 1988 State 1977 Enplanements Large Carriers Commuters Total 1988 Enplanements Large Carriers Commuters Total Percent Change in Total Enplanements Iowa 1,132,080 3,551 1,135,631 1,124,340 244,067 1,368,407 +20 Kansas 648,289 0 648,289 602,128 99,493 701,621 +8 Mississippi 601,954 13,321 615,275 491,206 120,750 611,956 0 Montana 787,283 0 787,283 680,949 131,281 812,230 +3 North Dakota 474,999 0 474,999 466,746 33,474 500,220 +5 South Dakota 481,932 0 481,932 362,520 87,168 449,688 -7 West Virginia 509,754 66,300 576,054 310,517 202,590 513,107 - 12 Wyoming 272,981 13,586 286,567 205,419 111,909 317,328 +11 SOURCE: DOT RSPA Forms 41 and 298-C
Passenger Fares and Airline Service 127 Median change in departures in sample 100% 80% 60% 40% 20% 0% In small community Within 50-mi radius Within 100-mi radius (61 /1 00 gained (63/1 00 gained (91/1 00 gained departures) departures) departures) FIGURE 3-17 Median change in departures in formerly-subsidized commu- nities and their region. 1977 to 1988 (formerly subsidized communities are those that received local-class subsidized service in 1977). impact on fares in small communities. Many economists predicted that fares for short-haul, low-density routes would initially go up in response to the elimination of these subsidies, as airlines adjusted to the lower traffic volumes in these markets. Few economists, however, could pre- dict the extent to which the traffic efficiencies created by hub-and- spoke systems might offset this effect. Unfortunately, the available data on fares paid by small-community travelers are limited. The most reliable source of fare data, the O&D survey, does not include data from commuter carriers. Accordingly, only small-community travelers who fly all or part of their trip on a large carrier are included in the survey, whereas travelers flying only on commuters are not. These data shortcomings were given serious consideration in the examination of fares in small-community city pairs for this study. Accordingly, no attempt was made to determine fare trends in low-density, short-h2iul, small-community markets, which are likely to have the highest share of passengers flying on commuters. Instead, a sample of 160 small-community markets was drawn from the 795-city-pair sample. For the most part, the city pairs involved one small community paired with a much larger city (e.g., Bismarck, North
128 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION Dakota, and New York City), in which virtually all trips were likely to have involved a large carrier for at least part of the journey. For this analysis, the small community was defined even more broadly than before to include many small cities with up to 500,000 enplanements in 1988. The purpose of this analysis was to determine if average yields in small-community markets varied significantly from those in larger markets during the sample period of 1988 to 1989. Each of the 160 small-community city pairs was grouped by mileage block, along with the remaining 635 city pairs. The average yields in each of the small-community categories and the other categories were then compared. As shown in Table 3-19, the results from this analysis show yields in small communities that are not noticeably different from those in TABLE 3-19 AVERAGE YIELD IN SMALL-COMMUNITY CITY PAIRS VERSUS ALL OTHERS, 1988-1989 Yield (cents per mile) by Distance (miles) 100â 500â 1,000â 1,500â Market Type 499 999 1,499 1,999 2,000+ Small community 38.4 21.3 13.4 10.4 10.0 (N=62) (N=40) (N=35) (N=12) (N=ll) Other 35.9 21.3 12.8 11.1 9.5 (N=148) (N=165) (N=95) (N=123) (N=104) Percent difference + 7 0 + 5 â6 + 5 NOTE: Small-community city pairs include one or more small cities, which are defined as having fewer than 500,000 enplanements per year. However, because of insufficient fare information for very small cities, only 20 city pairs involving cities with fewer than 100,000 enplanements are included in the sample. Data are for the last two quarters of 1988 and the first two quarters of 1990. N = sample size. SOURCE: DOT O&D Survey. other markets. For most markets sampled, the difference in yield was typically less than 10 percent, suggesting that on many long- and medium- haul trips (which usually involve a connection at a larger hub or spoke airport), travelers from small communities may be benefiting from the same efficiencies of hub-and-spoke systems as travelers from much larger markets.
Passenger Fares and Airline Service 129 SUMMARY General Trends Deregulation brought changes to the airline industry that have produced substantial benefits to air travelers. More travelers are flying now than ever before: the number of annual passenger trips (which excludes connecting enplanements) has increased by nearly 100 percent since 1977. This growth in travel has been accompanied byâand largely stimulated byâreductions in the fares paid by passengers. After ad- justments are made for inflation, the average passenger yield (fare divided by miles traveled) fell by about 15 percent between 1979 and 1989. This decline occurred despite several external events that caused sharp increases in airline costs, which probably would have resulted in even higher fares had the previous system of fare and route regulation continued. Overall, the pattern of fare and service changes since deregulation has been fairly consistent with prederegulation predictions. Fare and service options are much greater than before deregulation, resulting in a wider variety of fares. Hence, today approximately 25 percent of passengers are paying fares that are much higher (2.5 times higher) than the industry average yield compared with 19 percent 10 years earlier. By comparison, however, 15 percent of passengers are paying less than half the industry average fare, which is three times more than in 1979. Altogether, more than three-fourths of travelers are flying in markets that have experienced real fare declines since deregulation. The principal beneficiaries have been travelers in medium- and longer-distance markets, who have benefited from the elimination of fare cross subsidies (which favored short-haul travel) and from the proliferation of com- peting hub-and-spoke systems. On average, travelers in longer-distance markets (more than 1,000 miles) have enjoyed real fare declines of 10 to 35 percent. Effects of Competition on Fares and Service In recent years fares have climbed upward slightly throughout much of the airline industry. This trend toward somewhat higher fares has co- incided with an increase in airline costs as well as an increase in airline concentration nationally and in many individual cities and city-pair
130 WINDS or CFIANGE: DoMEsTIc AIR TRANSPORT SINCE DEREGULATION markets. In 1979 only 20 percent of markets had three or more com- peting carriers. However, by 1984 more than half did. Today, this figure stands at 40 percent, although the vast majority of passengersâab3ut two-thirdsâare still flying in large, multicarrier markets. Recently, concerns about competition have centered specifically around hub-and-spoke systems. On one hand, these systems have substantially increased competition, especially in medium- and long-haul markets in which travelers can choose among several competing airline hubs. On the other hand, they have provided opportunities for airlines to dominate local passenger traffic at their hubs. In 1989 approximately 40 percent of city pairs that included concentrated hubs (as an origin or destination) was dominated by a single carrier compared with approximately 15 percent of nonhub city pairs. In general, yields are 10 to 35 percent higher in single-carrier markets than in multicarrier markets. As might be expected, yields tend to be higher in concentrated hubs than in most nonhub markets, by about 20 percent on average. At least part of this differential is associated with differences in the level of service enjoyed by travelers in hub cities. Hub travelers, for example, can choose from many more flights to a greater variety of cities, enjoy less-circuitous routes, and have access to more nonstop service than travelers in most nonhub cities. Average yields in most hubs (concen- trated and unconcentrated) tend to be at least 10 to 15 percent higher than in nonhub cities. Although yields in concentrated hubs are some- what higher still (by 5 to 10 percentage points), it is not clear whether this difference stems from market power or variations in service costs. Because medium-sized cities, which are most readily dominated by hubbing carriers, are increasingly being utilized by airlines in their regional hub-and-spoke networks, single-carrier dominance and above- average fares will probably continue to characterize most airline hubs. Rural and Small Communities One of the explicit goals of deregulation was to allow carriers to move to a market-oriented pricing and route system. The elimination of cross subsidies and required service to small communities was expected to result in a more efficient airline industry, which would benefit the majority of air travelers, most of whom fly in larger, more heavily traveled markets. Since deregulation, most smaller communities have lost some or all service from large jet carriers. Between 1977 and 1988, the number of
Passenger Fares and Airline Service 131 major-carrier enplanements fell by about 60 percent in cities with fewer than 100,000 enplanements per year. At the same time, however, smaller commuter carriers picked up most of the slack, growing from virtual nonexistence to become the principal provider of most small-community service. The shift toward commuter carriage is evident in small-community enplanement and departure trends. Compared with the local-service jet carriers that once served many small communities, commuters use much smaller aircraft, which carry fewer passengers, but are flown more efficiently on a frequent and timely basis. Accordingly, whereas the number of enplanements in most small communities has declined by about one-sixth since deregulation, the number of scheduled departures has increased by about one-third. This decline in passenger traffic, coinciding with an increase in scheduled flights, suggests that commuter carriers may be concentrating on serving more time-sensitive business travelers, whereas many other small-community travelers are apparently driving to nearby larger airports to take advantage of lower fares and a wider array of services. Because of the economies associated with higher traffic volumes, fares are usually lowest in large cities. For most medium- and long- haul travel, however, small-community travelers are likely benefiting from the same traffic efficiencies created by hub-and-spoke systems as travelers from much larger markets. Indeed, yields in these small- community, medium- to long-haul markets, are similar to those in much larger markets, which have fallen sharply (in real terms) since deregulation. NOTES The SIFL index is not used as a deflator in this chapter in order to remain consistent with previous studies, which use GNP or other consumer-oriented price deflators. Also, the SIFL index itself has not been immune to the effects of deregulation, which have significantly affected airline industry cost structures from which the SIFL is derived (e.g., labor costs). The yield taper in Figure 3-7 reflects the high cost per mile for short trips, and the gradually declining cost per mile as trip length increases. The higher cost per mile for short trips is due to certain fixed costs of providing air service that do not vary by trip length, such as take-off and landing costs (which account for a large portion of fuel consumed) and terminal charges (e.g., gate space and baggage handling). Examinations of pricing at two-carrier hubs (DOT 1990b, Vol. 1, 156), as well as findings on fare levels by number of competing carriers that are presented
132 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION earlier in this chapter, suggest that markets with two competing carriers will have many characteristics similar to those of monopoly, or single-carrier, mar- kets; therefore, hubs dominated by either one or two carriers were defined as concentrated for the purposes of these analyses. It is possible that a higher proportion of lower market densities in concentrated- hub city pairs could have contributed to the larger percentage of one-carrier markets; however, in the city-pair sample, only 42 percent (95 of 226) of the concentrated-hub markets were low-density (less than 100 passengers per day), compared with 46 percent (262 of 569) of nonhub and unconcentrated-hub markets. By comparison, high-density markets (more than 400 passengers per day) accounted for 26 percent (62 of 226) of the concentrated-hub sample and 26 percent (143 of 569) of the nonhub and unconcentrated-hub sample. Never- theless, 35 percent (22 of 62) of the high-density markets in the concentrated- hub sample were served by only one-carrier, compared with only 4 percent (6 of 143) of the nonhub and unconcentrated-hub markets. In fact, the concentrated- hub markets in our sample had a much higher share (usually twice as high) of one-carrier markets for all six density categories examined. The value of increased schedule frequency to business travelers accounts for a large share (about two-thirds) of the total monetary benefits of deregulation estimated by Morrison and Winston (1986). Estimating the value of reduced time between flights, however, is a difficult task. On one hand, increased schedule frequency allows business travelers to better tailor their travel plans to the demands of the business day. This schedule flexibility results in the more productive use of time by business travelers, because less time is spent in transit before and after business meetings. On the other hand, not all of this transit time is likely to be unproductive, because business travelers can often plan ahead and make productive use of this time even if desired flight schedules are not available. Hence, whereas the committee recognizes that increased schedule frequency is valuable to business travelers, it is skeptical about the ability to quantify this benefit with precision. The accuracy of these yield-differential estimates might be affected by the small number of hub cities in the analysis. Even though all hubs were included in the comparison (by the hub definitions employed)âand yield estimates were based on a large and reliable sample of ticketsâthe total number of hubs in the comparison (15) is small, making it difficult to control the possible skewing effects of 1 or 2 hubs with special circumstances, such as slot controls or unusually high market densities (e.g., Chicago). REFERENCES ABBREVIATIONS AlA Air Transport Association of America CAB Civil Aeronautics Board DOT U.S. Department of Transportation FAA Federal Aviation Administration GAO General Accounting Office
Passenger Fares and Airline Service 133 ATA. 1988. Airline Deregulation 10 Years Later: What Has It Done for Consumers? Washington, D.C., Oct. ATA. 1989. Competition in the Airline Industry. Air Transport Report. Washington, D.C., Sept. Bailey, E., D. Graham, and D. Kaplan. 1985. Deregulating the Airlines. MIT Press, Cambridge, Mass. Bailey, E. E., and J. R. Williams. 1988. Sources of Economic Rent in the Deregulated Airline Industry. Journal of Law and Economics, Vol. 31, pp. 173-203. Borenstein, S. 1989. Hubs and High Fares: Airport Dominance and Market Power in the U.S. Airline Industry. Rand Journal of Economics, Vol. 20, No. 3, Autumn, pp. 344-365. Butler, R., and J. Huston. 1990. Airline Service to Non-Hub Airports Ten Years After Deregulation. The Logistics and Transportation Review, Vol. 26, No. 1, pp. 3-16. CAB. 1984. Implementation of the Provisions of the Airline Deregulation Act of 1978. Report to Congress. Washington, D.C., Jan. DOT. 1 990a. Secretary's Task Force on Competition in the U.S. Domestic Airline Industry. Executive Summary. Feb. DOT. 1990b. Secretary's Task Force on Competition in the U.S. Domestic Airline Industry. Pricing. Vols. I and II. Douglas, G., and J. Miller III. 1974. Economic Regulation of Domestic Air Transport: Theory and Policy. The Brookings Institution, Washington, D.C. Dresner, M., and R. Windle. 1991. Airport Dominance and Pricing in the U.S. Airline Industry. Working Paper. College of Business and Management, University of Maryland, College Park. Eads, G. 1972. The Local Service Airline Experiment. The Brookings Institution, Washington, D.C. FAA. 1974-1990. FAA Aviation Forecasts. U.S. Department of Transportation. GAO. 1985. Deregulation: Increased Competition Is Making Airlines More Efficient and Responsive to Consumers. GAO/RCED-86-2. Washington, D.C. GAO. 1988. Airline Competition: Fare and Service Changes at St. Louis Since the TWA-Ozark Merger. RCED-88-217BR. Washington, D.C., Sept. GAO. 1990. Airline Competition: Industry Operating and Marketing Practices Limit Market Entry. RCED 90-147. Washington, D.C., Aug. Levine, M. E. 1987. Airline Competition in Deregulated Markets: Theory, Firm Strategy and Public Policy. Yale Journal on Regulation, Vol. 4, Spring, pp. 393-494. MacAvoy, P., and J. Snow (eds.). 1977. Regulation of Passenger Fares and Com- petition Among the Airlines. Ford Administration Papers on Regulatory Reform. American Enterprise Institute for Public Policy Research, Washington, D.C. Meyer, J., and C. Oster, Jr. 1981. Deregulating the Airlines: The Early Experience. Auburn House Publishing Co., Dover, Mass. Meyer, J., and C. Oster, Jr. 1984. Deregulation and the New Airline Entrepreneurs. MIT Press, Cambridge, Mass. Meyer, J., and C. Oster, Jr. 1987. Deregulation and the Future of Intercity Passenger Travel. MIT Press, Cambridge, Mass. Morrison, S., and C. Winston. 1986. The Economics of Airline Deregulation. The Brookings Institution, Washington, D.C. Schwieterman, J. P. 1985. Fare Is Fair in Airline Deregulation: The Decline of Price Discrimination. AEI Journal of Government and Society, July-Aug., pp. 32-38. Stommes, E., and S. Beningo. 1989. Rural Air Service: Its Importance and Status Since Deregulation. Office of Transportation, U.S. Department of Agriculture.
rd Barriers to Competition One expectation of deregulation was that allowing low-cost competitors to enter the industry would bring down consumer costs and make the air transportation system more competitive and efficient (MacAvoy and Snow 1977). Substantial new entry did occur during the early phase of deregulation, but since the mid- 1980s new entry has slowed and the industry has become more concentrated. As in the prederegulation period, a few firms account for more than 95 percent of revenue passenger miles. In contrast with the prederegulation period, however, the existing air carriers are competing with each other much more vigorously. Preserving this competition is fundamental to retaining the benefits of deregulation. It is impossible to estimate at this stage how much additional concen- tration could occur without a significant reduction in competition, but clearly competition at some level is threatened partly because entry into the industry appears to have slowed to a standstill. An examination of whether barriers to competition are inhibiting entry by new carriers or competition among existing carriers is presented in this chapter. Barriers to entry are defined here in an applied sense; that is, a barrier to entry includes (a) the total inability of a firm to enter a market, (b) the existence of costs borne by the new entrant but not borne by the incumbent, or (c) advantages that accrue to incumbent firms because of economies of scale and scope (Bain 1968). The following potential barriers to entry in the commercial aviation industry are explored: (a) the financial risk of entry, (b) airport capacity constraints that diminish the prospects of entry, (c) environmental issues that affect airport use and the availability and cost of aircraft, and (d) the effect of airline marketing strategies on competition. The key points made in this chapter are that impediments to the performance of the deregulated marketplace could reduce the benefits enjoyed by consumers to date, but policy makers attempting to eliminate these impediments are faced with 134
Barriers to Competition135 a paradox. Many impediments result from management innovations or marketplace conditions that also serve consumer interests. Hence, in sev- eral cases, it is difficult for government to intervene without imposing costs in excess of benefits. Some policy changes that will help are iden- tified in the summary section. FINANCIAL RISKS Some proposals for deregulation were accompanied by the belief that carriers would have relatively free access to markets because of the mobility of the airlines' chief assetsâaircraft. Carriers dominating individual markets would not charge monopoly fares, according to this theory, because of the ease with which a competitor could enter at reduced fares and contest the incum- bent carrier (hence the name contestability theory). Thus the mere threat of entry was expected to discipline pricing. Contestability theory requires low costs of entry and exit. Access to many markets, however, is far from costless, partly because of the expense and difficulty in obtaining terminal space at many hubs (discussed later), but mostly because of the risk associated with competing with an airline at one of its hubs. A competitor that wishes to challenge another carrier at its hub is faced with considerable financial outlays. The cost of providing a competitive level of service at a hub is substantial: expenditures for advertising, per- sonnel, and aircraft operations are crucial during start-up, when the com- petitor attempts to win business away from the hub carrier. The risk of being unable to recover these outlays is the largest single deterrent to entry at hub airports. It is difficult to compete with a hub carrier during start-up because the hubbing carrier has inherent advantages: some result from the scope of its operations, others from marketing. The larger network of the dominant carrier allows it to increase service at a lower marginal cost (Caves et al. 1984). In addition, by having an extensive network, the dominant carrier is more likely to attract passengers, who then form impressions about the quality of service on other routes (Levine 1987). Marketing builds on these advantages. Frequent flier programs make it difficult to attract busi- ness travelers away from an incumbent carrier with which they may have already accrued a substantial balance in a frequent flier account. Finally, if the incumbent carrier has already established preferred- prov i der rela- tionships with most of the travel agents around the hub, the new entrant faces an additional competitive disadvantage. During the months in which a competitor first takes on a carrier at its hub, the competitor must offer
136 WINOs OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION substantial levels of service, which at a minimum include dozens of flights a day. It must confront the problems of luring frequent fliers away from an incumbent that offers them more opportunities to earn mileage and the fact that travel agents have preferred-carrier relationships with the incum- bent. Because most cities will not support competitive hub systems, com- petition at a hub tends to be a battle from which only one carrier emerges (Levine 1987). The attempt of Midway to compete with USAir in Phil- adelphia, for example, resulted in heavy losses for both carriers, and Midway, unable to gain sufficient traffic, ultimately withdrew entirely. A less risky alternative for a carrier wishing to establish a hub in a major market is to develop another airport to serve the same city; this strategy has been pursued in several cities, for example, by Midway at Chicago's in-town airport (Midway). These airports have considerably lower traffic volumes than the major airports that serve those communities, but they have allowed new entrants to develop niche markets. Another alternative is to operate a spoke into the major's hub (which requires one or two gates) to attract passengers originating from or destined for the dominated hub to make connections through the new entrant's hub. Providing sufficient capacity to allow additional spoke operations may not significantly reduce the competitive advantages of the hubbing carrier but will provide a marginal improvement in competition by adding more choices for passengers, especially leisure travelers. AIRPORT CAPACITY CONSTRAINTS As airport terminal capacity limits are reached, it can become more dif- ficult for new carriers to begin service and for existing carriers to enter new markets. Entering a market requires the ability to lease or develop gates, baggage handling and aircraft maintenance facilities, and ticketing and passenger waiting areas. According to airport managers, little under- used gate capacity and related terminal space is available at major airports in the short term. Over the longer term it is possible for carriers to enter many markets, but the experience of the last few years indicates that such entry is neither easy nor inexpensive. Terminal and Gate Capacity The lack of excess terminal space at many airports is partly a result of the way airports have been financed. Many airport terminals were devel-
Barriers to Competition 137 oped well before deregulation and were financed with 15- to 30-year bonds, which were guaranteed by the airport through long-term leases with the airline or airlines serving the airport. Whereas some of the leases have expired and have been renegotiated, others have several more years to run. In exchange for entering into these long-term commitments, which exposed the airlines to financial risk and locked them into serving the airport, the airlines typically required lease clauses that would protect them from additional costs. These agreements may well have been rea- sonable transactions before deregulation as a method to balance the power of two potential monopolists (the airline and the airport). One of the goals of deregulation, however, is to promote competition by allowing relatively free entry. The incumbent airlines have a considerable advantage if they are able to block entry into a market they serve or if they can substantially raise the cost of entry such that potential competitors would be deterred from entering the market. The results of a 1989 survey of the members of the Airport Operators Council International (AOCI), whose member airports represent 90 percent of domestic commercial aviation traffic, imply that the barriers to entry caused by air terminal capacity limits and long-term lease agreements at the 30 largest airports in the United States are substantial (D. Plavin, testimony before Subcommittee on Aviation, House Committee on Public Works and Transportation, Sept. 19, 1989). For example, member airports were asked about the availability of "competitive gates" during peak hours and off peak hours, and few gates were available in either the peak or off peak hours.' At the time the survey was conducted, airport terminal space may well have been in short supply. In the midst of the current recession, and following the demise of Eastern Airlines, terminal space is likely to be available in many markets. Atlanta, for example, currently has an excess of gates, whereas just 2 years ago there was no excess. Given the current lack of demand and the pressed finances of most carriers, few airlines are in positions to capitalize on the excess supply. Once the current recession is over, however, shortages may well reappear. Many airport terminals have been expanded over the years, and when these expansions occur, some excess capacity usually is built in to ac- commodate growth. In effect, capacity is provided in increments that exceed immediate demand, but airlines that wish to enter after the incum- bents grow into that capacity may not find space. From the vantage point of prudent fiscal management, neither airlines nor airport managers want too much unused physical capacity because the costs of unused space must be borne by all the tenants and users of the airport. From the vantage
138 WiNos OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION point of a potential new entrant, the short-term prospects of assembling a sufficient number of gates at two or three of these major airports (at times of the day that provide for reasonable connections with other flights) can be daunting. Contractual Agreements Airport operators believe that existing capacity limits are exacerbated at many airports because the incumbent airlines, holding long-term leases with majority-in-interest (Mu) clauses or exclusive-use agreements, are able to block airport expansions that would provide more capacity for new entrants. In addition, many airport-airline leases contain clauses that pro- hibit the airport from charging "additional rates, fees and charges" and from changing its method of calculating landing fees. The airlines can block expansions with these provisions, but only those that would increase their costs without their consent. MI! Clauses MII clauses, which vary from lease to lease, essentially give the airlines that lease airport facilities the authority to approve certain airport capital expenditures. Approval of expenditures is by the consensus of the majority (or a set percentage of votes). Each airline's vote is weighed according to the amount of space it leases or its share of airport operations. Mu clauses are typically included in financing mechanisms referred to as residual cost leases. In a residual lease, the airline is responsible for any cost that is not covered by all other sources of revenue available to the airport, such as concessions and parking. The MII clause protects the airline from having to pay for airport facilities that may not be cost- effective or in the interest of the airline, such as elaborate decorations in the airport. Of the top 30 airports in the country in 1984, roughly half had residual leases; most of those leases contain MII clauses that cover all or part of the airport (CBO 1984). Exclusive-Use Agreements Some airport leases give the incumbent airline exclusive use of the facilities it is helping to finance, even if it is not using them. When a new entrant
Barriers to Competition 139 wishes to serve such an airport, it must sublease facilities from the in- cumbent airline. If the new entrant is a potential competitor, the incumbent is in a position to extract monopoly rents. Fee Restrictions Airport leases also can prohibit the airport from charging additional fees or changing the methods by which existing fees are charged. Although it is a reasonable protection for airlines against unanticipated costs being imposed by the airport, this provision could also be used to block an expansion at the airport. The leases of half of the 30 largest airports in the AOCI survey contain such clauses. The General Accounting Office (GAO) also surveyed a sample of air- ports (n = 183) to investigate opportunities for entry (GAO 1990). Results of the GAO survey indicate that at large and medium-sized airports (n = 66), 87 percent of gates are on long-term leases, and about 60 percent of all gates are on long-term, exclusive-use leases that will not expire for another 10 years or more. At airports defined as concentrated, leases on 53 percent of the gates will not expire for 20 years. Because airlines do not always need the gates they lease, they occa- sionally sublease them to other carriers. Most subleases, however, are with regional or commuter airlines and not potential competitors (GAO 1990). The GAO also found that subleases for existing gates cost a premiumâat least 10 percent more (and usually much more) than the cost for the incumbents (K. Mead, testimony before Subcommittee on Aviation, House Committee on Public Works and Transportation, Sept. 21, 1989). In its review of the AOCI and the GAO data on airport capacity, the Secretary of Transportation's Task Force on Competition in the U.S. Domestic Airline Industry reported, "It is not possible to prove that market power is being exercised by sellers of gate capacity," but it did conclude the following (DOT 1990b): At certain important airports there is a shortage of groundside facilities including boarding gates, holding rooms, baggage handling areas, and aircraft service facilities. Exclusive use agreements and long term lease of gates and other facilities between incumbent airlines and the airport also make new entry difficult. Construction of new facilities takes on average two years. Further, in some cases provisions in existing contracts between airports and airlines can delay or prevent construction of new facilities.
140 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION The Air Transport Association of America (ATA) sharply disputes the claim that current airport capacity limits or lease clauses are barriers to competition. Underused facilities, for example, can be subleased, and even though lessors can earn a profit on the sublease, lessees may still be able to benefit from gaining access to the airport. As a practical matter, however, subleasing occurs infrequently. The 1989 AOCI survey found that only 8 percent of gates had been sold or subleased during the past 5 years, and most of these subleases were apparently with regional or commuter carriers and not with competitors (D. Plavin, testimony before Subcommittee on Aviation, House Com- mittee on Public Works and Transportation, Sept. 19, 1989; see also DOT 1990b, Table 3.2). There is also anecdotal evidence that by charging high sublease costs, in combination with costs for ground handling, incumbent carriers can effectively raise the price of entry for smaller competitors, and thereby reduce the cost advantages that the smaller carriers might have (Levine 1987; DOT 1990b). Lease contract provisions, according to the ATA, allow incumbents to refuse to support only those additions to capacity that would serve com- petitors and increase costs for incumbents. In other words, a competitor and an airport operator may expand facilities only if they do not increase costs for incumbents. The disadvantage of this financing arrangement is that during shortages of gates and related facilities, new entrants must be prepared to finance new capacity. Not only does this typically require 2 years, but the nominal cost to the new entrant may be higher than the cost for the incumbents; because incumbents lease facilities built in the past, they may be paying low unit costs compared with what a new entrant would have to pay. Although this latter phenomenon makes entry into airline markets more difficult, it is a common problem in durable goods industries. In the past when a new carrier wished to begin serving an airport, the existing airlines in most cases agreed to "cost equalization" to pay for runways, taxiways, and aprons. In some cases the airlines agree that a terminal expansion to accommodate a new carrier should be financed by all carriers, but this is typically limited to expansions that benefit all the carriers, such as the addition of baggage handling areas and ticket counters. Other parts of the terminalâgates, hold areas, and lounges for first-class and frequent fliersâmust be financed by the new entrant. In regard to the justification of current long-term leases, the airline position is that the carriers and the airports made a deal for which airlines bear the risk. Robert Crandall, Chairman of AMR Corporation has stated, "I don't care, I'll take a 30-day lease. But, if the community says to me,
Barriers to Competition 141 'We want you to make a $500 million or $600 million capital investment,' I'm not going to do that in exchange for a 30-day lease" (Airport Mag- azine, 1989). The contractual issues involving terminal space and use are complex and vary from airport to airport. Airports have different financing ar- rangements for various parts of their facilities, and airline leases contain varying provisions. Regarding these various lease provisions, the Secre- tary's Task Force on Competition concluded, "At best, the numerous contractual barriers make it difficult for a new entrant to obtain cost- competitive access to airports. At worst, contractual clauses such as Mu deter efficient development of new gate capacity, with a negative effect on new entry" (DOT 1990b, 3.14). Regardless of the specific role of Mu or other lease provisions, when new entrants must finance terminal ex- pansions, their competitive positions are weakened when they bear a higher cost than the incumbents and when their entry into a market is delayed during the design and construction period. In 1990 Congress approved legislation recommended by Secretary of Transportation Samuel Skinner that may permit airports to raise additional revenues by imposing a head tax on passengers [referred to as passenger facility charges (PFCs)]. Such taxes had been prohibited in the past because the city of Philadelphia (which owns the airport that serves the city) once used such revenues for nonairport municipal needs. Under the new law, airports would only be allowed to use such funds for projects directly related to airport development, such as the building of gates to enhance competition. This new revenue source would give airport managers an opportunity to increase service and competition at their airports, but Con- gress stipulated that the head taxes would not be allowed unless the Federal Aviation Administration (FAA) develops an acceptable solution to the problems of limited slots at slot-controlled airports. The FAA initiated rulemaking in 1991 to reconsider the slot allocation system. (Slot allo- cations are discussed later.) Dominated Hubs As carriers build the connection banks required to make a hub work, their presence in the local market can become so pervasive as to approach being a monopoly. Some airline executives claim that they do not set out to build monopolies. "You are not intentionally expanding it to make it into a for- tress," says Crandall. "You expand because it becomes more economical" (Airport Magazine, 1989). However, financial analysts and others assessing
142 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION airline profit strategies have concluded that airlines use hubs to shield some of their output from competition (Oum and Tretheway 1990). As more flights are connected to a hub, the number of passengers available to support ad- ditional flights grows. Making the connecting banks work for these flights requires many gates because of the desire to minimize the delay between connections. And higher-yield local passengers help provide density to sup- port frequent hub service. Because few airports have excess capacity in the short run (and few have enough local traffic to support more than one extensive network of nonstop service) hubs tend to become dominated by one or two major carriers who use up the existing capacity. The competitive advantages that dominant carriers enjoy, coupled with the airlines' ability to influence or control airport development, led the Secretary of Transportation's task force to conclude, "A major problem exists when incumbent air carriers have the ability to control airport de- velopment, because it is in their interest to restrict capacity in order to discourage new entrants" (DOT 1990b, v). At the same time, however, dominated hubs are likely to remain dominated for the competitive reasons cited previously and not because of a shortage of capacity. The ability to enter is a necessary condition for a competitor to enter a dominated hub but is not sufficient to ensure that competition will occur. The difficulty and expense of obtaining terminal space, however real, are much smaller deterrents to entry than the financial risks discussed previously. Slot-Controlled Airports Since 1969 airlines wishing to serve New York's LaGuardia or John F. Kennedy, Washington's National, or Chicago's O'Hare airports have had to hold a "slot," or time period, in which to land or take off. The FAA imposed the high-density rule in 1969 as a temporary measure to minimize delay and made the rule permanent in 1973. The slots were allocated to the carriers with authority from the Civil Aeronautics Board (CAB) to operate at those airports and were initially managed by scheduling com- mittees of airline executives. In 1986 the FAA began to allow carriers to sell or lease their slots to other carriers. About 128 of the total 3,800 slots changed hands each quarter initially, but the number declined consistently over time to about 20 per quarter in 1988 (GAO 1990). Leasing of slots now occurs more frequently than sales, and slots are leased usually to regional carriers for short periods, typically 90 days, thereby limiting the opportunities of competing jet carriers to serve the airport (GAO 1990).
Barriers to Competition 143 At Chicago O'Hare, United and American airlines increased their con- trol over the proportion of slots from 65 to 75 percent between 1986 and 1988. Slots at the other slot-controlled airports are much more evenly distributed among the major carriers. (The ATA is opposed to slot controls and has long advocated their abolition.) From time to time the FAA has reallocated a small number of slots to new entrants. The acquisition of slots, already difficult because of scarcity, is com- plicated by the need of new entrants to also obtain gates at congested airports. Despite the difficulty, new entrants have been able to acquire slots. America West acquired four at Kennedy and seven at O'Hare, and Midwest Express acquired four at National and LaGuardia and two at O'Hare. About half of these acquisitions were achieved through FAA reallocations (DOT 1990b, 2.16). Although incumbent carriers are apparently using slot controls to their advantage, the Transportation Secretary's task force concluded that suf- ficient evidence does not exist to prove that the current system of slot allocation is being used as a barrier to entry (DOT 1990b, 2.17). Problems do appear to exist in the function of the market, however, because some new entrants are not given an opportunity to bid on slots (DOT 1990b, 2.17). The present system of slot sales does not require any particular form of sale, and potential purchasers are not always informed of im- pending sales. As mentioned previously, in 1990 legislation Congress required the FAA to examine the existing slot allocation system with the aim of improving opportunities for new entrants. ENVIRONMENTAL CONSTRAINTS Although community opposition to aircraft noise contributes to airport capacity limits, which may impede competition at some airports, at present these constraints apply to all airlines, not just new entrants. Public actions to reduce the number of noisier aircraft now in use, however, can constrict supply and reduce the prospects of new entry. In response to growing public complaints about aircraft noise during the 1960s, Congress imposed noise limits on aircraft in 1968. In regulations pursuant to this legislation, the FAA defined three noise levels (Stages 1-3). Stage 1 aircraft, the noisiest, have been phased out of U.S. domestic op- erations. Newly manufactured aircraft must meet Stage 3 performance cri- teria. Legislation passed by Congress in late 1990 sets the stage for phaseout of Stage 2 aircraft. The law prohibits Stage 2 aircraft operations after 1999, but allows the upgrade of Stage 2 aircraft to Stage 3 if the engines are quieted
144 WINDs OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION with "hush kits." Hush kits cost from $1 to 3 million per aircraft and some impose payload penalties (FAA 1989). Air carriers unable to meet the 1999 deadline are allowed to continue to operate Stage 2 aircraft to 2003 if 85 percent of their fleet meets Stage 3 requirements by the end of 1999 and if firm orders exist to upgrade Stage 2 aircraft to Stage 3 by 2003. Airport Restrictions Federal regulations prohibit local restrictions that interfere with safety and interstate commerce, but airport authorities have been allowed to respond to community opposition to noise by imposing some restrictions. Most commercial airports have some noise abatement requirements, which are usually specifications to pilots for approach and departure paths or pref- erential runway use. Many airports have curfews during certain hours. Eight of the 20 largest airports have in-place or proposed restrictions for Stage 2 aircraft (DOT 1990b, 4.4). San Diego International Airport, for example, plans to ban Stage 2 operations by 2000 and similar proposals are being discussed for the Los Angeles International Airport and the major New York area airports. Two smaller airports have restrictive noise regulations: John Wayne Airport in Orange County, California, and Jackson Hole Airport in Wy- oming. The restrictions at John Wayne were disputed and litigated for many years until a compromise was agreed on to expand service until 2004. The noise limits at Jackson Hole, although quite restrictive, apply to an environmentally sensitive area (Grand Teton National Park). The review of local noise restrictions by the U.S. Department of Trans- portation (DOT) concluded that existing noise restrictions are relatively minor and do not represent a barrier to entry; they have little overall impact on new entry because they affect all carriers more or less equally (DOT 1990b, 4.4). Some proposed restrictions, or a proliferation of restrictions such as those at John Wayne and Jackson Hole, however, would further limit capacity and competition. Aircraft Availability During the early 1980s many new entrants were able to begin operations with used or leased aircraft, and despite the backlog on new aircraft orders, which extends to the mid-1990s for some models, a ready supply of aircraft is available through leasing or the used-aircraft market (DOT 1990b, 5.4).
Barriers to Competition 145 About 2,300 Stage 2 aircraft are currently used by U.S. airlines, account- ing for about 60 percent of the total fleet (FAA 1989). Before Congress's latest decision, the FAA projected that about 960 Stage 2 aircraft would still be in operation by 2000 and would account for about 20 percent of the fleet. All new commercial aircraft produced in the United States that weigh more than 75,000 pounds meet Stage 3 criteria, but new Stage 2 aircraft were still being delivered through 1988. The new restrictions on the operation of Stage 2 aircraft after 1999 will subsequently reduce the supply of aircraft or require carriers to retrofit or re-engine existing Stage 2 aircraft. Although hush kits and re-engine pro- grams have been developed for some aircraft and the FAA appears con- fident that they can be developed for others, the cost and availability of conversion programs for some of the major workhorses of the current fleet (e.g., the Boeing 737) are uncertain. In any event, the phaseout of Stage 2 aircraft will increase the cost of entry after 1999 by reducing the supply of used aircraft and increasing the cost of operating used aircraft. In the near term, however, Stage 2 aircraft may be available for a considerable discount. AIRLINE MARKETING STRATEGIES Deregulation has dramatically changed the methods by which airlines market their services. In the competition for passengers, the major carriers have invested heavily in computer reservation systems (CRSs), developed aggressive techniques to recruit and reward travel agents, built brand name loyalty through frequent flier programs, and developed affiliations with commuter airlines to provide feeder traffic to their hubs. These strategic moves by the majorsâspurred in part by the competitiveness introduced by deregulationâhave strengthened the competitive advantages of large carriers. Behind many of the strategic moves of air carriers since deregulation has been an effort to develop and exploit economies of scope. According to Levine (1987), economies of scope in the airline industry are generated by the number of points served by a carrier and should be distinguished from economies of scale, which are achieved in some industries as a function of size. Scope economies are generated as a result of traveler demand for service in more than one city-pair market. For example, a large carrier can enjoy scope economies by advertising on television be- cause it serves many markets, unlike a carrier that serves only a few city pairs. In addition, once a traveler obtains information about service quality,
146 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION cost, and convenience in one city-pair market, he or she forms an impres- sion about that carrier's service in other markets. Economies of scope also occur in the generation of information through CRSs, reward structures for travel agents, incentives built into frequent flier programs, and service patterns made possible by hub-and-spoke networks. Scope economies confer competitive advantages to large air carriers, even in the absence of economies of scale (Levine 1987). CRSs The economies of producing and distributing information in the airline industry are fundamental to postderegulation airline competition (Levine 1987). Air fares and service patterns have become much more complex and change much more often than in the past, contributing to the impor- tance of the CRS and the advantages that these systems confer on their owners. Although CRSs create opportunities for the smallest carriers to have their flights and fares displayed to travel agents nationwide, they also provide important marketing advantages to the carriers that own them. In turn, travel agents, who rely heavily on CRSs, have become a central part of the commercial air travel system, currently booking about 80 percent of all tickets compared with less than 40 percent before deregu- lation (DOT 1990c, 1 ,12). It is estimated that about half of leisure travelers and one-fourth of business travelers do not have a preference for an airline; thus travel agents can play a major role in influencing consumer decisions (DOT 1990c, 1). CRS owners, well aware of the importance of agents in influencing consumers, offer several incentives to influence their behavior, and, through frequent flier programs, offer incentives to travelers even when lower-cost flights might be available. Competitive Advantages of CRS Ownership A series of studies has been conducted on various anticompetitive aspects of CRSs, culminating in an extensive analysis by the DOT (DOT 1988). This study was subsequently updated in the recent report of the Secretary of Transportation's Task Force on Competition in the U.S. Domestic Airline Industry (DOT 1990c). In 1983 the CAB issued an advance notice of proposed rulemaking on carrier practices and requested comments on such issues as display bias, discriminatory pricing, and delayed posting of flights and fares by corn-
Barriers to Competition 147 petitors. In response, the U.S. Department of Justice (DOJ) reviewed the practices of the largest CRS vendors and concluded that these vendors could exercise substantial market power, some practices were anticom- petitive, and regulations were necessary to correct for display bias [both the CAB and DOJ reports were reviewed in detail by the DOT (1988)]. The CAB issued regulations in 1984 that have greatly reduced display biases, but other problems remain.2 A 1985 DOJ study found that CRS vendors could exercise market power and that pricing practices for par- ticipating airlines were discriminatory. It was concluded in a 1986 GAO study that CRSs generated incremental revenues for airlines (GAO 1986). Incremental revenues refer to the additional benefits (additional bookings) that carriers receive as a result of owning a CRS; these incremental rev- enues may have resulted from biased information before the 1984 regu- lations and may still result from a "halo effect," or the tendency for agents to book additional flights on the carrier from which it leases a CRS. (The existence of incremental revenues would indicate competitive advantages for carriers owning a CRS that would constitute a barrier to entry as defined previously.) Also expressed in the GAO report was con- cern that the booking fees charged other carriers were above competitive levels, but the GAO did not have sufficient information to determine whether pricing for participating airlines was discriminatory and called on the DOT to conduct a detailed review (GAO 1986). The 1988 DOT study was based on extensive financial information that CRS owners were required to provide. It was concluded that display bias had been largely corrected by CAB regulations, but carriers owning CRSs were found to have competitive advantages in three key areas: booking fees, halo effects leading to incremental revenues, and advantages such as being better able to coordinate yield management programs (discount pricing). Booking Fees Small carriers and carriers that do not own a CRS must pay a fee to participate in a CRS, and all major carriers except Southwest list their flights in all CRSs. In 1988 the DOT found that the fees charged by the three largest CRS vendors (Apollo, Sabre, and System One) were twice the average cost of providing the service (DOT 1988). In 1988 Apollo and Sabre alone earned roughly $430 million in booking fees from other carriers, a figure equivalent to 15 percent of the net income for the entire domestic industry that year.
148 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINcE DEREGULATION Because almost all carriers list their flights in all CRSs, it would appear that carriers that own a CRS would not want to overprice booking fees to avoid having to pay similarly high prices when their flights are booked through other systems. For American and United airlines, however, this is not the case; bookings on their systems, Sabre and Apollo, account for 70 percent of all booking fee revenues (DOT 1990c, 57). Thus they earn far more from other carriers than they have to pay other carriers for book- ing fees. Incremental Revenues CRSs also apparently generate a halo effect, or the tendency for agents using the systems to book more tickets with the carrier that owns the CRS than with other airlines. The halo effect may result from several different features of CRSs and related host services. In some cases agents and travelers may prefer the host carrier if that carrier has the largest number of flights originating from their community or because of its frequent flier program. The CRS of the host is believed by many agents to have the most recent information about the host's prices and seat availability, and therefore the host carrier may be preferred by agents who want to ensure that clients are buying tickets for seats that are available (one of a number of features referred to as architectural bias). Another cause of the halo effect may be the cooperative business relationship that develops between the carrier that owns the CRS and the agent, which may stem from preferential leasing arrangements, other business arrangements, or commission overrides (volume incentives). Brenner (1990) noted that in some cases travel agents were already favorably inclined toward a certain airline (United or American) before they switched to the CRS owned by that airline; to the extent that this is true the incremental revenues could not be attributed to the CRS but to some other feature of the airline's service. The most recent DOT review of incremental revenues noted that American had made such claims about Sabre, but also noted that most industry officials believe that the advantages of CRSs result from either the business relationship between the airline and agent or architectural bias (DOT 1990c). In addition, the predisposition of some agents toward a particular airline's CRS may have been influenced by other marketing strategies, such as the role of that airline's frequent flier program on local business traveler preferences, the airline's local presence, and the agent's hope to earn extra commissions (Borenstein 1989).
Barriers to Competition 149 Whatever the cause, halo effects appear to "exert a large influence on travel agency booking patterns" (DOT 1988). The DOT's analysis and studies by the carriers themselves indicate that firms that own CRSs appear to earn between 12 and 40 percent more revenues from travel-agent use of their systems than they would if they did not own the CRSs. The DOT's analysis of the rates of return to CRS owners concluded that investments in these systems were "extremely remunerative." The Secretary of Trans- portation's task force reviewed vendor studies of incremental revenues collected by CRS-owning firms; the incremental revenues estimated in these studies were even higher than those estimated by the DOT in its 1988 evaluation of the CRS industry (DOT 1990c). Yield Management In addition to earning excess booking fees from rivals and gaining the advantages of incremental revenues, carriers can use information from their CRSs to devise and continuously monitor sophisticated pricing strat- egies. A carrier can operate a yield management program without a CRS, but having one allows a carrier to observe continuously the booking pat- terns on other carriers and to devise counter-strategies regarding the size of discounts and the number of seats sold. Economies of Scale CRSs display considerable economies of scale because of the sheer scale of investment required to compete and the advantage that an airline that owns a CRS has over a nonairline investor that is interested in developing a CRS. The carriers that developed CRSs have spent hundreds of millions of dollars over many years to bring their systems to their current advanced state (the substantial profits being earned suggest that these investments have been recouped). The incremental revenues CRSs earn are apparently sufficient to allow the carriers to lease such systems to travel agents below cost. Hence, a potential competitor that is not also an airline would have to develop a system more efficient than those already in use in order to attract travel agents and would have to be able to support it without generating incremental airline revenues. Given the high cost of system development and the efficiency and economies of scale of the largest systems, and the contribution made by incremental airline revenues, such new competition appears unlikely. The Secretary of Transportation's Task
150 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION Force on Competition in the U.S. Domestic Airline Industry reviewed the CRS industry and previous DOT studies and concluded that whereas entry into the industry is not precluded, the scale of investment required to compete with existing CRSs and time needed to build up to that scale make entry unlikely (DOT 1990c, 50). Changes To Improve Competition Two major changes are under way in the travel industry that may reduce the competitive advantage of CRS owners. Technological advances in computing may reduce the halo effect, and structural changes in the own- ership and operation of CRSs may reduce some disadvantages for carriers that do not own CRSs (DOT 1990c). Technological advances in computer software are making it easier for travel agents to gain access to the internal reservation files of all carriers that list with a CRS. One probable reason for the halo effect is that agents trust the information from the host carrier because it is more timely. As a result they are less concerned that they might sell a ticket for a flight that is already sold out. Now that they can actually look at the internal files of other carriersâcalled "direct access"âthey have a better idea of the actual availability of flights. The DOT concludes, however, that these changes have not overcome some of the advantages enjoyed by host airlines (DOT 1991, 12598). Gaining entry to direct access is slightly more time consuming for the travel agent, and some routine functions in direct access are more cumbersome (e.g., confirmation of bookings, changes to complex itineraries, and the issuance of boarding passes). In addition, the accompanying booking fee when using direct access is more expensive (this cost, about 15 cents in the most recent estimates by DOT) is paid by the airline rather than the agent. Use of direct access is growing but "many agents are said to be unwilling to use it" (DOT 1990c, 69). Further advances in software may make direct access more popular with agents; what appears to be needed is a single integrated display that allows the agent to review simultaneously the most timely information on the carriers serving the market of interest. Given the pressures under which agents operate, such a display is more likely to be used in cases in which multiple options are available because it would relieve the agent of having to check each carrier's internal file individually. Structural changes in CRS ownership are also occurring. Apollo, the second largest CRS, is operated by Covia Corporation (Table 4-1). Half of Covia (actually 50.1 percent) is owned by United's parent corporation,
Barriers to Competition 151 TABLE 4-I COMPUTER RESERVATION SYSTEMS (DOT 1990c, Table 4.1; Airline Economics, Inc. 1989) Name Owner Percent of Total Market Share (1988) Bookings (segments) Revenue Sabre AMR 43.1 38.8 (American) Apollo' CovialUAL 27.9 27.6 (United, USAir, KLM, Swissair, British Airways, Alitalia) System One5 Texas Air Corporation 13.9 16.6 (Continental, Eastern) Parsc TWA/Northwest 9.4 11.1 Datas III Delta 5.7 5.9 "Affiliated and marketed outside the United States as Galileo; co-owned by British Airways, Swissair, KLM, Alitalia, Austrian Airlines, Aer Lingus, TAP-Air Portugal, Sabena, and Olym- pic. Also affiliated and marketed with Gemini; co-owned by Air Canada and Canadian Airlines International. bAffiliated and marketed outside the United States as Amadeus; co-owned by Lufthansa, Air France, Iberia, SAS, Air Inter, Linjeslyg, Finnair, JAT, Adria, Braathens, Iceland Air. 'Pars and Datas II are in the process of merging. United Air Lines, Inc. The other half is owned by a consortium of airlines, all foreign except USAir. In exchange for buying a share in Covia, USAir apparently attempted to replace some of the booking fees it was paying Apollo with profits from the CRS and the ability to represent Apollo to certain travel agents. Delta Airlines, owner of Datas II, the smallest CRS system, tried to merge Datas II with Sabre in 1989, but the DOJ would not allow it. In February 1990, Delta agreed to merge Datas .11 with Pars, the system owned by Northwest and TWA. The merged system, to be called WorldSpan, is expected to be oper- ational in 1993 and will be sold to travel agents by a company inde- pendent of Delta, Northwest, and TWA (Aviation Week and Space Technology 1990). In its review of these structural changes, the DOT concluded that the multiple ownership of CRSs results in their be- ing operated as separate profit centers; this gives individual airline mana- gers somewhat less ability to use the CRS to thwart competition (DOT 1990a, 17). These technological and structural changes reduce some of the com- petitive advantages of host carriers over other carriers that own CRSs, but several carriers (Pan American, Alaska Airlines, Air West, South- west, and Midway, among others) do not have ownership shares in
152 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION CRSs. These carriers (except Southwest, which does not participate) have to pay booking fees on all systems, and for competitive reasons cannot opt out of listing in each CRS. With all carriers virtually com- pelled to list with all systems, CRSs are able to charge high booking fees without fear of losing business. Other complaints from nonhost airlines about CRS operations include the slower loading of schedule and pricing information for nonhost airlines, fees for no-shows that host carriers are not charged, and the host's advantages in monitoring travel-agent booking patterns and designing commission overrides to influence agents' behavior (DOT 1990c). Whereas the ownership of a CRS gives the large carriers that own them advantages over smaller carriers, ownership of a CRS alone is not a sufficient condition for profitability or even survival. Eastern Airlines, for example, was one of the first carriers to develop a CRS, yet it still failed. In Eastern's case, the carrier's advantage in owning a CRS was offset by a variety of other problems, including poor strategic moves in the early deregulation period and a long history of labor discord. Ownership of a CRS, along with other marketing and size advantages discussed in this chapter, may confer relatively small ad- vantages in individual markets, but in a competitive industry that is operating with narrow margins, small advantages matter. Travel-Agent Incentives Travel agents provide an important service to consumers by supplying efficient access to a complex array of travel options. This importance has grown markedly during deregulation. Whereas travel agents booked only 38 percent of tickets in 1978, by 1988 they booked 80 percent (DOT 1990c, 1, 12). Nevertheless the role of agents as brokers of information and sellers of travel services to consumers is changing as agencies become more closely affiliated with individual air carriers through CRSs and supporting services (DOT 1990c, 8). Almost all travel agencies (94 percent) are automated (by means of CRSs), and 90 percent rely on a single CRS to influence agents; carriers pay com- mission overrides, which, combined with CRS systems, have "had much success in causing agencies to shift travelers to favored suppliers" (DOT 1990c, 2). Some forces at work in the marketplace mitigate the extent to which overrides reduce competition. Most agencies (60 percent) are small, and many do not earn overrides. On the other hand, the travel business
Barriers to Competition 153 is becoming somewhat more concentrated; firms with revenues of $5 million or more grew from 24 to 33 percent of the industry between 1983 and 1987 (DOT 1990c, 11). These large firms seek out commission overrides . Another market force that increases competition is the increasing con- cern of corporate clients about getting the best price. Corporations are more actively monitoring the travel costs of their employees. In addition, a fee system is being developed whereby travel agents earn fees directly from the corporation rather than through a commission. The fee system at present, however, is small: it is available from only about one-fourth of travel agencies and represents less than 1 percent of their revenue (DOT 1990c, 17). Some travel agencies even give large corporate clients rebates on the commission overrides that they earn from air carriers, and large corpo- rations increasingly are arranging discount fares directly with airlines. According to an article in the Wall Street Journal, firms with more than $5 million in annual air travel costs are able to get negotiated discounts for more than 16 percent of tickets (Dahl 1990). The average discount was about one-fourth of the full coach fare. These benefits, however, largely accrue to major corporations able to hire full-time travel managers to oversee their travel agents. Small-business and leisure travelers have less influence over travel agents, because the incentives for travel agents working with these cus- tomers are mixed. Agents may seek out the lowest price for the most acceptable routing in hopes of gaining repeat business. At the same time, the commission system rewards agents on the basis of the total price; hence there is a disincentive to seek out the lowest fare. In addition, some agents know that the commission will be even larger if they book the flight with a preferred supplier (one from whom the agent has been prom- ised a commission override). Through economies of scope, commission overrides can strengthen the competitive position of large carriers or carriers that serve a large number of city pairs from the travel agent's home city (Levine 1987). From a travel agent's perspective, the airline serving the largest network of cities from the agent's home city provides the agent with the most opportunities to sell tickets or to offer alternatives to consumers considering flying on an airline with a smaller local presence. This advantage in offering in- centives to travel agents, combined with frequent flier programs, reinforces the advantages of size (scope), and thereby makes it more difficult for new entrants to compete, be they major airlines entering a market or small, low-cost carriers.
154 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION Frequent Flier Programs Frequent flier programs have grown in popularity since American Airlines introduced its program in 1981, and every major carrier except Southwest offers one. The DOT, on the basis of conversations with airline executives, estimated that roughly 20 to 25 percent of airline passengers accrue fre- quent flier mileage (DOT 1990c, 31). Frequent flier programs create a principal-agent problem (Levine 1987). Carriers award bonuses to frequent travelers, who are mostly business travelers and whose travel costs are often borne by their employers. Most corporate travel managers believe that frequent flier programs, though extremely popular with frequent travelers, result in higher air fares than necessary. Stephenson and Fox (1987), on the basis of a survey of cor- porate travel managers, estimated that frequent flier programs increased air travel costs for businesses by 6 percent, resulting in $4 billion in excess costs in 1987. Some corporations audit air fares to be sure that employees choose low- cost fares, and some firms try to reclaim frequent flier benefits for the company. Most corporate travel managers, however, concede that em- ployees see frequent flier awards as compensation for having to travel extensively, and concerned about employee morale, managers are un- willing to try to reclaim awards for the company. Frequent fliers have agreed in the past that the program influenced their choice of airline (Toh and Hu 1988). At the same time, these frequent fliers stte that schedule frequency and on-time performance are the most important factors in selecting an airline. The majority of respondents in the survey by Toh and Hu (1988) belonged to more than one program, so they apparently had the option of maximizing schedule convenience while still earning mileage credit. Such a choice is becoming more dif- ficult, because some programs now impose expiration dates, thereby mak- ing it harder to win awards while participating in multiple programs. A more recent survey of travel agents indicates that most business travelers (roughly 80 percent) choose airlines in order to accumulate frequent flier miles at least half of the time (GAO 1990). The Secretary's task force on competition conceded that frequent flier programs influence consumer choice and that the carriers with the largest route networks have an advantage over their smaller competitors (DOT 1990c, 41). The task force's report noted that small carriers have formed alliances with other carriers to make their networks more attractive but concluded that frequent flier programs make it more difficult for small carriers to compete. Large carriers with extensive route networks will
Barriers to Competition 155 naturally have the most attractive systems because they can offer the traveler more trip choices with which to earn mileage and more exotic vacation possibilities as rewards. Cooperative arrangements between small carriers, who are sometimes competitors, tend to be short-lived. The scope economies enjoyed by larger airlines build consumer loyalty, particularly among business travelers who are less concerned about price and with whom carriers can exploit the principal-agent relationship, thereby conferring advantages to size that cannot necessarily be offset by a smaller airline that is attempting to compete only on the basis of price (Levine 1987). Code Sharing Code sharing, the listing of regional or commuter airlines with the code of a major or national carrier in a CRS, represents one of the more powerful examples of vertical integration in the deregulated airline industry and further highlights the importance of CRSs in airline competition. As dis- cussed in Chapter 3, commuter carriers moved into thin markets abandoned by the jet carriers in the early years of deregulation. The freedom provided by deregulation allowed carrier managers to allocate equipment and per- sonnel in line with costs, and many communities began receiving service by means of turboprop aircraft that carry 60 or fewer passengers. Com- muter carriers often provided a low-cost operation, principally because of their use of lower-cost equipment more suited to their specific market characteristics and to a lesser extent because of lower overall labor costs. In effect, the commuter carriers began feeding service to the jet carriers at costs below those that the jet carrier could achieve over the same route. With the development of hub airports, major carriers soon realized that they could attract passengers traveling beyond their hubs by advertising their affiliations with the commuters serving their hub. Affiliations be- tween carriers of differing sizes, in which schedules and baggage handling are coordinated, predate deregulation, but use of the larger carrier's CRS code by the smaller airline feeding traffic to the larger carrier places these affiliations in an entirely new context. Listing the commuter with the code of the major in the CRS provides itineraries to travel agents and allows little-known commuter airlines to benefit from the brand name of the major carrier (Levine 1987, 440-441). CRSs also give greater weight to itineraries involving code sharing, which means that they will be listed before other possible interline connections and are therefore more likely to be seen and chosen by travel agents (Oster and Pickrell 1988).
156 WINDS OF CHANCE: DOMESTIC AIR TRANSPORT SINcE DEREGULATION Code-sharing agreements became widespread by 1985, and the rela- tionships between the commuters and major carriers became so important that many commuters were acquired in whole or in part by their affiliates (FAA 1989). The majors were partly motivated by the desire to control the commuters' low-cost feed to their hubs but were also concerned about gaining enough control to ensure that quality and safety were being pro- vided consistent with the firm's image. As an example of the degree of vertical integration that has occurred in the commuter industry, the top 50 commuter airlines account for about 92 percent of all commuter airline traffic, and all but 11 of the top 50 commuters share CRS codes with a major or national airline (Regional Airline Association 1988). Consumers may receive some indirect benefits as a result of code sharing. Major airlines entering into code-sharing agreements are likely to impose commensurate service requirements on the commuters and may assist them in the purchase of higher-quality aircraft. Efficient, integrated service between large and small carriers, however, predates code sharing, and information about such relationships was displayed in CRSs before code sharing. Equally convenient connections, indeed even those that may better meet consumer preferences, however, may not be as fairly treated in CRSs because of the preference given to online connections (Oster and Pickrell 1988). Hence, consumers may not always benefit from code sharing. The effects of code sharing on competition are somewhat less direct, but still important. Because the commuters provide service on routes with relatively little traffic volume, many routes can only be served by a single commuter. By effectively controlling the commuter traffic arriving at its hubs through CRS listings or outright ownership of the commuter, the major further protects dominance at its hub (Levine 1987; Oster and Pickrell 1988). Code sharing also makes it difficult for other regional airlines to compete with the code-sharing partner in markets in which sufficient demand would support more than one carrier because only one commuter will be able to share the CRS code; this could ultimately lead to reduced competition in the regional airline segment of the industry and higher fares for consumers traveling from small communities (Oster and Pickrell 1988). In addition, by extending the service networks offered by large carriers, often at a cost lower than the large carrier could provide directly, code sharing enhances the economies of scope enjoyed by larger carriers (Levine 1987). The advantages that CRSs, travel-agent commissions, frequent flier programs, and code sharing confer to carriers may be small individually, but they have a more powerful influence by the extent to which they interact with one another and with hub-and-spoke networks to create econ-
Barriers to Competition 157 omies of scope. Large carriers can exploit advantages of size even when these advantages do not confer lower costs through economies of scale (Andrandi, etal. 1989). Levine (1987) makes a compelling argument that achievement of scope economies helps explain why incumbent airlines with higher cost structures than those of new entrants have been able to survive and how some have even extended their dominance during de- regulation. The development and exploitation of scope economies may also explain why the new entrants that have survived have focused on specific regional or niche markets. New entrants such as People Express and the reconstituted Braniff, which tried to compete nationally with the incumbent majors, ultimately failed. Whether America West will be able to do so remains to be seen. MERGERS AND ACQUISITIONS The inability of contestability theory to account for barriers to entry, the advantages that large carriers have through economies of scope, and the oligopolistic character of the industry raise concern about whether the continued consolidation of the airline industry will result in insufficient competition to discipline pricing. Enforcement of antitrust laws has long been the traditional policy instrument for protecting against excessive concentration in U.S. industries, and deregulation shifted policy to reliance on antitrust enforcement to protect consumers instead of reliance on a public utility form of regulation. The Airline Deregulation Act of 1978 required the CAB to treat airline mergers and acquisitions in a manner more consistent with the antitrust standards applied to almost all other industries (Meyer and Oster 1981, 228; Bailey et al. 1985, 175). According to the provisions of the act, application of the Sherman Act test would be used to prohibit mergers that would result in a monopoly in any region of the country. Application of the Clayton Act test would be used to prohibit transactions that would have the effect of substantially lessening competition or that would tend to create a monopoly. The Deregulation Act, however, did give the CAB (and later the DOT) somewhat more latitude in weighing the benefits of mergers (transportation convenience and needs) than is applied in antitrust cases in other industries. In the first few years, before the CAB was dissolved and its antitrust authority shifted to the DOT, several mergers were permitted that were "end-to-end" in character. These mergers allowed carriers to merge that did not serve overlapping markets, many of which actually enhanced
158 WiNos OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION service by reducing transaction costs, and apparently accomplished this end without reducing competition. For example, Pan American was al- lowed to merge with National, and North Central with Southern to become Republic, which then absorbed Hughes Air West. Several proposed merg- ers were rejected. Some were disapproved because the carrier's routes were "parallel mergers"; that is the carriers served too many overlapping routes (Eastern-National). Some were disapproved because of concern about hub dominance and barriers to entry (Continental-Western). When the DOT was given authority over mergers in 1985, the number of mergers and acquisitions increased from 8 between 1980 and 1984 to 18 in 1985 and 25 in 1986. Most of the mergers did not raise significant competitive issues; many of the small carriers involved in them were in financial difficulty and would have gone bankrupt had they not merged (CBO 1988). Some end-to-end mergers may have even facilitated com- petition because the combination of two carriers serving different markets helped build a broader, more competitive network (Meyer and Oster 1987). Two mergers among major airlinesâTrans World with Ozark and Northwest with Republicâwere opposed by the DOJ but approved by the DOT. The DOJ argued that these parallel mergers would facilitate hub dominance and create significant barriers to entry for competitors, but the DOT approved them, noting that both the Minneapolis and St. Louis airports had gates that new entrants could use. As indicated in the previous section, hub dominance, when combined with the influence that the hub- bing carrier has over local travel agents through volume incentives and leased CRS facilities and the influence that frequent flier programs have on frequent, mostly business, travelers, creates barriers to entry and com- petition (Levine 1987). The availability of gates is a necessary condition for competition, but is not a sufficient one. Mergers and acquisitions, which had occurred during the regulated era, increased after deregulation; almost all proposals put before the DOT, when it had antitrust authority, were approved. The mergers in the 1985- 1987 period returned the industry to the level of concentration that existed in 1978 (although not all increases in size have been beneficial to the carriers involved) (Jordan 1990). More careful scrutiny of future mergers by the DOJ may enhance competition or at least keep it from diminishing. Because of the complexity of airline networks and competition, it is difficult to specify in advance the conditions under which mergers or acquisitions will be anticompetitive. Some mergers, for example, may facilitate competition between hub networks but simultaneously create opportunities for hub dominance. An overarching question regards the total number of major carriers that are required to maintain an adequate
Barriers to Competition 159 level of competition. Although of the number of firms that are required to ensure adequate competition necessarily involves some speculation, the main criterion for the adequacy of competition nationwide is the level of competition for passenger flows between competing hub systems. As shown in Chapter 3, consumers receive the largest benefits when three or more competitors are competing in the same market, especially if one of the competitors is a new-entrant airline with a low-price marketing strat- egy. Having only three carriers nationwide, however, would probably not be adequate to ensure this level of competition. As indicated in the previous sections of this chapter, sufficient barriers to entry exist to prevent three major carriers from being able to compete with each other for over-hub traffic from every major spoke city. Five or six major airlines, however, would probably constitute a sufficient number of hub systems to ensure the presence of three or more competitors in most major spoke markets, especially when several additional healthy regional and national carriers are offering consumers alternatives in specific regional or niche markets. Five or six nationwide firms that compete with each other at all the large commercial airports may provide much stronger competitive pressure to hold down costs and fares than would 10 or 15 carriers competing in less extensive networks. The fewer the number of firms, however, the easier it is for them to form and enforce a tight oligopoly in which industry output is lower and fares are higher than would be the case in a competitive market. For fewer than five or six carriers, the results would depend on the circumstances of the carriers, their markets, and the vigor of the competition among them in the future, but, as indicated above, three major nationwide carriers are likely to be too few to ensure adequate competition. In a severe economic downturn or crisis affecting carrier costs, gov- ernment regulators could be faced with the problem of failing firms, which, if allowed to merge with major competitors, could substantially reduce the total number of carriers and competition. These cases have to be considered as they arise. In general, parallel mergers in which carriers have substantial overlap- ping routes or share a hub are likely to reduce competition. Such mergers give carriers too many opportunities to dominate markets in which the prospects for entry is limited. In contrast, a merger that combined carriers with complementary routes may not be anticompetitive even if it created a large nationwide carrier that would dominate a hub airport. The extended network created would help make the carrier more competitive in over- hub traffic, even though it may give it the opportunity to exercise market power in the hub market. As described in an earlier section of this chapter, there is a certain inevitability about hub dominance, although, as described
160 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION in Chapter 3, markets dominated by single carriers account for a relatively small share of total trips. A regulatory or legislative remedy for this problem that would not simultaneously undermine the service benefits and the competition in longer distance markets that hub networks provide is not apparent. One approach to minimizing the anticompetitive impact of such mergers would be to ensure the availability of gates and re- lated facilities that might be used for establishing spokes from other carrier's hubs. Because hub networks provide consumers with more choices in over- hub traffic and because the largest benefits occur in markets with three or more carriers, the effect that mergers or major acquisitions would have on the ability of three or more carriers to compete in over-hub traffic would serve as a useful measure of the Clayton Act test of substantially lessening competition. Some mergers can be imagined that would substantially reduce com- petition (Levine 1987, 486-487). In addition to the shared-hub issue discussed previously regarding parallel mergers, concentration among car- riers in regional markets that would reduce the choices of consumers to making connections through just one hub may not be acceptable. A merger that was otherwise acceptable that would give carriers monopolies in certain city-pair markets may justify divestiture of key assets (e.g., slots or gates) that limit competition in those markets. Some have argued that reliance on antitrust enforcement alone will not be sufficient to correct for the lack of contestability in airline markets (Dempsey 1990). Although the ease of entry required of contestability theory is not characteristic of the airline industry, and the industry is oligopolistic, the case for market failure that would warrant the more interventionist policy espoused by Dempsey (1990) has not been made. To the contrary, competition has increased since deregulation, there is little evidence of monopoly pricing, and most passengers are traveling in competitive markets. Antitrust enforcement serves adequately to check the anticompetitive features of other oligopolistic industries in the econ- omy (Neale and Goyder 1980). Reliance on antitrust enforcement for the airline industry (if such enforcement is vigorous) is likely to check the tendency toward cartel-like behavior and therefore help keep the industry competitive. The DOJ has been more vigilant on antitrust issues than was the DOT. Although deregulation has created benefits for most consumers, bar- riers to entry provide opportunities for carriers to exercise market power in specific markets. Dempsey (1990) has urged policies such as the setting of price ceilings to protect consumers in specific markets and a
Barriers to Competition 161 return to "enlightened" regulation. Neither of these proposals appears promising. The difficulty with setting price ceilings in specific markets is that pricing in hub-and-spoke systems is a function of a complex set of joint costs and pricing strategies. It is not apparent how one could establish the maximum fair price for one market without specifying the appropriate pricesâand mix of quantities of outputâfor all of them. Dempsey (1990) asserts that this might be accomplished through some new and improved form of regulatory practiceâone that would be more insulated from politics and the influence of special interests. The study committee, however, is not sanguine that another group of experts, even one more insulated from outside pressures, would be any more effective at second-guessing the marketplace than were the experts at the CAB. SUMMARY The financial risks of new entry, airport capacity limits at many airports, and the influence of airline marketing strategies on travel agents and consumers confer considerable competitive advantages to carriers able to exploit them. Incumbents can build powerful hubs to deter competitors, take advantage of scarcity by trying to raise their rivals' cost of entry at airport terminals, and limit the sales of slots to noncompetitive airlines. By building extensive hub-and-spoke service networks and through CRSs, incentives to travel agents, frequent flier programs, and code sharing with commuter carriers, major carriers can influence consumer choices through economies of scope in ways that smaller airlines cannot. Not all large airlines, of course, have been able to exploit scope econ- omies. Pan American, for example, has neither a dominated hub nor a share in a CRS. Some large carriers, such as Eastern (now liquidated), Continental, and Trans World, have some of these advantages, but they have been offset by poor management in the past, contentious labor- management relations, a heavy debt burden, or some combination of these disadvantages. Smaller, new-entrant airlines, such as Midway, America West, and Southwest, have been able to survive and grow by finding niches in which they are competitive but are finding expansion into more markets a slow and difficult process. This difficulty is likely to continue. Economies of scale deter new entry into CRSs, and airport terminal capacity is con- strained at major cities that are critical to the development of a national route network. The competitive advantages enjoyed by major carriers
162 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION through economies of scope appear to be contributing to the lack of entry into the airline industry. Corrections for the shortcomings in the competitiveness of deregulated air transportation markets, however, can be difficult to devise. Intervening in airport leasing arrangements, for example, would require voiding con- tracts that were entered into freely, something that Congress is loath to do and, indeed, might be illegal. In the case of competition at hub airports, there appears to be an in- evitable drift toward domination at hub airports with relatively small local traffic bases because of the nature of competition in hub-and-spoke sys- tems. A remedy for the opportunity to earn monopoly rents in some city- pair markets that does not simultaneously undermine the service benefits that hubs offer is not apparent. Public policy can strive to ensure that the opportunity for entry exists at dominated airports, something that Trans- portation Secretary Samuel Skinner has hoped to achieve by allowing airports to charge head taxes, but at many airports this may not be a sufficient condition to attract additional competition of significant scale to discipline pricing. For code sharing, which appears to inhibit competition in the commuter segment of the industry, prohibiting alliances between major or national carriers and regional carriers may only encourage the acquisition of the smaller carriers and thereby hasten industry concentration. Other competitive issues, however, may be more amenable to incre- mental adjustments that would improve competition. For example, pro- viding travel agents with the ability to shift freely among CRSs would give the CRS owners less of a competitive advantage. In addition, better consumer information about the presence of commission overrides would reduce potential biases in information provided by agents. (See Chapter 8.) The current system of slot allocations gives incumbent carriers a sub- stantial competitive advantage because of the size and importance of mar- kets such as Chicago, New York, and Washington, D.C. Slot allocations are a poor way to ration scarcity, and the current system for selling and leasing slots does not appear to work well. As discussed in Chapter 6, peak-period pricing is a better alternative for rationing scarce runway capacity. Finally, more vigorous antitrust enforcement is needed than has been practiced in the past and will help mitigate the anticompetitive features of the emerging oligopoly. Parallel mergers involving shared hubs, the acquisition of gates at a hub airport that would give control to a carrier already dominating the airport, and mergers resulting in regional mono- polies are likely to be anticompetitive. Some proposals for mergers or
Barriers to Competition 163 acquisitions, however, may improve competition. Because the largest consumer benefits occur in markets with three or more carriers, the effect that mergers or major acquisitions would have on the ability of three or more carriers to compete in over-hub traffic would serve as a test of the anticompetitive features of proposed mergers and acquisitions. NOTES The preliminary results of the AOCI survey are published in the Secretary's Task Force on Competition in the U.S. Airline Industry (DOT 1990b, 3.2-3.5). The DOT report notes that the results were only partially audited at the time of pub- lication and might be subject to change. The final results of the survey had not been published as this report neared completion. The definition applied to the availability of gates in the survey was restrictive but also reflected the conditions required for competitive gates. Facilities could not be remote, and had to include an aircraft parking position, facilities for passenger loading and unloading, a pas- senger holding area, check-in counters, space for baggage make-up and operations, and baggage claim facilities. The CAB regulations governing CRSs were due to expire at the end of 1990, but the DOT extended its deadline for making a decision about any revisions to the rules until November 1991. The DOT has rulemaking under way to extend the rules. The DOT's preliminary findings indicate that the CAB rules did not go far enough in ensuring easy switching between CRSs or in reducing host carrier's market power. Even "large" travel agencies can be relatively small in terms of total staff. For example, an agent with four or five CRS terminals can generate $5 million in gross sales annually. REFERENCES ABBREVIATIONS CBO Congressional Budget Office DOT U.S. Department of Transportation FAA Federal Aviation Administration GAO General Accounting Office Airline Economics, Inc. 1989. The Airline Quarterly. Washington, D.C., Spring, p. 28. Airport Magazine. 1989. Hubbing: A Way of Life? Vol. 1, No. 6, Nov/Dec., p. 35. Andrandi, B., G. Chow, and R. Gritta. 1989. The Market Share, Cost and Profit Relationship in the Airline Industry. Journal of the Transportation Research Forum, 1989. Vol. 29, No. 2, pp. 223-227. Aviation Week and Space Technology. 1990. Need for Information Processing May Spur Diversification of CRSs. Nov; 26, p. 90.
164 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION Bain, J. 1968. Industrial Organization, 2nd ed., John Wiley and Sons, Inc., New York, N.Y. Borenstein, S. 1989. Hubs and High Fares: Airport Dominance and Market Power in the U.S. Airline Industry. Rand Journal of Economics, Vol. 20, No. 3, pp. 344-365. Brenner, M. 1990. Analysis of the Airline Concentration issue. Melvin A. Brenner and Associates, Weston, Conn., July. Caves, D., L. Christensen, and M. Tretheway. 1984. Economics of Density versus Economies of Scale: Why Trunk and Local Service Airline Costs Differ. Rand Journal of Economics, Vol. 15, No. 4, Winter, pp. 47 1-489. CBO. 1988. Policies for the Deregulated Airline Industry. Washington, D.C., July. CBO. 1984. Financing Airports in the 1980s. Washington, D.C. Dahl, J. 1990. Big Companies Get Fat Travel Discounts, But Small Firms and Con- sumers May Pay. Wall Street Journal, March 12, p. BI. Dempsey, P. 1990. Flying Blind: The Failure of Airline Deregulation. Economic Policy Institute, Washington, D.C. DOT. 1988. Study of Airline Computer Reservation Systems. Report DOT-P- 88-2. May. DOT. 1 990a. Secretary's Task Force on Competition in the U.S. Domestic Airline Industry: Executive Summary. Feb. DOT. I 990b. Secretary's Task Force on Competition in the U.S. Domestic Airline Industry.' Airports, Air Traffic Control, and Related Concerns (Impact on En- try). Feb. DOT. 1 990c. Secretary's Task Force on Competition in the U.S. Domestic Airline Industry: Airline Marketing Practices: Travel Agencies, Frequent-Flyer Programs, and Computer Reservation Systems. DOT. 1991. 14 CFR, Part 255, Computer Reservation Systems; Notice of Proposed Rulemaking. Federal Register. Vol. 56, No. 58, March 26, pp. 12,586-12,634. FAA. 1989. Report to Congress: Status of the U.S. Stage 2 Commercial Airline Fleet. U.S. Department of Transportation. GAO. 1986. Airline Competition: impact of Computerized Reservation Systems. Wash- ington, D.C., May. GAO. 1990. Airline Competition: Industry Operating and Marketing Practices Limit Market Entry. RCED 90-147. Washington, D.C. Jordan, W. 1990. U.S. Airline Concentration Following Deregulation and Mergers. Paper Presented to the Transportation Research Forum, 32nd Annual Forum, Long Beach, Calif., Oct. 8. Levine, M. E. 1987. Airline Competition in Deregulated Markets: Theory, Firm Strategy and Public Policy. Yale Journal on Regulation, Vol. 4, Spring, pp. 393-494. MacAvoy, P., and J. Snow (eds.). 1977. Regulation of Passenger Fares and Com- petition among the Airlines. Ford Administration Papers on Regulatory Reform. American Enterprise Institute for Public Policy Research, Washington, D.C. Meyer, J., and C. Oster, Jr. 1987. Deregulation and the Future of intercily Passenger Travel. MIT Press, Cambridge, Mass. Neale, A., and D. Goyder. 1980. The Antitrust Laws of the U.S.A.: A Study of Competition Enforced by Law, 3rd ed. Cambridge University Press, New York, N.Y. Oster, C., Jr., and D. Pickrell. 1988. Code Sharing, Joint Fares, and Competition in the Regional Airline Industry. Transportation Research, Vol. 22A, No. 6, Nov., pp. 405-417.
Barriers to Competition 165 Oum, T., and M. Tretheway. 1990. Airline Hub and Spoke Systems. Journal of the Transportation Research Forum, Vol. 30, No. 2, pp. 380-393. Regional Airline Association. 1988. Annual Report of the Regional Airline Associa- tion. Washington, D.C. Stephenson, F., and R. Fox. 1987. Corporate Attitudes Toward Frequent Flier Pro- grams. Transportation Journal, Vol. 27, No. 1, Fall, pp. 10-22. Toh, R., and M. Hu. 1988. Frequent Flier Programs: Passenger Attributes and At- titudes. Transportation Journal, Winter, pp. 11-22.
PART HI Effects and Responses in the Public Domain
5 Commercial Aviation Safety Nearly every writer on aviation safety since 1978 has pointed out that the safety of the traveling public was not deregulated. Indeed, the Airline Deregulation Act of 1978 (Public Law 95-504, Sec. 102, Oct. 24, 1978) begins with this statement: "The Board shall consider the following, among other things, as being in the public interest. . . (1) The assignment and maintenance of safety as the highest priority in air com- merce." The changes in air travel since deregulation, however, have raised the question of whether the intensified competition and traffic growth stimulated by deregulation have increased the risks to the traveling public. Also of concern is the ability of the government to inspect air carrier operations, certify aircraft, and provide a safe air traffic control (ATC) system for a rapidly growing and evolving industry. The discussion begins with a summary of trend data on accidents and accident rates in commercial aviation. The intent of this chapter is to determine whether changes in safety can be traced directly or indirectly to deregulation. The main points are that systematic links between dereg- ulation and air carrier accidents and accident rates cannot be established, directly, nor can any linkage be established by using proxies for accidents, such as nonaccident safety indicators. Such linkages are difficult to es- tablish, however, because commercial aviation has attained such a high level of safety that small changes are difficult to distinguish from random fluctuations. Even so, there is reason for concern and action. Evidence of maintenance violations at major air carriers, even if isolated events, requires improved government action to deter future violations and to maintain the public's trust. (The difficulties faced in improving the federal response are addressed in Chapter 7.) 169
170 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION Fatal Accident Rate 0.4 0.3 0.2 0.1 19501955 1960 1965 1970 1975 1980 1985 1990 Year - All Scheduled - Pa8senger Service FIGURE 5-1 Safety record of U.S. airlines, fatal accidents per 100,000 departures, Part 121 scheduled domestic service (data from NTSB and FAA). SAFETY TRENDS Commercial air travel on major airlines has become progressively safer over time for a variety of reasons, including advances in aircraft design and ATC, and in pilot experience, training, and certification. The fatal accident rate shown in Figure 5-1 compares the rate of fatal crashes over time using the number of departures as a measure of the exposure to risk. The number of departures provides a better measure of risk of serious accidents than distance traveled because the majority of aircraft crashes that result in the destruction of the aircraft occur during take-off or landing; roughly 5 percent occur during the cruise portion of the flight (Lautman and Gallimore 1989). [Use of aircraft mileage or revenue passenger miles (RPMs) as a measure of risk exposure shows an even sharper decline than that shown in Figure 5-1, in part because of the trend toward increased load factors and longer trip lengths]. The fatal accident rate for all carriers operating scheduled domestic service under Title 14, Part 121, of the Code of Federal Regulations (14 CFR 121) is shown in Figure 5-1. (Fatal accidents that resulted from suicide or sabotage are not included in the data presented in this chapter.) Part 121 is the part of Federal Aviation Regulations that applies to carriers operating aircraft (almost exclusively jet aircraft) with seating for 60 or more passengers and cargo carriers operating jet aircraft.1 Throughout the 1980s the fatal accident rate for Part 121 carriers generally moved down- ward until 1989, a year during which several anomalous accidents occurred
Commercial A viation Safety 171 in which individuals were killed; three of these accidents involved all- cargo carriers. It is possible to separate the accident and fatal-accident trend involving only regularly scheduled passenger trips from 1962 on- ward.2 This trend follows the overall Part 121 trend fairly closely but does not increase as sharply in 1989. Subsequent sections of this chapter focus on the possible effects of deregulation on jet carrier safety in terms of changes in maintenance spending and pilot experience and the safety performance of new-entrant airlines. Studies that have attempted to disaggregate the effects of these changes on safety are reviewed. Although this chapter is limited to an assessment of changes that are plausibly related to deregulation, it is important to bear in mind that deregulation is but one of many possible influences over the safety of jet carriers. For example, part of the im- provement in the trend in Figure 5-1 is attributable to the improved safety performance of each new generation of jet aircraft. The first generation of jets introduced in the late 1950s (such as Comet IV's, 707's, and DC- 8's) had substantially higher fatal accident rates than those introduced in the 1960s (such as 727's, 737-100's/200's, and DC-9's) (Weener, n.d.). In general, the jets introduced during the early 1970s (mostly wide bodies) had even better safety performance than earlier generations, and additional gains were achieved by the aircraft introduced in the 1980s (MD-80's, 737-300's/400's, 757's, 767's, BAE 146's, and A310's). The improved safety performance of these aircraft is no doubt correlated with improved pilot training and ATC procedures. This changing mix of aircraft and improved safety features has contributed to improvements since deregu- lation, but the overall effects of the newest generation of jets have been gradual and, until recently, fairly small. The most popular aircraft used during the 1980s (727's, 737-100's/200's, 747's, DC-9's, DC-lO's, L- 1011's and A300's) have accounted for the vast majority of jet usage; all were introduced before deregulation. Toward the end of the last decade, the latest generation of jets, which as a group have the safest records, came into more widespread use, but they accounted for less than 15 percent of the total U.S. domestic jet fleet in 1989 (FAA 1989a). For this reason the subsequent analyses that disaggregate the changes in safety due to changes attributable to deregulation are probably not significantly affected by the changing composition of the jet fleet. If anything, the concern about aging aircraft discussed later is partly an outgrowth of the slow change in the composition of the fleet. Although the trends indicate that safety has improved since deregulation, public concern about safety remains high. In fact, a Gallup Poll conducted in mid-1989 indicated that 63 percent of the American public had lost
172 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION Accidents and Fatalities 400 300 200 100 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 Year Total Accidents M Fatalities FIGURE 5-2 Accidents and fatalities, Part 121 scheduled domestic ser- vice (data from NTSB). confidence in the safety of airline travel. The extensive media coverage of jet crashes may encourage the public to focus more on the many deaths that can occur in a rare catastrophe instead of the overall improvement in the rate of fatal accidents (Barnett 1990). Regardless of the cause, public concern directly influences both airlines and the priorities of elected of- ficials. Barnett (1990) points out that the 1988 terrorist bombing of Pan American flight 103 over Lockerbie, Scotland, caused sharp declines in international travel on U.S. carriers. Pan American estimated its losses as a result of the Lockerbie crash at more than $250 million. Public fears have increased pressure on elected officials to minimize further the risk of terrorism through improved airport security. Although the trends in- dicate that commercial aviation safety has improved over time (even when the fatal accident rate includes accidents attributable to sabotage), it does not appear from examples such as this that the public has accepted this level as safe enough. Accidents and Fatalities During Deregulation Since 1978 the total number of aircraft departures has increased sharply, and the absolute number of accidents involving airlines offering regularly scheduled passenger service under Part 121 regulations has fluctuated around roughly 20 each year (Figure 5-2). The number of fatalities varies considerably from year to year because of the large number of lives that
60 50 40 30 20 10 Accidents 70 Commercial A viation Safety 173 1978 1980 1982 1984 1986 1988 1990 Year Accidents M FataIitie FIGURE 5-3 Accidents and fatalities, Part 135 scheduled service (data from NTSB). can be lost in a single catastrophe. In 1980 and 1984, however, no fatal accidents occurred, and in 1986 the single fatal accident that occurred resulted in the death of one person. They are the only years in the history of commercial aviation during which such safe performance has been achieved. Commuter carriers operate under Title 14, Part 135, of the Code of Federal Regulations. Before 1978, these carriers, operating aircraft with 30 seats or fewer, had not been subject to regulation of fares and service by the Civil Aeronautics Board (CAB). After the Airline Deregulation Act of 1978 defined commuter carriers as those operating aircraft with 60 or fewer seats, the FAA revised the Part 135 safety standards and defined commuter carriers similarly. Although the effects of deregulation on com- muters were indirect, the subsequent abandonment of thin markets by the major and national carriers provided many new opportunities for com- muters, and as discussed in previous chapters, this segment of the industry has grown sharply since deregulation. Commuter carriers were involved in progressively fewer accidents during the early years of deregulation, but from the mid-1980s onward no trend is evident (Figure 5-3). Accident and Fatality Rates The trend in accident rates (per 100,000 departures) for major jet carriers has been fairly stable since 1978 (Figure 5-4). Fatal accident rates have
174 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION Accident Rate 0.5 0.1 1978 1980 1982 1984 1986 1988 1990 Year - Total Accident Rate - Fatal Accident Rate FIGURE 5-4 Accidents per 100,000 departures, Part 121 carriers, Sched- uled service (data from NTSB). fluctuated from year to year at a very low level and, with the exception of 1989, are generally moving downward. In contrast with the composition of the jet fleet discussed above, the marked improvement in commuter airline safety shown in Figure 5-5 is influenced by the improved operating characteristics of the aircraft in use and by a major tightening of the Part 135 safety standards that occurred in 1978 (Oster et al. in press). The largest commuter carriers, which account for the vast majority of passenger Accident Rate 3.5 o 1978 1980 1982 1984 1986 1988 1990 Year - Total Accident Rate Fatal Accident Rate FIGURE 5-5 Accidents per 100,000 departures, Part 135 carriers, sched- uled service (data from NTSB).
Commercial Aviation Safety 175 trips, have increasingly relied on turboprop aircraft instead of piston air- craft. Turboprop engines, which operate with turbines similar to those in jet engines, have fewer parts, operate with greater reliability, and are easier to maintain than piston engines. DEREGULATION AND ACCIDENTS Most of those expressing concern that safety has been compromised by deregulation have argued that airline management's closer attention to the "bottom line" will inevitably result in less attention to safety (Gray 1987; Nance 1988; O'Brien 1988; Lederer and Enders 1988). According to this line of reasoning, before deregulation the established trunk carriers, with protected markets and cash flow, were able to spend more for training and maintenance, and to maintain large staffs of engineering personnel in order to ensure a higher level of safety than that required by FAA regu- lations. The growth in travel, partly spurred on by deregulation, and the relatively open entry permitted under deregulation (for all types of carriers) have also greatly increased the demand for pilots, resulting in a work force of newly hired pilots who are less experienced than those in previous years. Other concerns have also been raised about deregulation; for example, Barnett and Higgins (1989) suggested that new-entrant airlines are less safe because of their lack of experience. In addition, Nance (1988) ques- tioned whether the replacement in some communities of the jet service of trunk airlines with the turboprop service of commuter airlines has increased the risk to travelers. Diminished safety as a result of deregulation could be manifested di- rectly and indirectly. The direct effect would be more accidents caused by inadequately maintained aircraft, or pilots with limited experience, or one of the other concerns mentioned previously. The indirect effect might be an increased probability of accident occurrence, which initially might not be directly observable, but could result in accidents in the future. Methodological Issues In any effort to analyze the possible effects of changes in airline operating practices on the number of accidents or the probability of accident oc- currence, one must recognize that (a) because of the excellent safety record achieved by domestic airlines, small changes in risk are difficult to dis-
176 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION tinguish from random fluctuations, and (b) carrier operations are but one of the major factors that affect safety. To amplify the second point, con- sider the many aspects of the aviation system that work together to prevent accidents. Commercial aviation movements in the airways are controlled by stringent separation standards to avoid collisions. Pilots undergo ex- tensive (and recurrent) training and certification procedures. Aircraft must meet many safety standards, including the provision of redundant safety features in design to help ensure that the failure of one component does not result in a catastrophe. Aircraft performance is closely tracked by manufacturers and the FAA to detect possible problems and correct for them. In addition, air carriers are required to follow maintenance practices designed to reduce the possibility that accidents may be caused by deferred or improper maintenance. These aspects of the system work together in a complex way to produce a remarkably safe mode of travel. Deregulation directly affected only air carriers, but during the same period in which the commercial aviation system was expanding rapidly most of the FAA's air traffic controllers were fired, and the FAA was subject to personnel ceilings throughout the early 1980s that reduced the size of its safety-related work force. Deregulation had no direct effect on the manufacture of aircraft, but, to the extent that it stimulated demand, in the short run it may have contributed to the use of more aircraft that are nearing the end of their service lives. (The safety issues associated with aging aircraft are discussed in a subsequent section of this chapter.) Safety-related incidents may occur because of the breakdown of one aspect of the system, but accidents usually occur because of a breakdown of more than one part of the system. For example, in February 1991 a USAir Boeing 737 that was cleared for landing at Los Angeles Airport crashed into a SkyWest Fairchild Metroliner that was on the runway, resulting in 34 fatalities. The controller, who was busy handling other aircraft and distracted by a misplaced flight strip, apparently lost sight of the Fairchild or forgot that it had been cleared to taxi into position. Ground radar that may have indicated the presence of the Fairchild on the runway was not working. A light pole and the glare from the light may have obscured the controller's ability to see the Fairchild. The pilot of the Fairchild had apparently turned on the strobe light, but it may not have been detectable at the angle of the approaching 737. The pilot of the 737 may not have been able to discern the Fairchild's taillight aiitid a confusing array of runway lights, and did not notice (nor would be expected to notice) that the Fairchild had been cleared to taxi into position for departure just 67 seconds before the 737 was cleared to land on the same runway.
Commercial Aviation Safety 177 A change in any one of these conditions might have been sufficient to avert the crash. In short, accidents and safety-related incidents are complex events in which primary cause is often difficult to determine. Carrier Finances and Accidents It is plausible that firms under extreme economic pressure caused by intense competition could cut expenses that might affect safety. O'Brien (1988) cited the example of Pan American during the late 1960s and into the 1970s (well before deregulation), a period during which the company was under financial pressure. A series of accidents led to an FAA inves- tigation that concluded, among other things, that changes in the company's training program were resulting in improperly prepared pilots. Neverthe- less, for the years preceding deregulation, a period during which some firms experienced financial difficulty, no linkage between the financial strength of carriers and accidents could be established (Golbe 1986). More recent econometric research has examined the relationship of profitability and air carrier accidents between 1957 and 1986, and some correlation has been found (Rose 1990). Rose (1990), however, reported that the evidence from her model on the effect of deregulation is inconclusive. Maintenance Expenditures and Safety Aircraft maintenance is critical to flight safety, and carriers under financial stress could decide to cut corners. Carrier maintenance practices are dif- ficult to compare for a variety of reasons, including aircraft fleets of different ages, differences in record keeping, and inconsistent FAA in- spection data (GAO 1988). Kanafani and Keeler (1988) examined the pre- and postderegulation experience to determine whether the maintenance expenditures and safety records of new entrants differed from those of previously established carriers. Their analysis indicates that new entrants actually spend more on maintenance as a proportion of total costs than the established airlines. They also found that new entrants have slightly fewer accidents per million departures than established carriers. Kanafani and Keeler (1988) hypothesized that the higher maintenance expenditures by new entrants may be attributable to the higher proportion of older aircraft used by the new entrants, which cost more to maintain, but they did not have any direct information with which to test this hypothesis. In a subsequent study, Kanafani et al. (1989) examined the incidence of
178 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION serious service difficulty reports (SDRs), which can be interpreted as indicators of maintenance effectiveness. This analysis did not find a higher rate of serious SDRs among new entrant airlines than among established airlines. If competition were somehow reducing carrier attention to safety and reducing essential maintenance, these factors would be likely to show up in the few accidents that did occur. Oster and Zorn (1987) reexamined the initial accident causes per 1 million scheduled departures from 1970 to 1978 and compared these rates with those from 1979 to 1985. The number of accidents per million departures that involved major carriers and were attributed to equipment failure declined sharply, a statistically significant difference. In other words, tighter operating margins have not resulted in an increase in accidents cause by equipment failures. Accident rates also declined for commuter carriers, and only the decline in equip- ment failures was statistically significant, perhaps because of the tightened federal regulations governing equipment maintenance issued in 1978. New Entrants Versus Established Carriers Although review of the trends in aviation safety does not show any increase in the long-term trend toward reduced accident rates since deregulation, it has been argued that the inexperienced carriers that entered the industry after 1978 kept the trend from improving as much as it might have (Barnett and Higgins 1989). Barnett and Higgins (1989) compared the death risk per flight from 1957 to 1986 for 12 major airlines in operation before deregulation.3 The risk to an individual passenger of being killed in a flight on one of these major airlines fell from roughly 1 in 1 million between 1957 and 1961 to 1 in 10 million between 1982 and 1986. The risk declined almost fivefold for this group during the decade of de- regulation, the largest drop between any time periods. According to Barnett and Higgins (1989), the new-entrant airlines had a substantially higher risk; for 19 carriers operating all-jet fleets that are roughly comparable with those of the previously established airlines, the risk of death from 1979 to 1986 was 1 in 870,000. This risk calculation, however, is derived from only three fatal accidents, which results in an unstable estimate. For one of the fatal accidents, the classification of the new entrants involved in these crashes is ambiguous because the accident involved World Airlines, a former charter carrier with extensive experience in jet operations. Although the higher fatal accident rate of the carriers classified by Barnett and Higgins (1989) as new entrants is cause for
Commercial Aviation Safety 179 concern, no explanation is given in their work for the remarkable gain in safety operation of the other carriers during the same period. If deregu- lation is somehow responsible for allowing new-entrant carriers, who provide about 5 percent of departures, to operate at a higher risk, then deregulation must also be given credit for the almost fivefold reduction in risk of the established carriers, who provide 95 percent of departures. Neither of these conclusions, however, seems plausible, especially in light of the studies examined above, which suggest that the risk definition and data aggregation techniques used by Barnett and Higgins (1989) do not measure the effects of deregulation on safety. Pilot Experience The growth in air travel, partly spurred by deregulation, has increased the demand for pilots and flight crews. The Office of Technology Assessment (OTA) estimates that between 1983 and 1986 the experience and quali- fications of new-hire flight crews declined substantially (OTA 1988). The 1982 crash of an Air Florida jet during a snowstorm immediately after take-off from Washington National Airport has been cited as an example of a crash in which a new-entrant carrier relied too heavily on a pilot who had insufficient experience in severe weather (Gray 1987). Although the captain had about 8,000 hours of flight experience (well over the FAA minimum), most of it had been gained from flying a DC-3 in and around Florida. He had nearly 1,500 hours of experience on the B-737, but little in cold weather. Poor judgment with regard to the deicing of the aircraft was implicated in the crash, in which 78 persons were killed. Because of the complexity of aviation accidents, assigning causality and defining the role of pilot error is an imperfect exercise. The various methods that have been developed and used for assigning primary causality result in considerably different estimates of the role of pilot error. The National Transportation Safety Board (NTSB) conducts detailed assess- ments of commercial aviation accidents and for each accident prepares a report that attributes cause to such factors as pilot error, equipment failure, and ATC. Weather conditions are usually not cited as a cause under the presumption that pilots are trained to fly in inclement weather. Individual researchers analyzing these data often make their own judgments about the primary cause. Pilot error, for example, is frequently cited as the main contributor to accidents; in some studies crew errors are cited to account for 70 percent of crashes in which the aircraft is destroyed (Lautman and Gallimore 1989). NTSB's approach is generally one of listing all the
180 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION factors, without necessarily determining the single most important one. Because pilots are involved in virtually all incidents, pilot error almost inevitably is cited as a contributing cause, and when all the causes are summed, pilot error is the most frequently cited. Another approach assigns cause to the event that precipitates an accident; this approach, used by Oster and Zom (1987) and adopted by the Aviation Safety Commission (1988), tends to give more weight to such factors as severe weather or mechanical failure. Neither approach to assigning cau- sality is perfect. The former approach results in more causes than acci- dents, and efforts to identify the primary cause are inherently subjective. The latter approach may not implicate pilot error when a pilot could have avoided an accident. The difficulty in assigning causality is illustrated by studies in which the role of pilot error and the effects of deregulation on pilot performance are examined. Morrison and Winston (1988) compiled and categorized NTSB accident reports for the 10 years preceding and the 10 years fol- lowing deregulation. They estimated that pilot error is the main contributor to accidents that involve major carriers but found that the absolute number of such errors leading to fatal accidents declined more than 50 percent during this 20-year period. Morrison and Winston (1988) found that pilots for major carriers involved in fatal crashes in the 10-year deregulation period are, on average, slightly older and have more hours of flying experience than pilots involved in accidents in the prederegulation period. Lauber (1988) compared the age and experience of all pilots involved in accidents between 1975 and 1978 with that of all pilots involved in accidents between 1982 and 1985. For major carriers, pilots involved in accidents have fewer hours of total flight experience (a median of 13,000 hours compared with 16,500 hours at the advent of deregulation), but they have more hours of experience in the particular aircraft they were flying at the time of the accident. For commuter pilots involved in accidents, hours of experience for the two periods were very close. It may appear from these studies that aggregate data do not show any significant change in the age (and by implication experience and ability) of pilots involved in accidents, but it is difficult to be conclusive, given the lack of comparative data on pilot age and experience. The key question is how the average age and experience of pilots involved in crashes com- pare with those of their peers. Such data, however, are not available. Lauber's (1988) findings are also limited because his analysis includes all accidents instead of those in which pilot error played a major role. Oster and Zom (1987) found that for the prederegulation airlines, the number of aircraft accidents per million departures in which pilot error
Commercial Aviation Safety 181 played a major part fell significantly from 1979 to 1985 compared with the number from 1970 to 1978. In contrast, the accident rate for new entrants, though not significantly higher overall, was significantly higher than the rate for previously established airlines for accidents in which pilot error played a major role (Oster and Zorn 1987). This finding, however, along with the previously cited findings of Kanafani and Keeler (1988) and Barnett and Higgins (1989) depends in part on the classification of new entrants, which in some instances is a subjective exercise (Levine 1988). Commuter Substitution Although safety has improved in commuter air travel since deregulation, the safety record of commuters is not as good as that of the majors, and communities whose service is changed might experience a higher risk. Oster and Zorn (1987) pointed out that the 20 largest commuter carriers, which transport most of the commuter passengers, have safety records comparable with those of jet carriers. Nonetheless, the smallest commuters have significantly higher accident rates than the top 20. The 20 largest carriers had a passenger fatality rate of 0.67 per million enplanements from 1979 to 1985 compared with 1.21 for the rest of the top 50, and 4.08 for the rest of the commuter industry. Comparisons of safety per- formance within the commuter industry are better made on the basis of enplanements instead of departures because individual commuter carriers are only required to report enplanement data. Oster and Zorn (1987) hypothesized that the relationship between size and safety might be a result of the greater use of small, piston-engine aircraft by the smallest carriers and of less aircraft maintenance experience. The hypothesis of less main- tenance experience deserves additional study because the level of main- tenance required and the amount of governmental oversight of maintenance by commercial carriers depends on whether the carrier is classified under Part 121 or Part 135 of the Code of Federal Regulations. Whether the safety record of small commuters raises a possible effect on safety caused by deregulation, however, is difficult to demonstrate. As Oster and Zorn (1987) have shown, the largest commuters account for the majority of traffic (about 92 percent) and have safety records com- parable with those of the majors. In addition, commuter flights from small communities to hubs involve fewer stops than jet carrier flights, which before deregulation typically involved more intermediate stops at other small communities to take on more passengers. Finally, some of the travel
182 WINDs OP CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION on commuter carriers replaces automobile travel, which, in general, is a much higher risk.5 If one compares the average risk of fatality in commuter aircraft with that of automobile trips, data for the top 50 commuter airlines indicate that safety is improved if travel shifts from automobiles to com- muter aircraft (Oster et al. in press). Communities that are not served by a top-50 commuter or in which jet service was replaced by commuter service may have experienced some increase in risk. In the latter case, however, the difference would not be as large as one might imagine because the differences in risk have narrowed considerably in recent years. It is also important to bear in mind that deregulation may have only accelerated the shift to commuter airlines serving small communities that was already occurring long before 1978, quite independent of deregulation (Eads 1972; MacAvoy and Snow 1977). FAA Performance and Safety ATC Capacity Limits The ATC system has had capacity limits caused by the gradual rebuilding of the controller work force since 1981 and the slower-than-anticipated implementation of the National Airspace System Plan (NAS Plan). The reasons for these capacity limits are discussed in Chapter 7 in greater detail; their possible effects on safety are reviewed in this chapter. The FAA maintains that it has not allowed ATC limits to result in safety risks; when demand exceeds supply, the FAA imposes "flow control," whereby controllers limit the number of airborne aircraft that they can move safely. Empirical evidence supports the FAA's claim that shortages of controllers have not resulted in reduced safety (Oster and Zorn 1987, McKenzie and Shughart 1987). Through a combination of limiting operations by imposing delays, requiring more overtime by ATC personnel, and the dedication and hard work of air traffic controllers, the FAA has maintained a high level of safety. Flow control, however, imposes considerable time penalties. When FAA officials find that demand for ATC is approaching capacity, they begin holding departing aircraft on the ground. The increased demand for air travel and the increased concentration of air traffic at airports has suggested that safety might be compromised by growing ground congestion at airports. The NTSB raised concern about the threat of increased runway incursions in a 1986 report. In
Commercial A viation Safety 183 1988 the Aviation Safety Commission observed that the FAA's own data on the risk of runway accidents indicated a growing problem. The commission urged a review of runway signage, directional indicators, and taxiway and intersection markings. Simply standardizing these run- way features would reduce the risk of pilot error and would do so at a low cost. In December 1990 and February 1991 accidents on runways at Detroit and Los Angeles cost 42 lives. Although the FAA has been working on new technologies and procedures to ensure safety on the ground, these accidents have intensified and focused the agency's ef- forts. The tardiness of the agency's handling of this and other issues, however, raises concern about its future performance. These issues are addressed in Chapter 7. Safety Inspections For budgetary and other reasons, the number of safety inspection personnel actually declined in the early years of deregulation even as the number of air carriers and the need for inspections increased. In 1984 the FAA began an intensive, in-depth, "white glove" inspection effort, known as the National Air Transportation Inspection (NATI) to supplement its ongoing routine inspections. Although most airlines were given a clean bill of health, serious deficiencies in maintenance practices that had been missed in the regular inspections were found at a few airlines, and as a result, the FAA has continued NATI in subsequent years (ASC 1988). The 1984 NATI report indicated that the FAA's safety inspections program was understaffed by 50 percent if the NATI program was to continue (ASC 1988). The FAA's intensified inspections program of the mid-1980s resulted in detailed inspections of major carriers, and in 1987 a record $9.5-million fine was assessed against Eastern Airlines for maintenance violations that occurred between 1985 and 1986 (Aviation Week and Space Technology 1987). Maintenance supervisors and employees of Eastern Airlines were subsequently indicted by a federal grand jury for falsifying maintenance records and allowing improperly maintained aircraft to be used in pas- senger service (Aviation Week and Space Technology 1990b). In March 1991 Eastern acknowledged that some of the violations had occurred. This was a turbulent period for Eastern, during which then-Chairman Frank Borman and the Eastern unions were in a contentious dispute; these labor- management disputes continued, and ultimately intensified, after Eastern was acquired by Texas Air Corporation.
184 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION The maintenance violations at Eastern would appear to indicate that deregulation may have contributed to reduced safety at this airline because of Eastern's attempts to cut corners on maintenance. Deregulation certainly influenced Eastern's efforts to cut costs, but its effects must be understood in the larger context. As illustrated in Chapter 2, Eastern was financially weak before deregulation. In addition, contentious labor-management dis- putes at Eastern predated deregulation, although after the carrier was acquired by Texas Air Corporation, the efforts of Texas Air's Chairman, Frank Lorenzo, to control costs surely exacerbated these disputes. These labor-management disputes, in turn, carried over to the criticisms of East- ern's maintenance practices. In 1986 and 1987 the carrier and its unions were engaged in a highly politicized and publicized debate about the safety of its operations, from which an independent observer would have diffi- culty sorting fact from fiction. Ultimately the FAA did take on this role and engaged in a detailed review of Eastern's maintenance practices. Although no accidents were attributable to Eastern's maintenance viola- tions, evidence that maintenance managers under pressure to dispatch aircraft were falsifying records and impeding the FAA investigation is disquieting. This case, though complex, raises an important question: If the federal government's normal oversight of carrier maintenance was more effective at deterring violations, would Eastern's violations have occurred? The FAA's focus on maintenance records in its routine inspections makes it difficult to detect maintenance violations if such records are systematically falsified. Although these violations were detected by the "white glove" inspections, without the special effort it is not clear whether the maintenance violations would have been detected. The limits of FAA staffing resources more or less force the agency to rely on the carriers to report maintenance practices faithfully. This reliance is not adequate to ensure compliance, however, and is cause for concern. In 1988 the Aviation Safety Commission recommended that the FAA rely more heavily on surprise inspections of carrier maintenance to deter maintenance violations such as the ones that occurred at Eastern (ASC 1988). Apparently the FAA has made some steps toward the safety com- mission's recommendation, but it is not clear whether these steps will deter adequately practices such as those that occurred at Eastern, because carriers still receive some advance notice. Although in recent years the FAA has received support for hiring more safety inspectors from both the Executive Branch and Congress, the FAA also faces a growing problem in the increasing number of aging aircraft that will require a more extensive inspections effort.
11-15 years, 515 aircraft 16-20 yeal 742 aircr 6-10 years, 74 Commercial Aviation Safety 185 )-5 years, 07 aircraft ore than 20 years, 960 aircraft FIGURE 5-6 Age of U.S. commercial jet aircraft in calendar years (3,671 aircraft, average age 12.7 years). Aging Aircraft By definition aging aircraft are aircraft that are being operated near or beyond their originally projected design goals of calendar years, flight cycles, or flight hours. Nothing in the Federal Air Regulations pertains directly to the life of an airplane as measured in calendar years, but it is customary for designers to assume 20 years as the calendar life of an airplane. For the first generation of jet transports, some designers believed that the aircraft would be technically obsolete within 20 years. Using this measure, a substantial portion of the U.S. fleet of large transport aircraft belongs in the category of "aging" (Figure 5-6). The numbers on the periphery of the pie chart denote the age bracket and the number of aircraft in each age bracket shown in the pie (e.g., 960 aircraft are more than 20 years old, whereas only 707 aircraft are less than 5 years old). The other two measures of age are the number of flight cycles and the number of flight hours accrued in service. Table 5-1 shows the design goals for U.S.-designed transports. Industry designers have used the phrase "design goals," but these goals have not represented the end of service life nor are they contractually binding. Commercial aircraft were designed and certificated in the United States after World War II in the belief that with proper inspection, maintenance, and repair, the life of the airframe could be unlimited. The foundation for this premise was the adoption of the principle of "fail-safe design" by the FAA and the industry in the early 1950s. This rule required that a specified level of residual strength must be maintained after "complete failure" or "obvious partial failure" of a "single principal structural element." The early U.S. jet and propjet
186 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINcE DEREGULATION TABLE 5-I DESIGN GOALS OF U.S.- DESIGNED TRANSPORT JET AIRCRAFT Aircraft Life Hours Life Flights DC-8 50,000 25,000 DC-9 30,000 40,000 DC-10 60,000 42,000 L-10I1 60,000 36,000 707 60,000 30,000 727 60,000 60,000 737 45,000 75,000 747 60,000 20,000 757 50,000 50,000 767 50,000 50,000 fleets were designed, tested, and FAA-certified to this rule without a specified life limit. The service experience acquired by this fleet by the mid- to late 1970s had generally shown a satisfactory level of structural safety and provided many documented instances of the validity of the fail- safe concept. However, a few instances in which failure to detect damage in a timely manner resulted in degradation of the residual strength to an unsatisfactory level. These instances and the recognition that economic factors combined with the rate of technology advancement would extend the desired airframe economic life beyond that anticipated at the time of design, testing, and certification prompted manufacturers, operators, and the FAA to agree that the fail-safe rule should be amended to require the use of fracture mechanics in defining the inspections and inspection intervals required for continued airworthiness. This agreement led to the current "damage-tolernce (fail-safe)" rule (14 c.F.d @ 25.571) of October 1978 and to the retroactive application of this rule to then currently certificated jet aircraft such as the DC-10 and the B-747. Supplemental Inspection Documents (SIDs), developed by the manu- facturers, serve as the bases for the continued airworthiness of the first generation of jet transports. Ten years later, the Aloha Airlines crash of 1988 focused public and congressional attention on aging aircraft. This accident and other structural failures stimulated a reexamination of the current approaches to the struc- tural integrity of aging aircraft. Fatigue-initiated damage is the primary cause of concern about aging aircraft. When aircraft are properly inspected and maintained, corrosion and accidental damage should be understood and controlled long before the design life is reached. On the other hand, the problem of fatigue-initiated damage increases with time or, more
Commercial Aviation Safety 187 properly, with use as measured by flight hours or flight cycles or both. Fatigue damage to the fuselage is caused primarily by the repeated ap- plication of the pressure cycle that occurs during every flight. Fatigue damage to the wings is caused by the ground-air-ground cycle that occurs during every flight and by pilot-induced maneuvers and turbulence in the air. Thus, the design-life goal of flight hours is 'more important for the wings than it is for the fuselage. Fatigue-initiated damage is a random phenomenon in which the probability of existence at any specific point in the structure increases with time. Because detecting the damage is also probabilistic, the success of the damage-tolerance process in preserving airworthiness lies in an acceptably low probability of the presence of damage and high probability of timely detection. A large transport airplane is a complex structure in which a large number of points are susceptible to fatigue cracking that could propagate to the point at which the residual strength would be less than the damage tolerance require- ment. The number of points in the structure at which cracks will initiate increases with the age of the structure. As the number of initiation sites increases, the probability increases of not detecting at least one before the strength degrades to the limit. In other words, at some time, the risk of not having limit load capability at some point in the structure may be too great for the airframe to be considered airworthy. Thus, even an airframe designed to be damage tolerant may reach a time in its life when additional inspections, maintenance, and repair will be required to maintain airworthiness. A common scenario for the fuselage includes the following: (a) cracks initiate at the edges of the fastener holes in the center of the panels along a splice; (b) with repeated flight cycles these small cracks link to form a patch spanning across several holes; and (c) the patch becomes sufficiently long so that it is detected during scheduled checks before rapid growth occurs. Rapid growth, if it occurs, will be arrested by "crack turning" at the tear straps or the frame or both, resulting in a safe depressurization that permits the pilot to safely land the airplane. Although this scenario has actually occurred, it is not the only possible scenario. At least two other scenarios can be envisioned in which cracking may not be arrested by the tear straps (e.g., if the fastener holes in the tear straps already contain small cracks or if there is widespread cracking in several adjacent bays). In these circumstances, the crack may continue to propagate rapidly along the splice and lead to uncontrolled depressurization, in which case neither the tear straps nor the frame is fulfilling its original fail-safe function. This, in fact, is what occurred during the Aloha Airlines inci- dentâthe uncontrolled crack propagation resulted in the loss of a large
188 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION portion of the fuselage. The structural condition previously described, wherein widespread cracking in tear straps and adjacent bays occurs is called multiple site damage (MSD). Clearly, MSD is more likely to exist in heavily used aircraft. A key factor in the maintenance of the safety of aging aircraft is the determination of age in flight cycles, flight hours, or both, of the onset of MSD. It is assumed in analyses of this phenomenon that the airframe was designed and manufactured as intended, such that aging is related to the "wear-out" phase of the well-known "bathtub curve." Because neither designs nor manufacturing control is perfect, cracks occur in airframes long before the design goal life is reached. This is not an aging phenom- enon, but should be considered as the population of locations in which the design or construction was deficient (i.e., local "hot spots"). Some are revealed by the airplane fatigue tests and others during the early service experience with the aircraft. As cracks are detected and corrective action is taken, the rate of new hot-spot cracking decreases with time; the net result is that the total population of crack locations at any given time would be small. Experience has shown that the risk of undetected cracks during this phase of the operational life can be controlled to a safe level by appropriate inspection and maintenance programs. In contrast, during the wear-out, or aging, phase of the airframe the frequency of cracking reaches the point at which the risk cannot be con- trolled solely by inspection and maintenance programs. In other words, the airframe has reached the age at which the probability of cracking in large areas of the structure is sufficient to produce an unacceptable risk of not detecting a critical level of damage. Current service with the postâ World War II transport aircraft fleet designed by U.S. manufacturers to meet specified fail-safe requirements and desired design-life goals indi- cates that most of the aircraft have not yet entered this wear-out phase in which they are subject to widespread cracking of structural details of normal fatigue quality. That is, the onset of MSD is beyond the originally projected design-life goals for these aircraft, and therefore MSD is not an issue at this time. The first 291 B-737 fuselages were manufactured using a cold bonding process to adhesively join the fuselage skins at the lap splices. The lap splices were also mechanically fastened together with three rows of rivets. Service experience has shown that some of these bonds have failed, that is, the lap splices have disbonded. A disbonded lap splice has the potential for developing MSD at a relatively early age compared with the original life goal. Under the conservative assumption that the disbond existed at the time of the initial delivery of the aircraft, it has been determined that
Commercial A viation Safety 189 the onset of MSD in these first 291 aircraft may occur at 30,000 flights. Had they not disbonded the splices clearly would not have been the sites for widespread cracking within the projected operational lifetime. The continuing airworthiness of the affected aircraft is being ensured through directives that mandate special inspections and a schedule for the replace- ment of several thousand fasteners in the lap splices with button-head fasteners of a larger diameter. In summary, the primary technical issues posed by the aging aircraft fleet are as follows: MSD, the undesirable condition caused by widespread cracking of the structure, negates fail safety and damage tolerance to discrete sources. Neither multiple-load-path nor crack-arrest fail-safe features of commer- cial transport aircraft can be depended on to protect the structural safety of the aircraft after the onset of MSD. Corrosion, a time-dependent process, decreases the size of structural members, leading to higher stresses and lower structural margins. Cor- rosion also has undesirable synergism with the factors that lead to cracking of the structure, factors that are not well quantified and are not considered in the damage-tolerance and fail-safe design of the structure. Nondestructive inspection is the key to assessing the health of the aging aircraft fleet. It should not be relied on for ensuring the continuing airworthiness of an aircraft that may be approaching the onset of wide- spread cracking, that is, the threshold of MSD as measured in calendar years, flight cycles, or flight hours. Structural repairs, which are more prevalent in the aging aircraft fleet, are generally made to regain static strength and may not adequately fulfill damage tolerance and fail-safety requirements. Terminating actions, the FAA language that denotes the structural actions necessary to eliminate MSD, do not have the data base in terms of component and full-scale testing comparable. with that on which the original structures were certified. Thus, neither the design life of the terminating actions nor the inspection intervals for continuing airworthi- ness can be established without further testing and analysis. The FAA is also addressing the potential aging problems of aircraft used in commuter service. Several factors make this a more difficult administrative task. In 1988 the commuter airline fleet consisted of 1,800 airplanes of 59 types, made by 17 manufacturers, and flown by 165 operators. The
90 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION major airline fleet includes far fewer types of aircraft, manufactured by a handful of companies, and flown by only a few carriers. Many commuter fleets are very small. Most of the aging commuter fleet was designed with 1950s tech- nology. By current standards the structures are neither damage tolerant nor fail safe. Design of these older aircraft was driven more by experience and intuition than analysis. The design staffs of these manufacturers are small and do not have the extensive interaction with the customers operating the aircraft that is characteristic of Boeing, Douglas, and Lockheed. Although risks to safety associated with aging aircraft are worthy of concern, no studies have been conducted that link the financial conditions of the airlines as a result of deregulation to the aging of the aircraft fleet, nor has any direct link been established between aging aircraft and in- creased risk. However, "With a fleet this large and this old," said Ben Cosgrove of Boeing, "we're entering an era that nobody's entered before" (Fortune 1989). The industry has responded to the issue of aging aircraft in two ways. An industrywide task force (Airworthiness Assurance Task Force) rec- ommended making mandatory many recommended maintenance proce- dures for older Boeing and Douglas aircraft and recommended requiring replacement of key components at certain points in an aircraft's service life (Murphey 1989). The FAA has adopted these recommendations. The cost of the task force's recommendations for Boeing aircraft alone is estimated at $800 million over 10 years (Business Week 1989). In addition the industry has placed massive orders for new aircraft. As of April 1989, U.S. airlines had placed orders for 940 additional aircraft and placed options for 1,000 more at a total cost of $95 billion (Airline Economics, Inc. 1989). Although the potential hazards of older aircraft should not be minimized, the life-cycle replacement and maintenance procedures now required by the FAA, if followed by the industry, will mitigate any additional risk associated with catastrophic failure of the airframe. The maintenance pro- cedures to detect corrosion recommended by the Industry Task Force, however, are quite extensive and have substantial cost implications for carriers and the FAA. Carriers are facing cost increases of roughly $1 million per aircraft. The FAA, already trying to expand its inspector work force to meet .existing standards, will be hard-pressed to ensure that the
Commercial Aviation Safety 191 maintenance and replacement schedules are carried out (Aviation Week and Space Technology 1 990a). SAFETY INDICATORS The use of safety indicators instead of total reliance on accidents has been proposed to give a better picture of whether the risk of accidents has increased since deregulation (Lauber 1988). The Aviation Safety Com- mission approached this safety issue by defining it as changes in "the margin of safety," which the commission defined as "that extra cushion built into the aviation system that allows failures (or errors) in mechanical, human, or technological components to occur without an accident nec- essarily resulting" (ASC 1988, Vol. II, p. 25).6 The Aviation Safety Commission, the OTA, and others have examined potential safety indi- cators in considerable depth (Oster and Zorn 1989; OTA 1988; ASC 1988; Golaszewski 1988). In lieu of an attempt to replicate these extensive reviews, their findings on key issues are summarized next. Two significant problems accompany the assessment of risk through the use of safety indicators. First, the overall level of safety in a carrier's operation is affected by many individual elements that cannot be precisely defined or measured quantitatively (Lederer and Enders 1988; GAO 1988). As demonstrated previously, little correlation, if any, can be established between indicators that can be measured relatively easily, such as main- tenance expenditures or overall financial condition, and safety. Second, many of the data used or proposed as indicators have biases in reporting. From its review of potential safety indicators the Aviation Safety Com- mission concluded, "While these data serve as rich sources of information upon which to base management decisions, they are not well suited as [safety] indicators" (ASC 1988, Vol. II, p. 39). Studies in 'hich the potential relationship between incident data and safety is examined have shown little correlation (Oster and Zorn 1989; Golaszewski 1988). After reviewing the same potential indicators and noting their weaknesses, the OTA concluded, "Non-accident data must be used in short-term safety analyses," but agreed that the current data are poorly suited for analysis. Believing that safety indicators are essential, the OTA recommended sub- stantially upgrading the data (and the FAA's management of the data) to make them more useful. Congress subsequently required the FAA to develop a set of safety indicators, an effort that has only recently begun (FAA 1989b, 1990). Despite the weaknesses in the current data, some of
192 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION which are noted in the next section, safety-related data can be used (or improved) to reduce risk in the commercial aviation system. Examples of problems and potential uses will be highlighted. Near-Midair Collisions Reports of near-midair collisions are among the best-known safety indi- cators. The OTA found a high correlation between total traffic and reports of near-midair collisions, but both the OTA and the Aviation Safety Com- mission found them inadequate for tracking changes in risk (OTA 1988; ASC 1988). Two data series are available, but both suffer from reporting bias. The FAA's Near Midair Collision Data Base consists of voluntary reports from pilots. Most near-midair collisionsâabout 75 percentâin- volve general aviation (GA) pilots, but GA pilots usually fail to report them. In addition to reporting bias, the FAA's data series had data man- agement problems through about 1984 and was interrupted by a definitional change in 1985. Total reports in this system reached a high of 1,063 in 1987 and subsequently declined to 549 in 1989 (FAA 1990). Whether the change in the reports reflected any change in risk is not clear because some of the increased reports may have been generated by the mention of the trend in the media. The National Aeronautics and Space Admin- istration operates another reporting system, the Aviation Safety Reporting System (ASRS), which attempts to improve on reporting by making the system confidential. Even this series, however, has fluctuations in re- porting that make trends difficult to interpret. In contrast to the FAA's reporting system, which showed a sharp decrease between 1987 and 1989, the number of reports in this series decreased much less (from 528 in 1987 to 455 in 1989) (data provided by ASRS staff). Although the data series on near-midair collisions appear to be moving closer together in recent years, neither series correlates with the actual number of midair collisions (which almost always involve two GA pilots). Pilot Performance Given the importance of pilots to the safe operation of aircraft, indicators of the margin of safety would presumably include trends in pilot capa- bilities. Pilot capabilities are determined by many influences: carrier se- lection, training, and staffing decisions; FAA licensing standards; and the health and well-being of the pilot. As noted above, however, little infor-
Commercial Aviation Safety 193 mation is available on the experience and qualifications of the current population of pilots. Extensive detail is available on pilots involved in accidents, but little empirical information with which to make comparisons is available on the general population of pilots. Indicators of pilot in-flight performance might be tracked using reports of pilot deviations from required separation standards or deviations into military airspace. Before 1985, controllers reported pilot deviations for inclusion in the FAA's Accident-Incident Reporting system, and reports were largely used to determine whether a violation required some action against the pilot. Many deviations were not reported because the incidents were resolved informally. Since 1985, deviations and the FAA investi- gations of them have been reported to the FAA's Office of Aviation Safety. Total reports of pilot deviations doubled between 1985 and 1987, but it is not known whether the increases are real or simply a result of better reporting over time (OTA 1988, 84; ASC 1988 Vol. 2, pp. 28-29). The number of pilot deviations declined in both 1988 and 1989 (FAA 1990). Flight Data Recorders for Crew Monitoring Carriers operating in other nations have begun collecting more and better information on pilot performance, but they have taken a more practical approach than trying to develop aggregate safety indicators. For example, British Airways, in conjunction with the British regulatory authority, has been using flight data recorders that provide a complete record of crew performance. Not only do these systems avoid reporting biases, they focus directly on correcting crew error, a major contributing factor in airline accidents. Although various on-board monitoring devices have been used in aircraft for many years, the latest technology, digital flight data recorders (DFDRs) can be programmed to monitor many dimensions of aircraft and crew performance. these systems are used by many airlines outside the United States. Air carriers have developed specific procedures for crew performance in the most safety-critical aspects of flightâtake-off, approach, and land- ing because roughly three-fourths of crashes resulting in loss of the aircraft (hull losses) occur during these phases. Carrier operating manuals require pilots to follow procedures during these phases to provide maximum safety. For example, to ensure a stable approach the aircraft should be descending at a specified rate. When it reaches a certain altitude, for example, 500 feet, it should be a specified distance from the runway and flaps should
194 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION be at required settings. DFDR programs provide measures of how closely the crews actually perform to operating standards. Carriers that have been collecting flight operating data for many years, for example, the Scandinavian Airlines System (SAS), have established operating procedures and have collected sufficient data to establish ranges of acceptable performance (Brandt 1990). When crews exceed some pre- established bench mark (usually some percentile of the distribution), the DFDR records an "exceedence." SAS, along with other carriers using these systems, has classified exceedences at different levels and has es- tablished follow-up procedures for each. Minor exceedences that occur during normal operations are simply added to the data base. Major ex- ceedences may require a discussion with the crew, but the confidentiality of the crew is maintained. Nevertheless, the most severe exceedences, even those not resulting in accidents, may trigger formal inquiries. Guarantees of confidentiality and nonpunitive use of the data collected by DFDRs are absolutely essential to gaining acceptance by pilots. In the United Kingdom, the Civil Aviation Authority (CAA), British Airways, and the British Airline Pilots Association have developed an arrangement whereby the identity of the involved crew is known by only one man- agement representative and by a pilot from labor (Leney 1990). The labor representative contacts the crew and reviews the case. The management representative does not divulge the names of the crew to anyone but the labor representative. The CAA has assured flight crews that it will not seek to use DFDR data for enforcement. In addition to being used to inform involved flight crews of potential problems, the data gathered in these programs are used to stress the importance of specific procedures, have led to changes in operating pro- cedures and different training, and have identified problems with specific airspace and airfields. At Air Portugal and Japan Airlines, younger pilots routinely ask for the data in order to assess their performance. Proposals for DFDR programs in the United States have not progressed in the past, partly because of labor-management disagreements, but both sides appear increasingly interested in DFDR programs (Flight Safety Foundation 1990). The Air Line Pilots Association, International (ALPA) appears to be developing a position supportive of DFDRs, but the pilots are concerned that the confidentiality of the data might be breached in legal proceedings. Attorneys already have access to the flight data re- corders used for accident investigations. Access to DFDR data and records of previous flights by pilots involved in accidents, however, might also be sought. ALPA is likely to require federal legislation protecting DFDR data and changes in federal regulations before agreeing to participate in
Co,nmercial A viation Safety 195 a DFDR program. The legislation would be patterned on the law protecting cockpit voice recorders from use in civil legal actions. The federal reg- ulations that refer to DFDRs (Part 121, Sec. 343 and 359) would need to be altered to ensure that the FAA would not use the data for enforcement (Gray 1990). Indicators of Controller Performance The ability of controllers to safely manage traffic has been a major issue of concern since the firing of the strikers of the Professional Air Traffic Controllers Organization (PATCO) in 1981 and the subsequent slow re- building of the air traffic controller work force. The number of fully qualified (full performance level) controllers, although increasing, remains below the prestrike level (GAO 1989). Current controllers handle more operations per year than in any year except the first full year following the PATCO strike (OTA 1988, 102). For examination of controller performance at the primary level, the FAA maintains a data base on operational errors made by controllers. Errors are defined as violations of applicable separation standards be- tween two aircraft in the air or on the ground that are attributable to the ATC system. Total reported errors increased substantially in 1984 (from 723 to 1,888), but some of the increase was due to the automation of the reporting system in that year. Since 1984, the operational error detection program (OEDP), or "snitch patch," automatically records violations of separation standards. As a result of these automated re- ports, controllers began separating traffic by distances greater than required by safe operating standards to ensure that the snitch patch would not catch small errors. The data are used by managers in eval- uating controller performance. Since reaching a peak in 1984, the number of errors and the error rate declined in each subsequent year through 1989, by which time total errors had declined 51 percent and the rate of errors (per 100,000 air traffic operations) had fallen by 58 percent (FAA 1990). It is not known, how- ever, what effect this reduction in reported separation standard errors has had on safety. For example, when two aircraft are headed in the same direction at approximately the same velocity, a minor violation of the separation standard would have no safety consequence. To the extent that controllers are extending separations of this kind merely to avoid minor errors, however, some capacity may be lost without gaining any genuine safety benefit.
196 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION Aircraft Performance and Maintenance Modem aircraft are increasingly sophisticated and reliable. Nevertheless, since 1979 about one-fifth of accidents involving scheduled flights by Part 121 carriers and about one-third of accidents involving scheduled flights by Part 135 (commuter) carriers are attributable to equipment failure (ASC 1988, Vol. 2, pp. 21-23). These equipment failures, in theory, could be tracked by following trends in mechanical reliability reports. The FAA and the aviation industry collect and monitor an extensive amount of data on aircraft component performance. Airline, safety inspector, and manufacturer reports on air- craft failures, defects, and malfunctions are collected in the FAA's Service Difficulty Reporting System (SDRS), which automatically tracks trends in performance. If these trends exceed bench-mark levels, the system alerts the FAA safety analysts. In response to these alerts, the FAA may issue airworthiness directives if it decides that the alert indicates a serious problem. The OTA reviewed the adequacy of the SDRS for analyzing safety trends and concluded that although useful for management purposes, the data have reporting limitations that undercut their utility as a safety in- dicator. According to the OTA (1988, 104), "FAA does not enforce reporting requirements or verify the accuracy of the data." Because some reports are triggered by safety inspections, which are not applied to all carriers in the same systematic fashion, the data are not considered com- plete. Despite the reporting problems, it is possible to use serious SDRs to compare the performance of different airlines, under the presumption that such reports are most likely to be filed (Kanafani et al. 1989). Greater emphasis by the FAA on reporting would further enhance the utility of these data. The reports of the FAA air carrier maintenance inspectors might be analyzed in search of both trend and cross-sectional data, but such data have considerable weaknesses. In previous reviews of FAA inspection data, the GAO concluded that (a) the data were not comparable because the FAA lacked standardized inspection procedures; (b) the data could not be used to gauge broad trends because the FAA did not have a comprehensive data base; and (c) the available information could not be readily classified (GAO 1988). Although the FAA headquarters has made some efforts to respond to the criticisms in the GAO reports, the GAO still finds that the FAA cannot guarantee the quality of its in- spections data because of inadequate and inconsistent reporting by in- spectors (GAO 1990).
Commercial Aviation Safely 197 SUMMARY Commercial aviation safety has improved during deregulation, roughly in line with long-term trends. Especially noteworthy is the complete absence of fatal accidents involving Part 121 air carriers in regularly scheduled passenger service in two years since 1978 and another year in which only one person was killed. These are the only years in modern aviation history when jet carriers have achieved such records in avoiding fatal accidents. Although concerns have been raised about the effects of more extensive competition on carrier costs and attention to safety since deregulation, no study has established an empirical link between carrier operating practices that might have been affected by deregulation and safety. Specifically, there is no empirical evidence linking carrier financial difficulty or cost cutting as a result of deregulation to accidents and no evidence that new entrants spend less on maintenance. The role of pilot experience in crashes is indeterminate with available data. For communities whose air service has changed from jets to turboprop, there may be a small increase in risk, but this increase is now negligible when the effect of improved commuter service is substituted for automobile trips and when the improved safety record of commuter carriers is taken into account. Capacity limits in ATC, due partly to shortages of controllers following the 1981 strike and partly to the slow replacement of ATC technology, have not resulted in reduced safety. The FAA has maintained safe per- formance in the air by imposing delays on departing aircraft. It has had problems keeping pace with the rapidly expanding commercial aviation system, however, as manifested by concern about increased risk of runway accidents and insufficient maintenance inspection staffing and procedures. The underlying causes of these problems are discussed in Chapter 7. The safety of aging aircraft has emerged as a major concern, but it is only indirectly related to deregulation. It appears that the recently adopted FAA regulations address concern about the catastrophic failure of older air- frames but will impose additional costs on the carriers and require addi- tional efforts by FAA maintenance inspection personnel. Reporting biases in safety indicators such as near-midair collisions in- dicate that they cannot be used to evaluate changes in risk since dereg- ulation. Many of the current problems with these data may be remedied with improved reporting procedures and by placing these measures under more scrutiny. Near-midair collision data, for example, have undergone significant changes in reporting and auditing of the reports since FAA began tabulating these data in the early 1980s. It could be expected that other safety indicators might also stabilize once they were subject to
198 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION analysis and reporting. Whether these data will ultimately be useful as safety indicators, however, remains in doubt. Media attention can still cause fluctuations in reports by increasing attention to these data and the relationship of changes in safety indicators to changes in risk has not been established. Carrier use of DFDRs appears to be a more direct and sys- tematic way of monitoring and improving crew performance. The devel- opment of such programs in the United States may require Congress to protect the data to ensure confidentiality and a change in federal regulations to ensure that the data would not be used by the FAA for enforcement. NOTES A small number of commuter carriers elects to operate propeller aircraft under 14 CFR 121 standards, but including data for these carriers does not affect the trend for jet aircraft. The National Transportation Safety Board (NTSB) provided tabulations on acci- dents and fatal accidents that exclude accidents caused by sabotage and suicide, during nonscheduled service, and during all-cargo trips. Departure data reported by the Civil Aeronautics Board (CAB) and FAA were adjusted to remove non- scheduled service and departures by all-cargo carriers. The death risk per flight used by Barnett and Higgins (1989), which they refer to as a "Q' '-statistic, is calculated as the ratio of the proportion of occupants killed to the total number of flights. It can be interpreted as the probability of any single passenger being killed in a crash. This ratio differs from other rates commonly used to analyze trends in airline safety, which tend to use total fatalities or accidents involving a fatality divided by total RPMs or total departures. By using the pro- portion of occupants killed in a crash, the measure is more sensitive than fatal accidents, because many fatal accidents involve the deaths of a small proportion of the passengers. The ratio is also less distorted by single fatal crashes in which 200 to 300 occupants might be killed. Difficulties in classification cloud the comparisons made by Barnett and Higgins (1989). Not all new entrant airlines, for example, can be considered inexperienced (Levine 1988). Some carriers that Barnett and Higgins (1989) classified as new entrants had operated jet aircraft as either charter or intrastate carriers for many years. In addition, some of the established carriers, such as Continental and Braniff, had been reconstituted as low-cost carriers by in many cases replacing unionized labor with new employees with less experience. Evans, Frick, and Schwing (1990) estimated that for trips of up to 600 miles, a middle-aged male driving on a rural Interstate has a lower risk of fatality if he is sober, driving during daylight, and wearing a safety belt. Access to the Interstate highway system from small communities, however, requires a trip of some distance to the Interstate highway system, and not all destinations are on the Interstates. The Aviation Safety Commission, among others, used the term "margin of safety" in a descriptive sense to apply to changes in risk, which, at some threshold, might result in increased accidents. The "margin of safety," however, has a specific
Commercial A viation Safety 199 meaning in structural analysis: Margin of safety = capability - requirement! requirement. Many of FAA's Airworthiness Standards have rules specifying factors of safety. For example, limit loads are those that an aircraft is expected to experience in its lifetime, but the structure must be designed to sustain ultimate loads, which are defined as limit loads multiplied by a factor of safety (usually 1.5). The 50 percent increase in loads would be the "extra cushion" in the parlance of the Safety Commission. Although the general definition of margin of safety and that used in structural analysis appear consonant, this is not the case. For example, a structural component with a strength 10 percent higher than ultimate load has a capability of 1. I ultimate load. If the requirement is 1.0 ultimate load, then this component has a margin of safety of 10 percent. In structural analysis a margin of zero means the design is perfect, in other words it is strong enough without any extra strength, which would mean extra weight. Because of the potential misunderstanding that might result from the use of margin of safety, the study committee preferred to avoid its use in this report. REFERENCES ABBREVIATIONS ASC Aviation Safety Commission FAA Federal Aviation Administration GAO General Accounting Office OTA Office of Technology Assessment Airline Economics, Inc. 1989. The Airline Quarterly, Spring. ASC. 1988. Final Report and Recommendations, Vols. I and II. Washington, D.C., April. Aviation Week and Space Technology. 1990a. Higher Maintenance Standards Sought for World's Aging Fleet. July 2, pp. 60-72. Aviation Week and Space Technology. 1990b. Eastern, Maintenance Heads Indicted by U.S. Grand Jury. July 30, pp. 84-86. Barnett, A., and M. Higgins. 1989. Airline Safety: The Last Decade. Management Science, Vol. 35, No. 1, Jan., pp. 1-21. Barnett, A. 1990. Air Safety: End of the Golden Age? Chance: New Directions for Statistics and Computing. Vol. 3, No. 2, pp. 8-12. Brandt, M. 1990. The Scandinavian Airlines System Flight Analysis and Aircraft Monitoring System. In Proc., Digital Flight Data Recording Workshop, Flight Safety Foundation, Arlington, Va., pp. 38-42. Business Week. 1989. Sure, The Plane is OldâBut Is It Dangerous? March 13, p. 36. Eads, G. 1972. The Local Service Airline Experiment. The Brookings Institution, Washington, D.C. Evans, L., M. Frick, and R. Schwing. 1990. Is It Safer To Fly or To Drive?âA Problem in Risk Communication. RiskAnalysis, Vol. 10, No. 2, pp. 239-246. FAA. 1989a. Report to Congress: Status of the U.S. Stage 2 Commercial Airline Fleet. U.S. Department of Transportation.
200 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION FAA. 1 989b. Aviation Safety Indicators: Concept Definition. Safety Indicators Di- vision, Office of Safety Analysis, U.S. Department of Transportation, Nov. FAA. 1990. Aviation Safety Statistics, Annual Summary, Airspace Incidents, CY 1989. Office of Assistant Administrator for Aviation Safety, U.S. Department of Trans- portation. Flight Safety Foundation. 1990. Proc., Digital Flight Data Recording Workshop, Arlington, Va. Fortune. 1989. How Safe Are You in the Air? May 22, pp. 75-88. GAO. 1988. Aviation Safety: Measuring How Safely Individual Airlines Operate. RCED-88-61. Washington, D.C., March. GAO. 1989. Aviation Safety: Serious Problems Continue to Trouble the Air Traffic Control Work Force. RCED 89-112. Washington, D.C. GAO. 1990. Aviation Safety: FAA Safety inspection Management System Lacks Ad- equate Oversight. RCED 90-36. Washington, D.C. Golaszewski, R. 1988. Air Carrier Safety Performance: Some Empirical Results. Gelman Research Associates, Inc., Jenkintown Station, Pa. Golbe, D. 1986. Safety and Profits in the Airline Industry. Journal of Industrial Economics, Vol. 34, No. 3, pp. 305-318. Gray, R. 1987. Aviation Safety: Fact or Fiction. Technology Review, Aug./Sept. Gray, R. 1990. Legal Issues. In Proc., Digital Flight Data Recording Workshop, Flight Safety Foundation, Arlington, Va. Kanafani, A., and T. Keeler. 1988. New Entrants and Safety: Some Statistical Evi- dence on the Effects of Airline Deregulation. In Proc., Transportation and Dereg- ulation Safety. The Transportation Center, Northwestern University, Evanston, Ill., pp. 237-258. Kanafani, A., T. Keeler, and S. Sathisan. 1989. Analysis of Airline and Aircraft Safety Posture Using Service Difficulty Reports. In Transportation Research Record 1214, TRB, National Research Council, Washington, D.C., pp. 43-51 Lauber, J., 1988. Assessing the Impact of Deregulation: The Benchmarks of Airline Safety. In Proc., Transportation Deregulation and Safety, The Transportation Cen- ter, Northwestern University, Evanston, LII., pp. 223-236. Lautman, L., and P. Gallimore. 1989. Control of Crew-Caused Accidents. Flight Safety Digest. Special Supplement: Crew Performance Monitoring and Training. Flight Safety Foundation, Arlington, Va., Oct., pp. 76-88. Lederer, J., and J. Enders. 1988. Aviation Safety: The Global Conditions and Pros- pects. In Proc., Transportation Deregulation and Safety, The Transportation Center, Northwestern University, Evanston, III., pp. 183-222. Leney, D. 1990. Special Events Search and Master Analysis. In Proc., Digital Flight Data Recording Workshop, Flight Safety Foundation, Arlington, Va. Levine, M. E. 1988. Discussant's Observation on the Evidence of Linkages Between Economic Deregulation and Safety. In Proc., Transportation Deregulation and Safety. The Transportation Center, Northwestern University, Evanston, III., pp. 33 1-334. MacAvoy, P., and J. Snow (eds.). 1977. Regulation of Passenger Fares and Com- petition among the Airlines. Ford Administration Papers on Regulatory Reform. American Enterprise Institute for Public Policy Research, Washington, D.C. McKenzie, R., and W. Shughart. 1987. Deregulation's Impact on Air Safety: Sep- arating Fact from Fiction. Center for the Study of American Business. Washington University, St. Louis. Morrison, S., and C. Winston. 1988. Air Safety, Deregulation, and Public Policy. The Brookings Review, Winter, pp. 10-15.
Commercial A via/ion Safety 201 Murphey, E. 1989. Aging Aircraft: Too Old to Fly? IEEE Spectrum, June, pp. 28-39. Nance, J. 1988. Economic Deregulation's Unintended but Inevitable Impact on Airline Safety. In Proc., Transportation Deregulation and Safety, The Transportation Cen- ter, Northwestern University, Evanston, III., pp. 335-351. O'Brien. J. 1988. Deregulation and Safety: An Airline Pilot's Perspective. in Proc., Transportation Deregulation and Safety, The Transportation Center, Northwestern University, Evanston, Ill., pp. 353-388. Oster, C., Jr., and C. Zorn. 1987. Deregulation's Impact on Airline Safety. Journal of the Transportation Research Forum, Vol. 28, No. I., pp. 1-12. See also Oster and Zorn. 1988. Oster, C., Jr., and C. Zorn. 1988. Airline Deregulation: Is It Still Safe To Fly? in Proc., Transportation Deregulation and Safety, The Transportation Center, North- western University, Evanston, III., pp. 265-294; Oster, C., Jr., and C. Zorn. 1989. Incident Data and Aviation Safety: Midair Collisions. Journal of the Transportation Research Forum, Vol. 29, No. 2, pp. 246-253. Oster, C., Jr., J. Strong, and C. Zorn. (in press). Aviation Safety. Oxford Press. OTA. 1988. Safe Skies for Tomorrow: Aviation Safety in a Competitive Environment. Congress of the United States, July. Rose, N. 1990. Profitability and Product Quality: Economic Determinants of Airline Safety Performance. Journal on Political Economy, Vol. 98, No. 5, pp. 944-964. Weener, E. (n.d.). Key Elements of Accident Avoidance. Boeing Commercial Air- plane Group, Seattle, Wash.
31 Airport and Airway Capacity Limits The performance of commercial airlines since deregulation is inter-related with the capacity of airports and the relevant airspace to accommodate the demand for air travel. Air carriers have attempted to serve the changing and accelerating consumer demand for air travel during deregulation by expanding the number of aircraft, increasing aircraft use and flight frequencies, and realigning route networks to take advantage of the service efficiencies made possible by hub-and-spoke networks. Provision of basic infrastructure by the public sector, however, has not kept pace. At some important points in the system, airport and airway capacity has not been increased sufficiently to meet demand. With the flexibilities given to airline managers to shift assets to better serve con- sumer demand, air carriers have responded to some of these capacity shortfalls by developing underused airports into hubs. Even so, insufficient capacity to meet demand at several points is resulting in increased delays throughout the system. The Federal Aviation Administration (FAA) has long contended that these capacity needs can best be met by building additional runways at these airports, but less expensive options need to be explored as well. Reviewed in this chapter are capacity limits in the existing system, expected increases in demand, and available options for meeting demand within the next 10 to 15 years or so. Strategies for the longer run (through 2050) are addressed in more detail in a 1990 Trans- portation Research Board (TRB) report on long-term airport capacity needs (TRB 1990). Two key points are made in this chapter. The first is that the public sector's response to the increased demand for airway and airport capacity that was stimulated by deregulation has been inadequate. The problems the FAA has had with the National Airspace System Plan (NAS Plan) are 202
Airport and Airway Capacity Limits 203 summarized in this chapter. (The causes for these problems are taken up in Chapter 7). The second main point is that, given the difficulties with expanding the supply of airway and airport capacity, the existing system should be used more efficiently. Assets have been used much more ef- ficiently by the private sector in aviation by greater reliance on the price mechanism. This approach deserves experimentation in the public sector. The application of existing research tools can also help the FAA manage airway capacity better. PAST AND CURRENT DEMAND FOR AIRPORTS Throughout the 1980s the number of air carrier operations (arrivals and departures) grew at an average annual rate of about 2.9 percent, but the shift to the hub-and-spoke service pattern changed the nature of the demand that air carriers placed on many airports and the airways around them. During this period relatively little physical capacity was added to airways or airports, but many air carriers were able to shift operations to underused airports to avoid capacity shortfalls at others. Although substantial over- capacity exists in some parts of the system (i.e., many airports and the airspace around them are underused), demand is pushing up against supply in and around some major airports, causing ripple effects throughout the system. Airport Use Since Deregulation Relatively few airports in the United States are used commercially. Some 3,200 airports are potential recipients of federal aid, but only about 400 have FAA towers to provide air traffic control (ATC) (OTA 1984). These 400 airports, however, account for almost all commercial operations and the majority of total operations. (The operations data used in Table 6-1 and Figure 6-1 are reports of arrivals and departures compiled by the FAA staff in each tower and were provided by the FAA.) Most commercial traffic occurs at a subset of the 400 airports with FAA towers. (Definitions of airport sizes are given in the following discussion.) The large and medium-sized airports in the country (n = 63) account for 82 percent of commercial jet operations and 72 percent of commuter aircraft operations. The overall share of operations and the size of the airports at which they occur are shown in Figure 6-1.
TABLE 6-I TOTAL OPERATIONS AT AIRPORTS THAT HAVE FAA TOWERS BY AIRPORT SIZE AND USER TYPE, 1978 AND 1988 Size° and Type of User Operations 1978 1988 Percent of Total by User Type 1978 1988 Percent Change, 1978-1988 Large (n=27) Jet carrier 5,709,435 7,216,905 57 57 26 Commuter/Air Taxi 1,172,232 2,405,059 31 29 105 GA 2,027,962 1,262,116 7 6 â38 Military 101,796 109,023 8 8 7 Localt 259,362 73,965 1 I â71 Subtotal 9,270,787 11,067,068 19 Medium (n36) Jet carrier 2,059,784 3,174,843 21 42 54 Commuter/Air Taxi 620,428 1,559,907 16 46 151 GA 3,315,020 2,385,540 12 ii â28 Military 173,717 225,248 14 6 30 Local' 1,699,154 764,995 7 5 â55 Subtotal 7,868,103 8,110,533 3 Small (n=76) Jet carrier 1,508,705 1,726,706 15 14 14 Commuter/Air Taxi 544,968 1,576,375 14 19 189 GA 5,248,337 3,838,478 19 17 â27 Military 354,610 366,962 29 26 3 Localb 3,593,077 2,468,170 15 15 â31 Subtotal 11,249,697 9,976,691 â11 Smallest (n = 259) Jet carrier 734,969 633,782 7 5 -14 Commuter/Air Taxi 1,429,147 2,716,836 38 42 90 GA 17,175,773 14,612,439 62 12 â15 Military 527,741 715,320 48 25 25 LocaV' 17,903,783 13,471,746 76 11 â25 Subtotal 37,816,408 32,150,123 â15 Total (n = 398) Jet carrier 10,102,888 12,752,236 100 100 27 Commuter/Air Taxi 3,766,775 8,258,177 100 100 119 GA 27,767,092 22,098,573 100 100 â20 Military 1,202,864 1,416,553 100 100 18 LocaV' 23,455,376 16,778,876 100 100 â28 Grand total 66,204,995 61,304,415 â7 'See text for definitions of airport sizes. bljjcal traffic is almost entirely general aviation (GA)
Airport and Airway Capacity Li,nits 205 35 30 25 20 15 10 Millions of Operations Large Medium Small Smallest Jet Carrier Commuter/Air Taxi Generai Aviation Military FIGURE 6-I Air traffic at airports that have FAA towers, 1988 (data from FAA). Between 1978 and 1988, considerable shifts occurred in the use of commercial airports. Although commercial traffic increased substantially, general aviation (GA) declined such that total operations declined 7 percent (Table 6-1). The 27 largest airports (defined in this report as those that averaged more than 275 commercial jet operations per day in 1978) ac- counted for 57 percent of total jet operations during the decade and ab- sorbed a 26 percent increase in such traffic between 1978 and 1988. The number of commuter and air taxi operations, slightly less than a third of operations in both 1978 and 1988, doubled during the 10-year period. GA use of large airports, in contrast, declined sharply (the local traffic listed in Table 6-1 is almost entirely GA). The number of GA operations at all airports declined more than 20 percent between 1978 and 1988; at large airports GA operations declined more than twice as much. The amount of commercial traffic at medium-sized airports grew more overall than the national trend in commercial traffic. Medium-sized air- ports are defined in this report as those with more than 100 and fewer than 275 daily jet operations in 1978. Jet operations increased by more than 50 percent, and commuter and air taxi operations increased by 151 percent between 1978 and 1988. These increases reflect the growth of secondary hubs as carrier managers responded to capacity shortages at major airports and took advantage of underused airports at which they could build hubs and gain market share. (The role of airport lease agree- ments and other issues affecting airport capacity are discussed in Chapter
206 WINDs OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION 4.) The amount of GA traffic declined at medium-sized airports as well but was more in line with national trends. Small airports, those that had fewer than 100 but more than 30 daily jet operations in 1978, received more commercial traffic during the decade, but most of the growth was in commuter service (Table 6-1). The number of jet operations increased 14 percent, but the number of commuter op- erations increased 189 percent. The number of GA operations at small airports declined, more or less in line with national trends. The smallest airports, those that had fewer than 30 daily commercial jet operations in 1978, had fewer jet operations by 1988, but the number of commuter and air taxi operations almost doubled. The amount of com- mercial traffic (jet, commuter, and air taxi combined) increased by about 50 percent at those airports. The downward trend in the amount of GA traffic during the 10-year period was similar to the national trend. Declines in GA The substantial decline in the number of GA operations and the shift in GA traffic away from large and medium-sized airports relieved some of the demand on those airports and the airspace around them. The gain in capacity, however, tends to be less than proportional to the decline in GA traffic because at many airports GA users are required to use short runways with approach and departure paths separate from those used by jets. Many reasons are given for the drop in GA traffic. New-aircraft prices more than doubled between 1978 and 1988, which apparently dampened demand (FAA 1990a). Shipments of single-engine piston aircraft, more than 14,000 in 1978, dropped to several hundred in 1989. Operating costs increased dramatically after the 1979 fuel embargo and, even after leveling off in the mid-1980s, were still twice as high in 1988 as in 1978. The number of GA pilots, a reasonably good barometer of demand, declined from 327,400 in 1978 to 300,900 in 1988, and the number of student pilots, a barometer of future demand, has declined by about 4.5 percent annually since 1980 (FAA 1989a). Although overall GA traffic has been declining, some segments of GA have been increasing. Use of turboprop and turbojet GA aircraft, which is largely for business, increased from 3.4 million hours flown in 1980 to 3.9 million hours flown in 1988 (FAA 1990a). This segment of com- mercial GA, however, only accounts for about one-fourth of all com- mercial GA traffic. The majority (72 percent) of business GA hours are flown in piston aircraft (FAA 1987). This segment has decreased sub- stantially since 1978, such that overall business GA hours flown decreased
Airport and Airway Capacity Limits 207 Total Number 80 1 234567891011121 23456789101112 Hours of the Day - 1980 -i-- 1988 FIGURE 6-2 Raleigh-Durham average daily operations, August 1980 and 1988 (data from OAG data tapes). 25 percent between 1978 and 1987 (estimates provided by National Busi- ness Aircraft Association). Although the number of turboprop and turbojet GA hours flown is increasing, it represents only about 10 percent of all GA hours flown, which is relatively modest. This traffic, however, tends to occur at large and medium-sized airports, partly to make possible connections with commercial airline service and partly to provide access to related services, such as car rental. Hubbing and Demand The shift to hub-and-spoke service patterns fundamentally changed the nature of demand at airports that carriers developed into hubs by increasing peak demands at these transfer points, but capacity problems are not limited to hubs. Peak demands at some major origin and destination (O&D) airports are also approaching capacity limits. For the hub-and-spoke system to work with a minimum of passenger time lost during connections, numerous flights must arrive and depart in a relatively short period of time. This causes a substantial peak demand on airport facilities at the times that passengers most prefer to fly (typically in the morning and early evening). Average daily scheduled operations (departures and arrivals) by hour for August 1980 and 1988, as estimated from computer tapes of the Official Airline Guide, are shown in Figures 6-2-6-5. In Figure 6-2, the increased peaking characteristic of airports developed into hubs since deregulation is shown by the operations at
208 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION Total Number 250 200 150 100 50 04 23456789101112123456789101112 Hour8 of the Day - 1980 -6- 1988 FIGURE 6-3 Atlanta Hartsfield average daily operations, August 1980 and 1988 (data from OAG data tapes). Raleigh-Durham airport. Before deregulation, the airport mostly served O&D traffic into the Raleigh-Durham area. American Airlines has since developed it into a hub, and the tightly spaced incoming and outgoing connecting flights create the sharp peaks in the morning, early afternoon, and early evening hours. In the off-peak hours, total traffic is little different from that in 1980. Similar patterns exist for airports in Charlotte, Cin- cinnati, and Dayton. In contrast to the sharply peaked demand at airports developed into hubs, airports that were already heavily used hubs, such as Atlanta Harts- field and Chicago O'Hare, show relatively little additional peaking because use of these airports was already approaching runway capacity during peak periods (Figure 6-3). For airports that primarily serve O&D traffic and were already ap- proaching capacity before deregulation, such as Boston Logan Airport, peak-hour use has not increased (Figure 6-4). Boston now has morning and midday peaks, however, that approximate the early evening peak. Although most airports with peak-period capacity limits are constrained by runway capacity, airspace capacity limits are equally important in some metropolitan areas. The increased peak-hour demand for airspace for the three major airports in the New York City area (Kennedy, LaGuardia, and Newark) is shown in Figure 6-5. Of the three airports, Kennedy and LaGuardia are primarily O&D airports, and they have experienced rela- tively uniform increases in demand. The increased peak demand in the morning and evening for the area is largely driven by the increased peaking
Airport and Airway Capacity Limits 209 Total Number 140 120 100 80 60 40 20 0 â¢â¢â¢â¢I....___ -.â¢7â¢â¢.... ' I I I I 1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12 Hours of the Day -1980 __11_ 1988 FIGURE 6-4 Boston Logan average daily operations, August 1980 and 1988 (data from OAG data files). at Newark. From 1980 to 1988 overall operations increased at these three airports by 28 percent, but the peak-hour demand in the evening increased by 55 percent; by 1988 the ATC system had to accommodate 100 additional jets in the peak hour. The adoption of hub-and-spoke networks is occasionally singled out as the cause for capacity shortfalls in the aviation system, but this attribution Total Number 350 300 250 200 150 100 50 0 1I I I I I 123456789101112123456789101112 Hours of the Day -1980 1988 FIGURE 6-5 Major New York City area airports' (EWR, JFK, LGA) average daily operations, August 1980 and 1988 (data from OAG data tapes).
210 WINDs OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION is too simplistic. Hub systems allow carriers to more closely serve con- sumer demand, both in terms of price and frequency, than do the linear route systems they replaced. The same level of service with a linear route system would not be feasible at a reasonable cost, but even if it were, it would worsen the capacity problems because more, smaller aircraft would be required to meet this level of demand. The increased number of aircraft operations would place more demand on O&D airports and increase the complexity of ATC as well. Hubbing does increase peak demands at hub airports, but as outlined in the following section, most of the worst capacity problems occur at destination airports and at hub airports that were at capacity before deregulation. Delay as a Measure of Capacity Airport runway capacity is not a fixed value. It varies with, among other things, changes in weather and visibility, the combinations of runways in use, the mix of aircraft using the airport, the timing of arrivals and de- partures, and ATC rules, procedures and safety precautions. Estimates of delay provide a proxy for the extent to which demand is approaching capacity. In the short term, as demand for airports and airways exceeds supply, queues develop, and passengers experience delay, which is ex- acerbated by increased peak demand at hubs. Delay, however, results from numerous causes, many of which cannot be controlled (e.g., weather). Thus, no airport can be completely free of delay. At low levels of demand, delays tend to be brief and are not of great concern. As congestion builds, however, delay escalates. As queues grow, the frequency and average length of delay increases exponentially, as shown in Figure 6-6. Alter- native measures of delay are provided from two systems maintained by the FAA and one maintained by the Office of the Secretary of the U.S. Department of Transportation (DOT). These sources and available trends are discussed in turn. The FAA's Air Traffic Operations Management System (ATOMS) Air traffic controllers report delay statistics to this system for operations that are delayed more than 15 minutes. The ATOMS data report total delayed flights instead of the duration of delay per flight. The delay measured in this system is based not on the schedule, but on the difference
Airport and Airway Capacity Limits 211 DEMAND FIGURE 6-6 Conceptual relationship of ca- pacity, demand, and delay. between optimal and actual travel time. The controller also makes judg- ments about the cause of the delay. Although the absolute number of delayed flights measured in this system is not assumed to be preciseâ controllers are unlikely to worry about collecting data during hectic pe- riodsâthe trends in the causes of delay are assumed to be fairly reliable. The data, available in a consistent fashion since 1984, indicate that weather is the principal cause of delay, followed by delays attributed to the volume of traffic being handled by the ATC center, and then by the traffic being handled by the airport terminal (Table 6-2). Of course, the delays in poor weather are caused by a combination of traffic and TABLE 6-2 DISTRIBUTION OF DELAY GREATER THAN 15 MINUTES BY CAUSE (FAA I989b, I990b) Cause Percent by Year 1984 1985 1986 1987 1988 1989 Weather 60 68 67 67 70 58 Center volume 16 Il 10 13 12 6 Terminal volume 18 12 16 II 9 30 Runway construction 3 6 3 4 5 5 NAS equipment 2 2 3 4 3 2 Other I I I I 1 1
212 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION ATL BOS ORD CLE DFW DEN DTW FLL IAH MCI LAS LAX MIA MSP LGA JFK EWR PIT PHL STI SF0 DCA - 1988 1989 20 40 60 80 100 120 140 FIGURE 6-7 Delays per 1,000 operations at 22 major airports, 1988-1989 (see Appendix C for airport codes) (FAA 1989c). weather. With more slack in the demand, the effects of weather would not be so great. If the delays attributed to center volume and terminal volume are combined, it appears that roughly one-third of total delays result from the inability of either airports or ATC to handle total demand at peak periods. Although the absolute number of delays appears to be relatively constant between 1984 and 1989, the data from 1984 are believed by the FAA to be distorted by start-up problems. With 1985 as an alternative base for comparison, the number of total delays increased between 1985 and 1989. Whether the upward trend is continuing, however, is debatable. Delays in 1989, though about 17 percent higher than delays in 1988, are still 6 percent below the total number of delays experienced in 1986 (FAA 1989c). Whereas total delays appear to have increased between 1988 and 1989, the worst delay problems, according to the ATOMS data, are concentrated at a few major airports, as shown in Figure 6-7. Although several airports have occasional delay problems, a disproportionate share of the problem occurs in a few places: Atlanta (ATL), Boston (BOS), Chicago (ORD), DallasâFort Worth (DFW), Denver (DEN), the New York area airports (LGA, EWR, and JFK), Philadelphia (PHL), St. Louis (STL), and San Francisco (SF0). Each of these airports had more than 20 delays per 1,000 operations in 1989, and those delays have a disproportionate effect on passengers. Those 11 airports account for 36 percent of all commercial enpianements annually and are major nodes in the network. Of these
Airport and Airway Capacity Limits 213 airports, seven are hubs and four are primarily O&D airports. Five (ORD, LGA, EWR, JFK, and SF0) have more than 50 delays per 1,000 oper- ations and account for about 17 percent of all enplanements nationally (FAA 1989b). The FAA 's Standardized Delay Reporting System (SDRS) SDRS is considerably limited compared with the ATOMS data because it is based on reports from only three major carriers (American, Eastern, and United). The advantages of this data series are that it extends back to 1976, gives information about specific airports and at what point the delay occurred (taxi in, airborne, taxi out, or gate hold), and estimates the duration of delay per flight. As with the ATOMS data, delays are measured against optimal instead of scheduled times. According to esti- mates developed by SDRS, average delays increased 38 percent between 1979 and 1987, from roughly 5.8 to 7.8 minutes per operation (FAA 1988). As with the ATOMS data, the SDRS estimates indicate that average delays are the worst at the New York City airports and at Chicago, but in contrast with the ATOMS data, the SDRS estimate shows average delays that are almost as bad at Boston and DallasâFort Worth (Table 6- 3). Because the SDRS data are not representative of all carriers using these airports, however, the results are likely to be biased. In addition, the Eastern strike during 1989 and the subsequent demise of Eastern in 1991 substantially reduced the quality of the trend data and subsequent years because Eastern was one of only three carriers reporting. The DOT's Air Travel Consumer Report Since 1987 the DOT has required the 13 largest air carriers to report their on-time performance to and from 29 airports (airports with at least one percent of domestic scheduled service enplanements). "On time" in this report means within 15 minutes of the schedule as given in the carrier's computer reservation systems. The data are published monthly. In its monthly report, the DOT ranks the on-time performance of each airline as well as the on-time performance of the airlines at each airport. Although these data are useful to consumers insofar as they encourage carriers to publish realistic schedules, they do not provide an accurate gauge of delays because carriers have built many of these delays into their schedules.
TABLE 6-3 DELAYS AT SELECTED AIRPORTS REPORTED TO SDRS, 1976-1988 Location Identifiction City Total Operations (thousands) 1976 1988 Percent Change Average Delay (minutes per operation) 1976 1985 1988 ABQ Albuquerque 227 230 1.3 6.7 7.8 ATL Atlanta 490 779 59.0 8.7 8.5 9.0 AUS Austin 165 189 14.5 3.9 4.1 BDL Windsor Locks 140 180 28.6 5.2 7.5 BNA Nashville 215 271 26.0 4.8 8.1 BOS Boston 307 438 42.7 6.4 8.3 9.2 BWL Baltimore 233 308 32.2 4.2 4.6 6.8 CHS Charleston 129 134 3.9 3.7 4.6 5.1 CLE Cleveland 235 254 8.1 4.4 5.1 5.1 CVG Covington 148 272 83.8 2.9 3.6 6.6 DAY Dayton 168 213 26.8 5.3 8.5 DCA Washington, D.C. 326 328 0.6 6.2 6.6 8.3 DEN Denver 419 502 19.8 6.4 8.7 8.4 DFW Dallas-Fort Worth 360 675 87.5 5.1 8.7 9.4 DTW Detroit 247 375 51.8 4.0 7.2 6.2 EWR Newark 193 380 96.9 7.5 10.7 11.0 HNL Honolulu 321 377 17.4 4.6 6.5 6.4 lAD Washington, D.C. 188 231 22.9 5.2 5.2 8.6 IAH Houston 208 295 41.8 4.1 5.0 5.7 IND Indianapolis 215 220 2.3 3.6 4.4 5.8 JAX Jacksonville 128 156 21.9 3.7 4.1 4,7 JFK New York 332 330 -0.6 10.5 9.4 11.0 LAS Las Vegas 300 371 23.7 4.8 5.5 LAX Los Angeles 483 624 29.2 4.7 7,3 7.8 LGA New York 345 365 5.8 9.2 9.8 11.0 MCI Kansas City, Mo. 179 232 29.6 6.4 6.8 MCO Orlando 94 295 213.8 4.8 7.4 MEM Memphis 310 353 13.9 3.3 5.1 4.6 MIA Miami 301 399 32.6 5.2 5.9 6.9 MKE Milwaukee 229 193 - 15.7 4.9 6.7 MSP Minneapolis 252 380 50.8 2.7 5.5 6.4 MSY New Orleans 156 148 -5.1 3.0 3.5 4.4 OAK Oakland 399 400 0.3 5.8 4.6 ONT Ontario, Calif. 156 141 -9.6 6.3 6.4 ORD Chicago 718 803 11.8 9.0 9.0 10.8 PBI West Palm Beach 220 231 5.0 5.3 6.2 PDX Portland 217 272 25.3 4.6 5.2 PHL Philadelphia 311 416 33.8 6.8 5.5 8.2 PHX Phoenix 425 464 9.2 3.4 5.7 6.9 PIT Pittsburgh 310 390 25.8 5.4 6.0 6.2 RDU Raleigh-Durham 198 275 38.9 3.5 4.7 8.6 SAN San Diego 207 206 -0.5 5.7 6.4 SAT San Antonio 195 198 1.5 4.5 4.6 (continued on next page)
Airport and Airway Capacity Limits 215 TABLE 6-3 (continued) Total Average Delay Operations (minutes per Location (thousands) Percent operation) Identifiction City 1976 1988 Change 1976 1985 1988 SEA Seattle 174 316 81.6 3.7 4.6 5.5 SF0 San Francisco 343 455 32.7 5.3 8.5 8.2 SiC San Jose 470 347 â26.2 7.0 5.7 SLC Salt Lake City 255 290 13.7 6.1 6.5 SMF Sacramento 140 183 30.7 3.7 4.4 STL St. Louis 321 433 34.9 4.7 6.9 7.6 TPA Tampa 192 244 27.1 3.7 4.1 5.1 NOTE: Data provided by the FAA. Airport identification codes are listed in Appendix C Summary Because the data series measure different aspects of delay, they yield somewhat different results; nonetheless, the FAA data tend to show that the largest delay problems occur at a few airports. Most of these airports, however, are critical nodes in the network, and three of them (LGA, JFK, and ORD) were already under slot-control limits because of capacity problems before deregulation. Of the 11 airports with more than 20 delays per 1,000 operations, 7 are hubs and 4 are primarily O&D airports. The delay data do not distinguish between airport and ATC capacity limits, but they do indicate the areas in which problems occur. It appears that roughly one-third of delays result from peak demands that exceed the supply of ATC and runways. Most delay, roughly 60 percent, occurs in poor weather, but the effects of weather interact with the heavy traffic. With more slack in the demand, the effects of weather would not be so great. PROJECTED INCREASES IN DEMAND On the basis of the federal government's estimates of demand and supply, delays can only be expected to worsen at many airports in the years ahead. The FAA projects that the number of operations (arrivals and departures) by major commercial domestic carriers will increase by one-third between 1988 and 2000 (FAA 1989a, 1990a). This substantial increase results from projecting an average annual growth rate of 2.5 percent, which corresponds with projected growth in the economy and is slightly slower than the growth experienced during the last decade (Table 6-4). Air taxi and corn-
216 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION TABLE 6-4 HISTORIC AND PROJECTED AIRCRAFT OPERATIONS (FAA 1989a) Operations by Year Average Annual (millions) Growth (%) Activity 1980 1988 2000 Historic Projected Air carrier operations 10.1 12.7 17.1 2.9 2.5 Air taxi and commuter 4.6 8.3 12.6 7.6 3.5 GA 49.0 37.4 47.7 (3.4) 2.1 Military 2.5 2.8 2.8 1.4 0.0 muter operations, which have been growing more than twice as fast as other forms of commercial traffic, are expected to double. GA is projected to reverse its decline and rise to nearly the peak levels reached in 1980 by the year 2000. At present, 21 airports experience more than 20,000 hours of annual flight delays. The FAA estimates that even if planned improvements to primary airports are made, 33 primary airports will ex- ceed 20,000 hours of delay in 1997 (see text box). Although these predictions imply substantial increases in demand and potentially large increases in delay, forecasts of aviation activity are in- herently difficult to make. The business cycle directly influences demand, and fuel price instability directly affects operating costs and indirectly affects pricing and demand. The timing of such influences is unpredictable. In addition, the accelerated phaseout of Stage 2 aircraft to reduce airport noise could cause a quicker shift to larger aircraft with more seating capacity than the FAA currently assumes. A faster-than-anticipated shift to larger aircraft would reduce demand on airports. Although no long-run forecast is certain, the FAA's forecasts of com- mercial operations for 2000 appear to be within reason, given the 20-year trend (Figure 6-8). The FAA's long-range forecasts of commercial op- erations have tended to be somewhat low, but GA projections have tended to be too high. The forecasts for 1989, for example, made in 1978 to 1980, estimated commercial revenue passenger miles (RPMs) and en- planements within a range of error of about 10 percent, which is fairly good, given the uncertainties of future commercial activity in 1979 (Table 6-5) (FAA 1978-1980, 1990a). Forecasts of total commercial operations at airports with FAA towers are quite close (within 2 to 3 percent). Although the commercial aviation forecasts made in the late 1970s were fairly reasonable, forecasts for GA were in error by a wide margin. The FAA's 1978-1980 estimates were made during a peak of GA activity and were about 50 percent greater than what actually occurred (Table 6-5).
AIRPORTS EXPECTED TO EXCEED 20,000 HOURS OF ANNUAL DELAYS IN 1997 BY REGION Area Airport Northeast Boston Logan New York City: Kennedy, LaGuardia, Newark Pittsburgh Philadelphia South Washington: National, Dulles Charlotte Nashville Memphis Atlanta Orlando Miami DallasâFort Worth Houston: Hobby and Houston Intercontinental Midwest Cleveland Columbus Cincinnati Detroit Chicago O'Hare St. Louis Minneapolis West Salt Lake City Las Vegas Phoenix Seattle-Tacoma San Francisco San Jose Los Angeles: Los Angeles International and Ontario Honolulu
218 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION Million8 35 30 25 20 15 10 Actual Projected '1 1970 1975 1980 1985 1990 1995 2000 Year - Air Carrier --- Air Taxi/Commuter - General Aviation FIGURE 6-8 Trends in aircraft operations and FAA forecasts for the year 2000 (excluding local operations). The current forecasts for GA also appear optimistic in light of the structural changes in GA and the flat trend in GA operations during the last several years. Even though commercial GA is growing in terms of turbojet op- erations, this segment makes up a small share of total activity and its growth rate cannot drive the projections shown in Figure 6-8. From these projections, it appears that most of future demand will occur in the commercial segment of the industry at commercial airports. Some increased business GA will also occur at commercial airports but will account for a small share of total operations. LIMITED SUPPLY EXPANSION As major airports and the airways around them become increasingly sat- urated, how and where can the increased demand forecast by the FAA occur? The answer to this question depends on how different options are pursued in coming years by the FAA, local governments, and air carriers. Major expansions in physical capacityâbuilding new airports or expan- sion of most existing congested onesâdo not appear likely within the next decade or so. Although the federal government distributes funds for airport development, the FAA has little direct control over decisions to expand airport capacity, which are made at the local level. In addition, federal aid for airport development makes up a relatively small share of total investment (CBO 1988). Expansion of major airports is unlikely in
TABLE 6-5 FAA AVIATION FORECASTS COMPARED WITH ACTUAL ACTIVITY (FAA 1978-1980, 1990a) Itinerant Operations Commercial Air Taxi! Enplanements RPMs GA Hours Flown Air Carrier Commuter GA No. Error No. Error No. Error No. Error No. Error No. Error Year (millions) (%) (billions) (%) (millions) (%) (millions) (%) (millions) (%) (millions) (%) Forecast 1978 446 â7.7 376 â15.1 65 66.2 12.7 â1.6 8.1 â2.5 42.5 48.0 1979 523 8.1 425 â2.1 60 63.3 12.7 â1.6 8.2 â1.2 41.8 47.1 1980 506 5.1 407 â6.6 59 62.4 12.1 3.2 8.4 1.2 41.8 47.4 Actual - (1989) 480 434 22 12.5 -- 8.3 22.1
220 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION some cases because of limited space, but in most cases because of com- munity opposition to noise. Given the difficulty of expanding supply at congested airports, more emphasis will be needed on finding ways to use existing airports more efficiently and encouraging use of currently under- used airports with excess capacity. Few New Airports Likely No new major national commercial airport has been developed since 1974, but local governments at Denver and Austin are developing new airports. Denver's, in particular, will offer substantial additional capacity during the 1990s. New airports are also being considered in Los Angeles and in Farmington, New Mexico, but are unlikely to be operational by 2000 (TRB 1990). Five other cities (Atlanta, Chicago, MinneapolisâSt. Paul, San Diego, and St. Louis) are exploring the possibility of purchasing land for future airport sites, but the TRB study committee that examined long- range airport capacity needs concluded that the development of additional new airports near these metropolitan areas is unlikely because of high costs, scarce land, and community opposition to noise (TRB 1990). Few New Runways at Existing Airports The FAA has long believed that more capacity gains are possible from building or expanding runways than from any other single measure, and additional runway capacity is being planned for 50 of the top 100 airports (FAA 1989b). Runway expansions, along with revisions to approach rules (discussed later), would increase capacity during poor weather at a number of other major airports by essentially enabling them to regularly exceed or operate close to their best instrument flight rule (IFR) capacity. Cur- rently congested airports, however, such as those in New York, Boston, Washington, D.C. (National), and San Francisco, do not have plans for additional runway capacity. Some of the delay may be eased by more capacity at destination airports, which would result in fewer aircraft being held at these congested airports awaiting opportunities to depart for their destinations. On the basis of planned expansions of runways, major air- ports with minor to moderate delay problems are addressing some of their capacity limits. The FAA senior officials, however, have expressed skep- ticism about whether many of these planned expansions will take place
Airport and Airway Capacity Limits 221 because of community opposition to additional noise (Del Balzo 1990). Even if they do, the airports with the worst delay problems are the least likely to add runway capacity during the next decade. BETTER USE OF EXISTING CAPACITY Airport and ATC capacity are closely related. At present, the ATC system can barely manage present demand during good weather when visual flight rule (VFR) conditions prevail. Preliminary runs of simulation models developed for the FAA of the national demand for airspace indicate that, even with clear weather nationwide, some schedule delays will occur. Thus, at present, when the weather is not good delays mount rapidly. Several options reviewed in this section indicate some additional capacity gains through better use of existing facilities: Modifications of ATC procedures, Replacement of outmoded ATC technology, Systems approach to demand management, and Marketplace strategies to encourage greater efficiency. Modifications of ATC Procedures As noted previously, about two-thirds of current delays in the system are attributable to the combined effect of inclement weather and total traffic. During poor weather, delays mount rapidly because IFR conditions require greater spacing of aircraft than do VFRs, which apply in clear weather. Large aircraft arriving at an airport during peak periods, for example, typically have following distances of 2 to 7 nautical miles during VFR conditions and 4 nautical miles during IFR conditions (OTA 1982, 104). Instrument conditions also affect the capacity of runways. Under VFR, two runways that are less than 2,500 feet apart might handle 125 operations an hour. Under IFR, however, the capacity can drop by roughly half because runways this close together are not permitted to handle operations simultaneously (OTA 1982, 105). Philadelphia's airport, used as a hub by USAir, has a problem similar to this; in poor weather its peak capacity is essentially halved. For larger airports with more runways, the effects of poor weather on runways are less dramatic because more combinations of runways are possible. Nonetheless, capacity is reduced.
222 WINDS OF CHANGE: DOMESTiC AIR TRANSPORT SINCE DEREGULATION One alternative to adding runway capacity is to modify the ATC rules governing approach procedures during inclement weather. The benefits of such modifications would be quite substantial and could drastically reduce delay during poor weather. The potential problem with such mod- ifications is an increased risk of a midair collision if simultaneously ar- riving aircraft abort their landings. The FAA is introducing new radars with more rapid response times and is developing procedures for aborted landings to minimize the risk of midair collisions. Changes in IFR ap- proach rules would increase peak capacity during poor weather by up to 40 percent at congested airports such as Boston and San Francisco (FAA 1989b). Allowing triple approaches during inclement weather would have substantial benefits for large airports with considerable delay problems, such as Chicago O'Hare, DallasâFort Worth, and Atlanta. Considerable analysis and careful demonstration projects will be required to ensure that ATC procedures can be modified without compromising safety (Kanafani 1986). Some ATC modifications depend on the development of new technology and new procedures to reduce the gaps in arriving aircraft. The FAA estimates that reducing trailing distances, runway occupancy times, and departure separations could increase the efficiency of single runways by 20 percent during instrument conditions (FAA 1989b, 8-2). Achieving these goals, however, depends on research to develop better (real-time) information on wake turbulence, aircraft braking systems, and techniques for reducing uncertainties in navigation and control. (More accurate radars are part of the FAA's technology development and acquisition program, discussed next.) Replacement of Outmoded ATC Technology Since 1981 the FAA's primary response to the traffic growth accelerated by deregulation has been a massive overhaul and replacement of the ATC system. The FAA's program, referred to as the National Airspace System Plan (NAS Plan), was originally promulgated as a 10-year, $12-billion program to replace outdated technologyâliterally vacuum-tube compo- nents in some casesâand to provide radars, computers, and controller work stations that are more reliable and less costly to maintain. Also included are advanced aircraft sensors and automated weather infuriiiation systems. The plan has evolved in recent years as the FAA has recognized the need for additional components to respond to the demands for more capacity and is currently a multiyear $27-billion capital improvement
Airport and Airway Capacity Limits 223 program (FAA 1989d). This program is one of the largest and most complicated (nonmilitary) technology acquisition programs ever under- taken by a federal agency. Almost all of the original NAS Plan components are now under contract, although most have been delayed 2 to 3 years for a variety of reasons. Major components will be phased in during a 10- to 15-year period beginning in the 1990s. Most of the individual components of the FAA's capital improvement program are designed to enhance reliability, safety, and employee pro- ductivity, and not to add capacity. Some of the radar acquisitions, the completed replacement of computers at the en route centers, and improved weather reporting systems are examples of safety and reliability enhance- ments. Other aspects of the program will provide for additional capacity as well as safety and reliability, for example, the Advanced Automation System (AAS) and the Microwave Landing System (MLS). The AAS, a $4.4-billion program, is the largest single element of the FAA's capital improvement program. The five phases of the AAS will improve computer hardware and software and controller work stations at all ATC facilities. All work stations will be completely revamped to provide more computer capability, monitors able to display more infor- mation, and greater back-up capability. The hardware will be supple- mented with software that will provide controllers with automated decision aids and automated conflict resolution to reduce controller workload and increase efficiency (Celio 1990). The greatest future constraint on airspace capacity will not be radar or work station capacity; instead, capacity will be defined by the number of aircraft a controller or team of controllers can manage at one time. Auto- mated procedures for ATC, which have been developed and are being refined, can increase capacity by overcoming the limits of an individual controller (Celio 1990). The FAA's initial NAS Plan projected that the AAS, which is essential for the application of these automated procedures, would be completed by 1994. Current projections estimate initial imple- mentation in 1994 and complete implementation by 2000 (GAO 1990). Further delays may occur; during 1990, when this report was being pre- pared, the FAA and the prime vendor were negotiating how to handle unanticipated technical problems that did not have apparent resolutions. The MLS, a $1.1-billion system, also promises capacity gains; it is designed to be more flexible than current instrument landing technology and could provide for a wider variety of approaches, including curved approaches, that would reduce some of the delays caused by current approach procedures. This capability could improve airport runway ac- ceptance rates, particularly during inclement weather. Simulation of the
224 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION traffic for the New York area, for example, indicates that in inclement weather, curved approaches could allow from 3 to 20 additional landings per hour (depending on weather conditions) (Aviation Week and Space Technology 1990). Acquisition of the MLS has already begun, but tests of the first systems did not meet FAA specifications, and the FAA sub- sequently cancelled the original contract. Final implementation of the MLS was projected for 1992, but problems in the initial testing phase indicate further delay. Meanwhile, new satellite technologies have been developed that many airline industry officials believe would work better. Although the NAS Plan has been delayed, the FAA projects that major enhancements will begin coming on line in the next few years. These systems, along with additional technologies, such as precision runway monitors (Quick Scan Radars), could produce 100 percent more IFR ca- pacity at selected airports, according to the FAA staff. However, given the past difficulties and delays the FAA has had in procuring and delivering. new technology, these technologies may not be available to ease conges- tion problems for some time. Systems Approach to ATC The growing saturation of the airways is posing a more complex ATC problem than the FAA has previously had to face. Throughout its history, the FAA has handled ATC at most airports on a first come, first served basis. The exception occurs at slot-controlled airports, but even at these airports controllers handle traffic as it arrives. Thus, in the past the man- agement of airspace demand has been handled by individual controllers. The FAA is forced increasingly by traffic saturation to manage demand throughout the system. For example, the total amount of aircraft from around the nation headed to the three New York area airports must con- verge at a single navigation point to be lined up for the airway into the New York area. This requires the FAA to begin spacing the distance between traffic as far away as Colorado and Atlanta to ensure that the navigation point does not become saturated. Similar problems must be managed for traffic approaching the San Francisco Bay area airports. Experience with the East Coast Plan illustrates the difficulty that the FAA faces. By adding navigation points for departing traffic from the New York airports, the FAA was able to allow more traffic to leave these airports during a shorter period of time (they no longer had to be lined up over a single navigation point). This alleviated departure delays and reduced the queue of traffic trying to land at Kennedy. However, the
Airport and Airway Capacity Limits 225 increased rate of departures from New York increased delays at Phila- delphia because one of the departure paths intersects the Philadelphia airspace, and the north-south traffic along the East Coast (except traffic to and from New York) was delayed because flights had to be routed on a wider path around the New York area. The FAA also underestimated the amount of political opposition to the program that would result from exposing many more residents of New Jersey to relatively low levels of aircraft noise (GAO 1988). The FAA can no longer rely on demand management at individual airports, but must begin developing a systemwide approach. Doing so will require a major reorientation of the FAA's approach to ATC. The sim- ulation models required to analyze the system effects on capacity resulting from changes in procedures are only now being developed. Capacity models can help the FAA analyze potential solutions to ATC problems by simulating the consequences of such changes. More intensive use of existing models and further improvements in them would cost the FAA less than $10 million annually; the benefits of such an investment, if applied to FAA management and planning, would far outweigh the cost. It would allow the FAA to define its air traffic operations on an analytical, systems performance basis, which would give the agency more systematic techniques for planning and evaluating changes in procedures and the potential benefits of new technology. The new ATC role for the FAA is also likely to have an effect on future capital improvement needs. The NAS Plan technologies were designed to upgrade the FAA's ability to perform its traditional approach to ATC. The airspace saturation effects now being encountered at major airports were not anticipated. NAS technologies can be adapted, but this could further delay the delivery of these systems and increase their costs. Marketplace Strategies Congestion Pricing for Runways Congestion at many airports occurs only during peak periods (typically in the morning and early evening). Few airports are congested throughout the day, but some, such as New York Kennedy and Chicago O'Hare, experience a high level of sustained demand throughout the day. In general, however, the cost of using congested airport runways or ATC does not vary with demand, resulting in excess peak-hour use. Most runway landing fees are
226 WINDS OF CHANCE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION calculated based on the weight of the aircraft, because the weight of the aircraft approximates the cost of design and construction imposed by users. These fees, however, do not reflect the opportunity cost of using the airport during periods of peak demand. More than 20 years ago Levine (1969) pointed out that the underpricing of a scarce resource like runway capacity will result in its overuse and that airport congestion would be reduced by pricing runway use to reflect its value to the user. Kanafani and Ghorbrial (1985) modeled the consequences of increasing the fee for using a hub airport on the basis of a sample of city pairs connecting at Atlanta. Their analysis indicates that as the landing fee increases, some direct service is substituted for connecting service, but the benefits of hubbing for most flights continues to outweigh the increased fee, even when it is increased from $50 to $1,800 per landing. Although some schedule frequency is lost, delays decline considerably, and the operational savings for carriers offsets the increased fee. Passengers lose a small amount of schedule convenience, but delay is reduced by as much as half. In addition, the airport would gain considerable revenue. Morrison (1987) and Morrison and Winston (1989) modeled airport runway pricing from an economic welfare perspective and demonstrated considerable ef- ficiency gain as well as revenue gain for the airport from the adoption of marginal cost pricing. Boston Logan Airport recently experimented with higher fees (not true congestion fees) but was forced to return to its previous fee system because the DOT found that the fees discriminated against GA. The pricing scheme, referred to as the Program for Airport Capacity Efficiency (PACE), was not peak-period pricing, but a general increase in the flat fee and an additional increment to the fee based on aircraft weight. GA activity at Logan, in contrast with other major airports, actually increased during the mid-1980s; in 1987 alone, GA flights increased 30 percent over the year before. The Massachusetts Port Authority (Massport) estimated that in 1987, GA flights accounted for 10 percent of total op- erations and carried 1 percent of passengers (Massport 1987). Most of the flights originated less than 150 miles from Logan and were from airports that also had scheduled commercial service. Like most airports, Logan's landing fee had been developed to recover the cost of building the runways and was based on the weight of the aircraft. Under the weight-based system the lightest GA aircraft would pay about $10 per landing and a commuter aircraft would pay about $23 per landing. In 1980 the minimum fee at Massport was raised to $50, but this was discounted to $20 for GA aircraft in off-peak hours. More recently Massport officials argued that because the runways were largely paid for,
Airport and Airway Capacity Limits 227 the fee should be allocated on the basis of use of the airport facilities, with an incremental additional cost for weight (Massport 1987). When costs were allocated in this manner, the minimum landing fee for a GA aircraft rose to approximately $100 and the minimum fee for a large jet dropped from $800 to about $450. The Massport analysis of the effect of this fee increase indicated that about 5 percent of operations would be affected, and they were assumed to be entirely GA. Although commuter aircraft would also experience a substantial increase in landing fees, the incremental cost per passenger was only $2 to $3; hence it was assumed that this cost would be passed on without affecting commuter operations (Massport 1987). During the period that PACE was in effect (July to December 1988) operations by smaller aircraft declined by about 34 percent, according to Massport officials. Turboprop and small jet operations, both air taxi and private, appeared to be relatively unaffected, but piston aircraft operations, both air taxi and private, fell sharply (correspondence with Massport officials). Most of the latter were private piston aircraft. There was some evidence that air taxi operators reduced service frequency and increased aircraft size. Use of the airport by lighter aircraft during the summer months declined by two to three operations during each peak hour. GA representatives sued Massport and were joined by the DOT, which argued that the pricing scheme was developed with an eye toward pricing GA out of the market instead of being based on an equitable allocation of costs. Although the appeals court supported the DOT in the legal battle, the court's ruling essentially was that it was not in a position to second- guess the DOT in such matters. The DOT, however, appears to be re- ceptive to a fairer cost allocation and pricing system (The Public's Capital 1989). U.S. Transportation Secretary Samuel Skinner's National Trans- portation Policy endorses greater reliance on congestion pricing strategies (DOT 1990). Thus it appears that the door is still open for pricing runway use, and Logan's experience suggests that there are benefits to be had. GA users are likely to continue to oppose such strategies, but the business segment of GA users is not opposed to peak-hour pricing (though it opposes any strategy aimed specifically at shifting classes of users to relieve air- ports) (Howe 1990). impediments to Congestion Pricing As indicated by Massport's attempt to raise runway landing fees, adoption of runway congestion pricing is impeded by both legal and practical ob-
228 WINDs OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION stacles. Federal aid to airports has long been conditioned on an "open access" basis that prohibits airports from discriminating against classes of users. GA supporters used this argument in their opposition to the increased fees at Boston Logan and were supported by the DOT in arguing this position. As noted above, however, the DOT was open to congestion pricing that was based on a fair allocation of costs to users of different kinds. Even a more fair proposal for reallocating costs may be opposed by GA users, however, because they could be required to pay considerably more for the use of airports than they now pay. Some airports are financed by airline leases that include provisions that prohibit the airport from charging any additional fee or changing the manner in which fees are charged. Most of the 30 largest airports have lease clauses that include one or both of these provisions (DOT 1990b). Thus, short of negotiating to change these leases when they expire, some airports would be unable to implement runway pricing without specific legislation enabling them to do so (for example, as was done with pas- senger facility charges, or head taxes).' Other practical impediments must be addressed as well, although in most cases some kind of solution appears achievable. Some airports, including slot-controlled airports, operate at near peak demand throughout the day, which indicates that off-peak capacity might not be available. In this instance congestion pricing might shift some users to alternate, under- used airports or encourage carriers to use larger aircraft. In the latter case more passenger throughput would be achieved without increasing the number of flights. Although some proposals envision fees that would encourage users to shift to off-peak hours, additional operations at many airports during off-peak hours in the early morning or at night would be resisted because of noise. Indeed, shifting aircraft to other airports might engender additional neighborhood resistance and therefore be unsuccess- ful. Whereas the difficulties that neighborhood opposition to noise on the efficient use of airports should not be minimized, increased use of alternate airports around Los Angeles and New York is occurring. Increased congestion fees might also affect the connections for travelers whose trips include segments on small, commuter aircraft and larger jets. A simulation of the impact of marginal cost pricing at one airport, for example, indicates that, among classes of commercial users, commuter aircraft would be the most disadvantaged by a shift to pure marginal cost pricing because small light-weight aircraft receive the largest implicit subsidy in current runway landing fees (Morrison 1987). The cost of landing during the peak for the commuter aircraft would have a higher per passenger cost than for a large jet even though the large jet may have
Airport and Airway Capacity Limits 229 a much larger fee than the commuter. The effect the fee would have on passenger preferences would depend in large part on the size of the fee. The passengers making connections from a commuter to a national or major airline are usually flying on joint fares (a single price for the com- muter segment and the major carrier segment of the journey). A peak charge that increased these joint fares by $10 (assuming a $300 increased landing fee for a plane with 30 passengers) may have only a marginal effect on passenger choices; substantially higher fees, however, would have a larger effect. If the effect of congestion pricing was deemed too disruptive on commuter flights originating from small communities, sep- arate fees could be established with an implicit subsidy to ensure adequate and timely service from small communities. One potential problem with runway pricing by individual airports is that they have a monopoly position over the airlines and other users. Some of the most congested airports have little room (or ability, for environmental and political reasons) to expand but might use pricing strategies to extract additional revenues from the airport users or, under pressure from community groups, to limit operations. Analyses by Kan- afani and Ghorbrial (1985) and by Morrison and Winston (1989) indicate that congestion pricing, though designed to shift demand from the peak, would nonetheless be a major source of revenue for airportsâso much so that the airports could find such a strategy attractive independent of whether the funds were reinvested in capacity expansion. In this respect congestion pricing, instead of serving as a market signal for the need for new capacity, could be used as a device to ration capacity and earn revenue. A considerable problem is how to define the optimal fee, which will necessarily involve a certain amount of experimentation. One straight- forward approach to solving this problem is to increase the cost of operating in the peak hour until the congestion is eliminated. In this approach, the cost rises to a level that the user with the least value for operating in the peak is unwilling to pay. For this operator the choices are to shift to a less congested hour or another airport, or to not operate. Commercial jet flights that depend on connections at hub airports are unlikely to be affected unless peak hour charges were extremely high, but some might shift. In addition, many major airports continue to have a small percentage of operations by private and recreational users who are likely to be among the first to shift their operations to avoid a peak- hour fee. Although the number of operations that would shift may be small, given the shape of the delay curve, such shifts would have a more than proportional effect on delay. The proposal that is outlined
230 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION next builds on this approach and incorporates features to deal with some of the issues raised previously. One Approach to Congestion Pricing The first step in such an approach would be to define the operational capacity of the airport through an engineering study or based on historic use (a process most airports are already engaged in for other reasons). For any airport at which demand during the peak hour exceeds the agreed- upon capacity, a federal congestion charge could be imposed (which would not disturb existing airport charges and landing fees). The charge would be increased until the congestion was eliminated. The charge would only be imposed for any hour in which demand exceeded supply. In order to ensure that revenues from congestion pricing were used to increase capacity, a condition for imposing congestion pricing might in- clude specific requirements for how the funds would be spent. This might be handled in several ways. One proposal that would help create incentives to increase capacity systemwide would be to place all the funds raised by this congestion charge into a central fund. All airport operators that im- posed a congestion charge could apply to the fund for capacity expansion. The use of the funds for capacity would not be limited to runway expansion; the fund would be broadly conceived to include expenditures on new radars or other technology that might increase throughput capacity. It would also include expenditures for sound-proofing and other investments that might minimize community opposition to capacity increases. By having some authority award the funds to airports with the most cost effective proposals for expanding authority, the funds would be in- vested in the most efficient manner. The central fund would be an off- budget entity. The potential problem with any federal mechanism, of course, is whether the expenditures of funds would be restricted by Con- gress as is often argued to be the case with the transportation trust funds. This problem could be handled by establishing a revolving fund or en- terprise fund with a permanent appropriation. Both kinds of funds are within the power of the congressional appropriations committees to control in any year, but they can be structured to minimize the amount of control exercised in the annual appropriations cycle. By imposing a price mechanism on the users that add to congestion, the charge would promote more efficient use of airports. By charging a price more closely associated with the true value of peak-hour use, the proposal would raise substantial revenues for capacity expansion. This
Airport and Airway Capacity Limits 231 example of a congestion pricing proposal indicates that it is possible to imagine solutions to many of the potential impediments to congestion pricing. Whether this or other proposals for congestion pricing would actually work in practice deserves to be tested. Airports should be allowed and encouraged to experiment in alternative schemes until some workable solution is found. Congestion Pricing for ATC Marginal cost pricing could also be applied to the provision of ATC facilities (CBO 1988). Relatively little research has been conducted on congestion pricing for ATC, but Golaszewski (1987) estimated the unit cost of providing these services. The estimates indicate that optimal pricing for ATC would have little effect on the aggregate cost to commercial carriers because such costs are roughly the same as the revenues generated from current aviation taxes. More optimal pricing of ATC use, however, would shift the timing of demand and may relieve some congestion. GA users, if charged for ATC use, would have to pay considerably higher fees. The fuel tax charged GA users is estimated by Golaszewski (1987) to represent less than 20 percent of the cost of the ATC services used by this group. Charging for ATC use at congested airports could be expected to shift GA away from major airports. An important issue pointed out by Golaszewski (1987) that must be considered in pricing for ATC use is whether safety would be compromised because of GA pilots shifting their operations to airports without FAA towers. Incentives To Shft to Underused Airports Satellite Airports To Relieve O&D Traffic For each of the airports that the FAA forecasts to be congested by 1997 there is another airport (sometimes two or more) within 50 miles that is currently underused and could absorb some future growth (FAA 1989b). These "satellite" airports have adequate runways for commercial jet ser- vice but may require additional investments in terminal facilities if sub- stantially more service is scheduled. For example, Islip and Newburgh, New York, are both within 50 miles of the three major New York City airports and have the capacity to absorb about 165,000 operations per
232 WINDs OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION year, which represents about a 10 percent expansion of the combined capacity at Newark, LaGuardia, and Kennedy. (Islip has fairly restrictive noise limitations, however.) Five airports are within a 50-mile radius of Los Angeles International, although only one, Palmdale, has significant underused capacity. Part of the incentive of shifting to these airports is to reduce delay both in the air and on the ground. Travelers headed to major metropolitan areas could save ground travel time by flying into airports closer to their actual destination and away from the most congested ground traffic in urbanized areas. Some satellite airports may not be used as they might because the use of congested airports is underpriced; charging the true marginal value of peak use of congested airports would encourage users who place rel- atively little value on using the airport during the peak to either shift to the off-peak or use nearby satellite airports. Despite community opposition to noise, which is a major impediment to all types of airport expansion, use of some satellite airports is increasing. Major carriers are already providing service to the satellite airports around Los Angeles International, and American Airlines has begun offering service to Stewart International Airport, a former Air Force base 60 miles north of New York City (Labich 1990). New Hubs The FAA has identified 28 airports around the country that are more than 50 miles away from major airports that have current or forecast delay problems (FAA 1989b). These airports have the capacity to absorb more operations and may have sufficient O&D traffic to make a hub airport economical. Carrier success in developing airports at medium-sized cities such as Charlotte and Dayton implies that other potential hubs will be developed to help meet the growing demand for commercial aviation. Even at these potential new hub airports, community opposition to in- creased noise may be a significant impediment, but opposition might be offset by an equally powerful local interest in the perceived economic benefits of hub airports. This has certainly been the case at the medium hubs already developed. Reliever Airports Reliever airports can include the types of traffic discussed previously, but GA is the focus for reliever airport development. Federal policy
Airport and Airway Capacity Limits 233 for many years has supported funding for reliever airports in the hopes of separating GA and commercial traffic, both for efficiency and safety. Most of the funding for GA airports (75 to 80 percent) has come from federal aid, and many such airports now exist (CBO 1988). Every airport that the FAA projects to be congested by 1997 has at least one reliever airport, and many have several. Federal subsidies to GA reliever airports in the past, however, have apparently had little success in attracting GA away from major airports (FAA 1989b; CBO 1988). Charging for the use of congested airports would help accomplish this policy goal. Use of Larger Aircraft Kanafani and Ghorbrial's (1985) analysis of hub congestion indicates that higher fees imposed on congested hubs would encourage more direct flights in markets that can sustain such traffic. Large city pairs may be served directly (bypassing congested hubs) with future deliveries of aircraft with long-range capability and competitive economics, such as Boeing 737-300's, MD-82's, and Airbus 320's. The majority of additional future seat capacity for domestic use will come from larger aircraft. (Although many narrow-bodied aircraft are on order, they will mostly replace the existing narrow-body fleet while the wide-body fleet increases) (FAA 1990; Boeing Commercial Airplanes 1989). Larger aircraft may also be scheduled into congested hubs. The latter option has the benefit of increasing passenger throughput at the hub. Because of the extra seating capacity, more passengers can be served without adding flights and congestion. Future development of small jets, which are fuel efficient in longer-range use, is being actively explored by manufacturers and may also make hub overflight strategies attractive for city pairs unable to support direct service by large jets. Although these adjustments to capacity constraints by the private sector are al- ready occurring, they could be hastened by pricing the use of congested facilities to better reflect their value. SUMMARY The accelerated growth in demand for air travel during deregulation has begun to exceed capacity limits at some airports and the airspace around them. Although these shortages in capacity occur at relatively few places, they occur in and around the major airports that serve a large share of total demand. The prospect for several new airports around major met-
234 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION ropolitan areas to alleviate this shortage does not appear promising in the near term. Some efforts are under way to expand capacity. A few major cities are exploring the development of new airports, and many existing airports are planning to add or expand runways. However, community opposition to noise, scarce land, and high development costs suggest that many of these plans will not be realized. Some technical measures can provide more capacity within the existing supply of airports. It appears that some modifications of ATC operating procedures during inclement weather would accommodate more growth at many major airports, although some of these modifications await tech- nology development and additional research to ensure that they will not compromise safety. In addition, future deliveries of NAS Plan technologies and the supporting software would allow for greater efficiencies in ATC and the use of existing runways. Even so, the FAA would need to develop its ability to provide ATC from a system's perspective, instead of focusing on individual bottlenecks. As the airways become increasingly saturated, changes in procedures at one point have the unintended consequences of moving the bottleneck (or delays) to some other part of the system. The FAA should develop and use simulation models to analyze the consequence of such changes in advance. More time and funding may also be needed to adapt NAS Plan technologies to this new role for the FAA. Even with these technical enhancements, the current level of delay at congested major airports may not be much reduced. Indeed, delay at most major airports is expected to increase. Demand management strategies such as peak-period pricing for runways have only had a limited trial in the United States, but the benefits were positive. Pricing to relieve peak-hour congestion would also hasten the shift to underused airports, some of which could be developed as hubs to absorb the demand for connecting traffic. Given the flexibility displayed by car- riers since deregulation in developing new hub airports as capacity limits at existing airports are approached and competitive advantage is defined, this additional hubbing would be a useful way of helping the nation absorb expected additional demand for the next decade or so. Other existing airports within 50 miles of currently congested airports can be relied on more heavily by both commercial and GA users to serve O&D traffic. Near every airport that the FAA projects to be congested in the future there is an existing airport, in some cases several existing airports, that could absorb some future demand. These adjustments would likely take place more or less by themselves as carriers and GA users seek alternatives to congestion, but the process would be hastened and delays reduced by
Airport and Airway Capacity Limits 235 pricing for peak-hour use of congested airports. One of the potential disadvantages of congestion pricing is that the revenues earned may not be reinvested to expand capacity. This disadvantage might be countered by creating an institution, a central authority of some kind, that would collect peak-hour fees and return them to airports willing to invest the proceeds in new capacity. NOTE To the extent that congestion fees would have to be based on opportunity costs rather than simply fully allocated actual costs, there may be a federal legal issue. One interpretation of the Supreme Court's decision in the case of Evansville Van- derburgh Airport Authority District etal. versus Delta Airlines etal. is that airports may not charge aeronautical users more than the airport's historic costs for providing capacity. Airlines disadvantaged by a congestion fee, or unwilling to pay it, could argue that fees based on opportunity costs would generate revenue in excess of the costs of providing the airport facilities. On the other hand, it could be argued that the fees, if reinvested in capacity, were charged appropriately to the users that require the added capacity. REFERENCES ABBREVIATIONS CBO Congressional Budget Office DOT U.S. Department of Transportation FAA Federal Aviation Administration GAO General Accounting Office Massport Massachusetts Port Authority OTA Office of Technology Assessment TRB Transportation Research Board Aviation Week and Space Technology. 1990. MLS Simulations Show Curved Ap- proaches Can Boost Traffic Flow. Aug. 20, p. 94 Boeing Commercial Airplanes. 1989. Current Market Outlook: World Travel Mar- ket, Demand and Airplane Supply Requirements. Seattle, Wash., Feb. Celio, J. 1990. Controller Perspective of AERA 2. The MITRE Corporation. Mc- Lean, Va. CBO. 1988. New Directions for the Nation's Public Works. Washington, D.C., Sept. Del Balzo, J. 1990. Reauthorizing Programs of the Federal Aviation Administration (Future Airport Capacity Needs and Proposals to Meet Those Needs). Hearings before Subcommittee on Aviation, Committee on Public Works and Administration, U.S. House of Representatives. Report 101-37. Feb., p. 82. DOT. I 990a. Moving America: New Directions, New Opportunities. Feb.
236 WINDS or CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION DOT. 1990b. Secretary's Task Force on Competition in the U.S. Domestic Air- line Industry, Airports, Air Traffic Control, and Related Concerns (Impact on Entry). Feb. FAA. 1978. FAA Aviation Forecasts, FY 1979-1990. U.S. Department of Transportation. FAA. 1979. FAA Aviation Forecasts, FY 1980-1991. U.S. Department of Transportation. FAA. 1980. FAA Aviation Forecasts, FY 1981-1992. U.S. Department of Transportation. FAA. 1987. FAA Statistical Handbook of Aviation Calendar Year 1987. U.S. De- partment of Transportation. FAA. 1988. Airline Delay: 1976-1986, Based on the Standardized Delay Reporting System. Report FAA-APO-88-13 (updates to 1987 provided by FAA staff). U.S. Department of Transportation. FAA. 1989a. FAA Aviation Forecasts Fiscal Years 1989-2000. Report FAA-APO- 89-1. U.S. Department of Transportation. FAA. 1989b. Airport Capacity Enhancement Plan. Report DOT-TSC-FAA-89- I. U.S. Department of Transportation. FAA. 1 989c. Air Traffic Activity and Delays Report for December 1989. Report DOT- TSC-FAA-89-I. NAS Analysis Branch, ATO-130, U.S. Department of Transpor- tation, Dec. FAA. I 989d. National Airspace System Plan: Facilities, Equipment, Associated De- velopment and Other Capital Needs. U.S. Department of Transportation, Sept. FAA. 1990a. FAA Aviation Forecasts Fiscal Years 1990-2001. Report FAA-APO- 90-1. U.S. Department of Transportation. FAA. 1990b. Air Traffic Activity and Delays Report. U.S. Department of Transpor- tation, Dec. GAO. 1988. Aircraft Noise: Implementation of FM's Expanded East Coast Plan. RCED-88- 143. Washington, D.C., Sept. GAO. 1990. Air Traffic Control: Continued Delays Anticipated for the Advanced Automation System. GAO/IMTEC-90-63. Washington, D.C., July. Golaszewski, R. 1987. The Unit Costs of FAA Air Traffic Control Services. Journal of the Transportation Research Forum, Vol. 28, No. 1, pp. 13-20. Howe, J. 1990. The Airport Access Issue. NBM Airports Handbook, National Busi- ness Aircraft Association, Washington, D.C. Kanafani, A. 1986. The Analysis of Hazards and the Hazards of Analysis: Reflections on Air Traffic Safety Management. Accident Analysis and Prevention, Vol. 18, No. 5, pp. 403-416. Kanafani, A., and A. Ghorbrial. 1985: Airline Hubbing-Some Implications for Airport Economics. Transportation Research, Vol. 19A, No. 1, pp. 15-27. Labich. K. 1990. Airport 2000-A Horror Story? Fortune, Vol. 121, No. 14, June 18. Levine, M.E. 1969. Landing Fees and the Airport Congestion Problem. The Journal of Law and Economics, Vol. 12, pp. 79-108. Massport. 1987. Massport's Program for Airport Capacity Efficiency. Dec. Morrison, S. 1987. The Equity and Efficiency of Runway Pricing. Journal of Public Economics, Vol. 34, pp. 61-85. Morrison, S., and C. Winston. 1989. Enhancing the Performance of the Deregulated Air Transportation System. Brookings Papers: Microeconomics 1989. The Brook- ings Institution, Washington, D.C.
Airport and Airway Capacity Limits 237 OTA. 1982. Airport and Air Traffic Control System. Congress of the United States, Jan. OTA. 1984. Airport System Development. Congress of the United States, Aug. The Public's Capital. 1989. A Rough Landing: Boston's Airport Experiment Shows Limits of Pricing Policies. Vol. 1, No. 2. Oct. TRB. 1990. Special Report 226: Airport System Capacity: Strategic Choices. Na- tional Research Council, Washington, D.C., 134 pp.
7 Constraints on the Performance of the Federal Aviation Administration From the discussion in Chapters 5 and 6 it is apparent that some safety and capacity issues are affected by the performance of the Federal Aviation Administration (FAA). The political and organizational reasons for these problems are outlined in this chapter. The FAA was not changed by deregulation, but throughout the 1980s the FAA's abilities to cope with a transformed industry were limited by a variety of constraints, and as a result it fell behind on safety inspections, air traffic control (ATC) services, and replacement of ATC technology. Through the mid-1980s, limits on personnel, attributable to efforts to control the burgeoning federal budget deficit, contributed to a declining FAA safety inspection work force even as the demands on it increased. Because of civil service limits on pay differentials, the agency has been unable to attract sufficient numbers of experienced air traffic controllers to high-cost areas such as New York City, and continues to have problems recruiting top-notch technical staff in all of its work forces, particularly in high-cost areas. As the decade ends, the FAA projects further delays in the massive 10-year technology acquisition effort on which it embarked in 1981, partly to help it overcome staffing problems. Cumbersome governmental procurement regulations and ineffective management have both been blamed for the continued delay and escalation in cost. Finally, during the 1980s, the agency found itself being second-guessed increasingly on technical and administrative matters by the Office of the Secretary, the Office of Management and Budget (0MB), and Congress. 238
Constraints on the Performance of the FAA239 During the mid-1980s some critics became skeptical that the FAA, operating under constraints such as the ones mentioned above, which are typical of federal agencies, would be able to respond quickly enough to the demands being placed on it by a rapidly growing and evolving aviation system. The Air Transport Association of America commissioned a study of the FAA's ATC service by the National Academy of Public Admin- istration (NAPA); this study urged Congress to reorganize the FAA as a government corporation to allow it to discharge its responsibilities more effectively (NAPA 1986). The 1988 report of the Aviation Safety Com- mission also recommended creating an independent governmental au- thority (ASC 1988a). Other proposals have gone even further, recommending the privatization of the ATC system. In a study for Congress, the Office of Technology Assessment (OTA) suggested more cautious reforms. In contrast with the NAPA and the Aviation Safety Commission studies, the OTA did not recommend changing the FAA into a governmental enter- prise, but instead recommended that streamlining the FAA, clarifying its mission, and increasing its funding would address needed reforms (OTA 1988). (The studies and options for reform are reviewed in Appendix B.) Although legislation has been introduced to address some of the recommendations of these studies, little action has been taken on the bills, and significant changes in the FAA's institutional status have not been made. Nonetheless, the FAA has begun making progress in several major areas of concern: support and funding for personnel has been turned around; Congress has given the agency more flexibility on pay in high-cost areas; Administrator James Busey has started to improve the FAA's management of its technology acquisition program; and U.S. Department of Transportation (DOT) Secretary Samuel Skinner has eschewed the intervention in FAA administrative and technical issues exercised by his predecessors. The key question at this stage is whether these changes will be sufficient to ensure substantial improvements without being accompanied by institutional reform to reduce the ad- ministrative and political constraints within which the FAA operates. Most studies of the FAA have argued for reforms that can be catego- rized under four main headings: funding, personnel management and pay, procurement, and management. Examples of problems that have emerged in each of these areas are reviewed in this chapter in order to demon- strate the seriousness and complexity of some problems. The extent to which problems in one area are reinforced by others suggests that the prospects for improvement in one area depend on improvement in the others as well.
240 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION FINANCING Appropriations Despite the austerity in the federal budget during the last decade; the FAA's overall budget appropriations have increased in current dollars almost every year since 1971 (Table 7-1). When measured in real terms (in 1982 dollars), the FAA's funding declined markedly at the beginning of deregulation, but has since rebounded (Figure 7-1). Most of the in- crease, however, has been for the National Airspace System Plan (NAS Plan) and for grants to state and local authorities for airports. Appropri- ations for personnel and research and development have either declined or remained constant. The largest single budget item is for the FAA's operations and main- tenance account, which supports primarily the operation of the ATC system and personnel, but also includes the FAA's safety programs and personnel (Figure 7-1). This account declined in real terms between 1979 and 1982, and then remained essentially flat through the mid-1980s, even though operations handled by the air traffic service increased substantially, as did demands on the FAA's safety inspection work force. Funding for this account began to increase with the rebuilding of the FAA's work force during the latter 1980s and, in real terms, was roughly the same in 1989 as it was in 1978. Discussed in more detail in a subsequent section of this chapter are the effects of political disputes and shifting budget priorities on the FAA's personnel. Worthy of mention in this section, however, is how the annual appropriations cycle and the accompanying disputes about the budget weaken management effectiveness. The protracted debates between Con- gress and the executive branch over the budget, which have become standard practice in recent years, have resulted in tremendous inefficien- cies. Instead of being able to devote their full attention to the provision of aviation safety and ATC capacity, FAA managers are frequently dis- tracted by the need for contingency planning for changes in funding. By far the largest increase in the FAA's funding occurred in its facilities and equipment account, which covers the cost of the NAS Plan and the acquisition of additional technology to replace obsolete equipment until the upgraded NAS Plan technology is available. In real terms this account has more than quadrupled since deregulation. As an example of the extent to which the problems faced by the FAA are interdependent, as funding in this area increased dramatically, the FAA was unable to expand the
TABLE 7-1 HISTORY OF FAA APPROPRIATIONS Year Airport Improvement Program Facilities and Equipment Research, Engineering, and Development Trust Fund Operations and Maintenance Trust Fund° Total General Tax Revenue Operations and Maintenance Total FAA" Appropriations 1971 260 48 24 34 366 1,223 1,589 1972 295 302 63 989 1,649 252 1,901 1973 295 303 66 0 664 1,098 1,762 1974 300 250 62 0 612 1,322 1,934 1975 355 227 58 0 640 1,438 2,078 1976 441 246 85 0 772 2,031 2,803 1977 545 200 74 250 1,069 1,530 2,599 1978 555 200 81 275 1,111 1,659 2,770 1979 629 291 - 75 300 1,295 1,834 3,129 1980 640 293 75 325 1,333 1,899 3,232 1981 450 350 85 525 1,410 1,882 3,292 1982 450 261 72 810 1,593 1,538 3,131 1983 804 625 103 1,277 2,809 1,466 4,275 1984 800 750 263 0 1,813 2,644 4,457 1985 925 1,358 265 1,110 3,658 1,649 5,307 1986 885 895 237 427 2,444 2,381 4,825 1987 1,025 805 142 621 2,593 2,361 4,954 1988 1,269 1,108 153 826 3,356 2,368 5,724 1989 1,400 1,384 160 471 3,415 2,976 6,391 1990 1,425 1,721 170 807 4,123 3,017 7,140 1991 1,500 2,500 190 2,846 7,036 1,242 8,278 (requested) NoTE: Data were provided by FAA and are in millions of current dollars. Appropriation from the Airport and Airway Trust Fund. bAppropnatiOn from both the Airport and Airway Trust Fund and general tax revenues of the U.S. Treasury for the total FAA budget
242 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION BiIIion8 1971 1973 1975 1977 1979 1981 1983 1985 1987 1989 Year - Airport Improvement Program âe-- Operations and Maintenance â4â Facllftles and EquIpment .---- Total â4'-- Research, Engineering, and Development FIGURE 7-1 FAA appropriations in constant 1982 dollars, 1971-1989. technical staff needed to manage adequately the acquisition contracts. Indeed, by some reports the most competent FAA technical personnel needed to manage these contracts either retired or were lured away by higher-paying jobs in private industry. Thus it appears that some of the delays in the NAS Plan, which are attributed in part to cumbersome federal procurement regulations, can also be traced to insufficient staffing, both in numbers and expertise. Funding for airports (the airport improvement program account) has also increased. The FAA is not the recipient of these funds, but instead provides grants to state and local authorities for airport improvements that enhance safety, capacity, and security. The FAA's research, engineering, and development funding increased during the mid-1980s, but has since declined. These funds are primarily used to conduct research and development (R&D) to improve the safety and capacity of the ATC system, including research in aerospace medicine and human factors. As an indication of a change in the administration's budget priorities, the President's 1991 and 1992 budgets recommended increases in both fiscal years for R&D (Executive Office of the President 1990, 1991). Whereas the FAA's budget has increased in real terms during a period of federal budget austerity, most of the increase has resulted from the
Constrainis on the Performance of the FAA 243 FAA's investment in developing the NAS Plan and related technology. Throughout much of deregulation, funding for R&D has remained fairly constant, if not below historic levels. Funding for FAA staff declined immediately after deregulation, and only began to increase (in current dollars) during the mid-1980s, well after the surge in aviation activity that was stimulated by deregulation. When measured in constant dollars, even as the expanding commercial aviation system increased the demands on the FAA, funding for staffing and research by the end of the decade was little different from what it was at the outset of deregulation. Although some of the technologies of the NAS Plan have been implemented during this period (e.g., new Host computers and new ATC software), few of the ultimate benefits of the NAS Plan investments have yet been realized. Revenue Sources The FAA relies heavily on the Airport and Airway Trust Fund as a source of revenue, but during the last 2 decades about half of its total budget has been funded from general tax revenues (CBO 1988). The Trust Fund receives revenues from several aviation taxes, but most (two-thirds) comes from the Passenger Ticket Tax, which, until recent increases, has been an 8 percent tax on commercial airline tickets (CBO 1988). Because of the steady growth in commercial aviation during the last 2 decades, the Trust Fund has provided a constantly growing stream of revenue. In recent years, the Trust Fund has developed an unobligated balance of several billion dollars, and considerable controversy has developed over the unwillingness of the past and current administrations to draw down the surplus. Whether this balance is a true surplus depends on which aviation costs are expected to be covered by the Trust Fund, a matter of disagreement since the inception of the Trust Fund. If only capital ex- penditures are expected to be covered, then the Trust Fund can be con- sidered to have a surplus; but if the cost of the FAA that supports the demands of commercial and recreational aviation for airports, ATC, and safety oversight is included, then the issue is not the appropriation of insufficient Trust Fund monies but overreliance on general tax revenues (CBO 1988). In 1990, after debating this issue for several years, Congress agreed to fund 75 percent of the FAA operations from the Trust Fund to better align the FAA's costs attributable to the private sector with the revenues earned from the private sector.' Congress also raised the Passenger Ticket Tax from 8 to 10 percent of each airline ticket. Despite these recent policy
244 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION changes, a considerable share of the FAA's budget (25 percent) will continue to be dependent on general tax revenues. Given that the federal budget deficit is projected to remain well above historic levels for many more years, budget strategies by the administration and Congress to contain the size of the federal deficit, including the use of the Trust Fund to mask the true size of the deficit and simple restrictions on total spending, are likely to continue to restrain funding provided to the FAA. In addition, in the near term major spending increases will be needed for the NAS Plan, which will also place additional pressure on expenditures from other accounts. PERSONNEL The FAA is a large, diverse agency with employees at work around the clock at 2,000 sites across the nation. More than 50,000 people work for the FAA in work forces classified as ATC specialists, electronic techni- cians, safety inspectors, engineers, or administrative personnel. Some problems in personnel pay and management are unique to a single work force, particularly ATC, whereas others are more generic problems within the civil service system. Both types of problems are highlighted in the following sections. ATC Problems in the controller work force receive most of the attention in the media and in congressional hearings, but the FAA's maintenance tech- nicians are also critical to the operation of the ATC system. As reviewed in the following sections, the FAA has struggled to staff both these work forces, even as the demand for them has grown. Past Shortages of Controllers The problems leading up to the 1981 controllers' strike had been devel- oping for many years before deregulation. The 1981 strike culminated from a series of disputes between the conirollers' union and the FAA management in which the two had become increasingly polarized over staffing levels, pay, working conditions, and the management style of the agency. As a result of the strike, the FAA began the lengthy and difficult
Constraints on the Performance of the FAA 245 task of hiring and training several thousand persons to replace the fired controllers and has made efforts to improve communications and relations between managers and employees. In 1980, before the strike, the FAA had about 16,250 controllers, of whom about 81 percent (about 13,200) were classified as full performance level (FPL) (ASC 1988b, 136). In addition to being seasoned employees, FPL controllers are trained and certified on all positions within a given area.2 During the strike the FAA managed the airspace with controllers who were not members of the Professional Air Traffic Controllers Or- ganization (PATCO), nonstriking PATCO members, supervisors, and mil- itary controllers. Demand was managed by limiting the number of aircraft in the air, the precursor to the present system of flow control. After the strike, the total number of FPL controllers fell to about 5,000, but by 1985 the number had increased to 8,300, about 60 percent of the size of the prestrike FPL work force. By this time the FAA's ATC staff, both FPL and less senior personnel, were handling more traffic than before the strike. By the end of 1990 the FAA had 10,800 FPL controllers, about 2,400 fewer than the prestrike FPL work force (GAO 1991). Political Disputes over Staffing Levels Although the FAA immediately began hiring and training new controllers to replace the fired strikers, the Reagan administration did not believe that it was necessary to replace all of them. Before the strike, the FAA had been perceived by some as over staffed (NAPA 1986, 158). For several years, Reagan administration officials argued with Congress that it did not need to replace all the fired controllers, and the debate with Congress over the appropriate number of controllers became increasingly fractious. To end the debate Congress mandated that the FAA rebuild the work force by 1988 to 15,900, of whom 10,450 were to be FPL (GAO 1988b). The administration ultimately agreed that additional controllers were needed, but an issue that might have been resolved on its technical meritsâwhich would focus on the ability of the ATC service to perform its taskâbecame instead a lengthy political argument over the number of FPL controllers. Hiring and Training Backlogs By the end of 1990 the FAA had largely rebuilt the controller work force but still had fewer FPL controllers than in 1980 (GAO 1991). It
246 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION had fewer FPLs in part because of the time it takes to recruit, hire, and train an air traffic controller. It takes about 3 to 5 years for a controller to reach the FPL designation once accepted into the training program, but few applicants make it to the FAA Academy (GAO 1988b). Until changes in the FAA recruiting were instituted in 1989, only 25 of every 100 applicants passed the exam administered by the Office of Personnel Management, and the FAA was able to hire only 2 of 5 with high scores (the other 3 found jobs before the FAA could hire them). Almost one of two individuals entering the training program washes out, and about 20 percent of those who become controllers never make the full per- formance designation (GAO 1988b). In other words, until recently it took more than 100 applicants for the FAA to achieve a single FPL controller. The FAA has made several efforts to meet the congressional mandates. It has expanded the training program and streamlined its recruiting and hiring practices to get more applicants into the pipeline. Whereas it once took the FAA 18 months to hire an applicant, it can now do so in 45 to 60 days (U.S. Congress 1989). Success in moving applicants into training is but one phase of the course toward becoming a controller. Most of the training is performed on the job, and because it takes several years for an applicant to reach the FPL designation, the rebuilding process is nec- essarily slow. Even so, the General Accounting Office (GAO) has been quite critical of the FAA's management of its training programs, which is contributing to the slow pace (GAO 1988b; Mead 1990a). Shortages in High-Cost Areas Although the number of FPL controllers is currently still smaller than before the strike, a review of the overall staffing level of controllers by the Aviation Safety Commission concluded that there was "no evidence of significant system-wide staffing shortages." The report, however, also found, "At some of the busiest facilities there are significant problems. The problems are especially prevalent in high-cost-of-living areas." These high-cost areas (e.g., Chicago, New York, San Francisco, and Los An- geles) are also among the most heavily traveled sectors in the ATC system. Controllers in these areas must manage the most complex and densely traveled airspace. Staff shortages in high-cost areas are caused by the desire of controllers to move to less expensive areas, the high washout rate of trainees, the difficulty recruiting controllers from other areas, high training workloads, and excessive overtime (ASC 1988b, 137).
Constraints on the Performance of the FAA 247 Rigidities of Civil Service Pay Part of the problem in high-cost areas is that the FAA has been unable, under civil service regulations, to pay sufficient wages to attract and retain employees.3 Given the approximate equality in pay across the nation, controllers have naturally wanted to move to less stressful areas where the cost of living is lower and the pay is roughly the same. Congress has responded to this problem by authorizing an experimental program to pay controllers and some technicians wage differentials of up to 20 percent to attract them to high-cost cities that have staffing shortages. What Congress allows with one hand, however, it restricts with the other. Although the FAA was given the authority to pay higher salaries to attract personnel to understaffed centers and towers, Con- gress did not provide any additional funds. The complete results of this demonstration program are not in (implementation was delayed by a lack of funds to pay employee relocation expenses), but early returns are mixed. Some participating ATC facilities showed slight improve- ments in staffing, but in other high-cost areas employees simply trans- ferred from nonparticipating facilities to participating facilities within the same area. At least one major facility targeted by the program showed no improvement at all (K. Mead, testimony before Subcom- mittee on Aviation, House Committee on Public Works and Transpor- tation, March 14, 1990, and Subcommittee on Investigations and Oversight, House Committee on Public Works and Transportation, June 6, 1990; Aviation Week and Space Technology 1990a, 1990c; GAO 1991). Current and Future Controller Work Force The current ATC work force, which includes developmental controllers, FPL controllers, supervisors, managers, and support personnel, totaled 24,195 in the spring of 1990. Of the 24,195 air traffic specialists, about 17,000 are either first-line supervisors, FPL or other controllers, or de- velopmental controllers (individuals in on-the-job training). Because of the hiring during recent years to replace the fired strikers, this work force is relatively young, in sharp contrast to the age distribution of electronics technicians (discussed next). The FAA estimates that by 1999 more than a quarter of the work force, about 6,750 of 24,195 ATC specialists, will have become eligible to retire (about a quarter of newly eligible retirees will be controllers presently
248 WiNos OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION classified as FPL). Most of the ATC work force that becomes eligible to retire by the year 2000 is likely to retire by then; indeed, 2,500 of these workers are already eligible to retire. Although the FAA's personnel staff perceive little problem recruiting additional air traffic controllers, personnel planning in the federal gov- ernment must deal with unanticipated changes in policy that can have direct effects on personnel. Until 1990, for example, federal employees were able to receive their retirement contributions in two lump sums, and for many employees this has been an attractive option. Indeed, the fear that Congress would eliminate this benefit as part of the 1990 negotiations over the federal deficit reduction compromise caused an extraordinary exodus of senior FAA personnel in August 1990 (Washington Post 1990). Roughly 340 FAA employees decided to retire in August 1990, including 117 senior air traffic controllers and 84 maintenance technicians. The loss of senior managers was even more striking: three of the FAA's five newly appointed executive directors retired, as did three of the FAA's nine regional administrators. For the future, the FAA personnel staff report a considerable inventory of qualified individuals interested in becoming controllers. Whether this future work force will be adequate depends on the projected benefits of the NAS Plan. The FAA projects that between 1990 and 2000 the number of operations (departures and landings) at airports that have the FAA towers is expected to grow by 25 percent (FAA 1989). The new NAS Plan technologies are expected to increase controller productivity sub- stantiallyâby a factor of twoâbut given past and projected delays in the NAS Plan, the new technology may not be on line in time to meet projected demand. In addition, as discussed in more detail later, the new technology also may change controllers' tasks substantially, with implications that are as yet not fully understood for the number of controllers and the skills needed. Maintenance Technicians FAA's maintenance technicians repair and maintain the ATC system's electronic equipment and power plants. During the early 1980s the FAA grudgingly allowed the number of maintenance technicians to decline as the 0MB systematically ratcheted downward the number of technician slots in view of plans for installing new, more reliable equipment and in view of the millions that the FAA was already spending each year on the NAS Plan. From a high of about 10,200 technicians in 1979, the total
Constraints on the Performance of the FAA 249 Total No. in 1990 2500 2000 1500 1000 500 Le88 than 10 10-19 20-29 30. Current Yeare of Experience Retired by 2000 FIGURE 7-2 FAA techniciansâexperience of work force in 1990 and projected retirees by 2000 (data from FAA). number declined to 7,560 in 1986 (GAO 1987a, 20). This level of attrition in the work force exceeded the FAA projections, and by 1987, the GAO (1987a) concluded, "Critical technician vacancies currently exist through- out the field." The GAO report also noted that by 1990 one-third of the technician and engineer work force would be eligible for retirement and that the FAA did not have a sufficient incoming stream of employees in training to replace the expected number of retirees. The electronic technician work force, which accounts for more than two-thirds of the field maintenance staff, is the FAA's oldest. This group is responsible for maintaining radars, communications equipment, and computers. As of spring 1990, roughly half were over the age of 45 and had 20 or more years of experience (Figure 7-2). Nearly 60 percent of these highly experienced workers will become eligible to retire within the neXt decade. The projected retirements in Figure 7-2 are based on the assumptions that (a) all workers who reach age 60 or older with 40 years or more experience will retire by 2000, and (b) half of workers eligible for retirement by age 60 with between 5 and 40 years of experience and half of workers who reach age 55 with 30 years experience will retire by 2000. Under these assumptions, 40 percent of the current work force would retire, including virtually all the workers who presently have 30 or more years of experience and 40 percent of workers who have 20 to 30 years of experience. Although these estimates depend in part on arbitrary assumptions, they may underestimate actual retirements because in the past most FAA personnel have retired within 2 years of reaching eligibility.
250 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION Thus, the number of actual retirements could be a greater proportion of those eligible than estimated in Figure 7-2. Although the FAA was aware of the pending staff shortages in earlier years and requested more positions, the Reagan administration did not support the request (GAO 1987a). During 1986 and part of 1987 the administration imposed a hiring freeze on technicians, but Congress in- tervened. A House and Senate conference committee provided more funds for fiscal year 1987 than the administration requested and strongly rec- ommended that the FAA increase the number of technicians to 8,551. The FAA nearly met this goal in 1987, and was expected to reach 8,800 in 1990 and 9,000 by 1991 (U.S. Congress 1990). Current FAA estimates indicate that it will need about 9,700 technicians to maintain the NAS Plan technologies once they are in place. The Federal Employees Pay Comparability Act of 1990 (P.L. 10 1-509) offers some incentives to help retain and rehire retired employees. A retention allowance of an additional 25 percent of base pay can be paid and rehired employees can receive full pay and retirement benefits. The provisions in the law may help the FAA maintain an adequate number of employees, but the procedures are cumbersome. The FAA will have to prove the need for rehiring or retaining on a case-by-case basis and must prove that an exceptional difficulty exists. In order to meet future needs, the FAA has streamlined the training program to move more technicians into the field faster than it has in the past. The GAO, however, has expressed considerable skepticism about the ability of the FAA Academy to accommodate the number of incoming employees needing training (GAO 1988b; K. Mead, testimony before Subcommittee on Aviation, House Committee on Public Works and Trans- portation, March 14, 1990, and Subcommittee on Investigations and Over- sight, House Committee on Public Works and Transportation, June 6, 1990). Given the projected level of retirements, training needs are likely to continue well into the future, as is the current, backlog of new hires who must be trained. The FAA is the only major employer of the controller work force, but it must compete with other federal agencies and private industry for the pool of potential technicians. The FAA personnel staff do not believe that the FAA's salaries for entry-level technicians are competitive with private industry or other federal agencies. Competition for electronic technicians will continue and be heightened by the expected shortage of entry-level personnel in technical fields by the turn of the century (Johnston and Packer 1987). This is one of the most fundamental difficulties the FAA has in its current institutional form. As an agency of the federal government
Constraints on the Performance of the FAA 251 the FAA is required to operate within a civil service system that does not provide adequate flexibility in pay for the technical and managerial ex- pertise that the agency needs. As discussed in the next section, one of the FAA's principal strategies for dealing with shortages in its maintenance work force is to rely more heavily on outside contractors to perform maintenance on the new ATC technology. NAS Plan Effects on ATC Personnel The FAA's initial optimistic projections of future NAS Plan deliveries during the early 1980s contributed to shortages in staff, particularly tech- nicians. For example, regional managers complained that the FAA reduced technical personnel slots before some systems were even in place and made inadequate provisions for training and debugging during transition to the new equipment (GAO 1987a). GAO officials continue to express doubts about the sufficiency of the FAA staff to handle the transition period when the new ATC technology must be phased into FAA operations (K. Mead, testimony before Subcommittee on Aviation, House Committee on Public Works and Transportation, March 14, 1990). The FAA's demand for air traffic personnel may be eased by the future delivery of technology, but the changing nature of the work required poses some complex problems for the FAA in planning for both the controller and technician work forces. The productivity improvements for control- lers, for example, assume that advanced technology will allow automated procedures to replace many of the separation assurance tasks now handled manually by controllers. Extensive testing and simulation of automated procedures with controllers indicates that these productivity benefits can be realized, but will entail substantial changes in the controllers' tasks (Celio 1990): Reliance of the controller on AERA [Automated En Route Air Traffic Con- trol] will shift the controller's focus of attention to strategic planning tasks, with less focus on tactical monitoring, problem analysis, and coordination. These changes will create a need for new controller procedures, training, performance evaluation criteria, and a review of controller selection criteria. The FAA currently has little research under way to prepare for these potentially fundamental changes in the ATC task and the implications it has for total staffing, how new employees will be identified and trained, and how existing employees will be retrained. (The FAA has a plan under
252 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION development that agency staff believe will address these concerns; the plan is expected to be released in 1991.) In addition to acquiring new technology that will have less labor- intensive maintenance requirements, one of the FAA's primary strategies for handling future maintenance and repair is to rely more heavily on contractors. Beginning in the early 1980s, following the administration's emphasis on privatization, the FAA began contracting out certain main- tenance tasks, and, to prove the cost savings, proposed a large-scale ($130 million) demonstration program. Neither the House nor the Senate Ap- propriations Committee was convinced of the merits of using these pro- posals to reduce the maintenance work force, and both refused to appropriate funds for the demonstration (GAO 1987a, 1987b). The FAA maintenance employees have expressed great skepticism about proposed FAA contract maintenance, which, in addition to the impingement of political disa- greements on what could be considered a technical question, reportedly has reduced the morale of the technical work force (GAO 1987b). Contract maintenance is also a major component of the NAS Plan. The FAA's profile of the expertise expected of electronics technicians and its training programs do not yet reflect the ways in which the NAS Plan will change the demand placed on its technicians, even though many of the technicians being hired and trained currently may become responsible for maintenance when the new technology is phased in. In light of this con- cern, the FAA has made provision to contract out maintenance for some advanced technology, recognizing that the vendors who develop the tech- nology will be more familiar with it than the FAA staff. The technology will also be considerably more dependent on software, an expertise not expected of current electronics technicians. Maintenance technicians and managers have raised questions about the efficacy of these proposals and have expressed concern about whether the FAA will have staff technically capable of providing adequate oversight as contract monitors. At present the FAA technical staff have a complete understanding of the individual components of the ATC system, how they all fit together, and how they can be fixed. As more expertise is acquired through contracting out, particularly software expertise, the technical managers are worried that this integrated understanding will diminish. The FAA is attempting to respond to this concern with its plan for long-range, human-resource training. Although the NAS Plan technologies may reduce the need for the num- ber of technicians in the future, the slower-than-expected deployment of the NAS Plan means that older technology will be in service and require maintenance and repair for longer than anticipated in the FAA's staffing
Constraints on the Peiforinance of the FAA 253 plans (K. Mead, testimony before Subcommittee on Aviation, House Committee on Public Works and Transportation, May 9, 1989). This raises some concern about the adequacy of the technician work force during the next decade when most senior technical employees will retire and when the potential labor saving technologies are not yet in place. The lack of longer range planning for personnel needs at the FAA, though worthy of concern, is not surprising. The entire agency, along with the rest of the federal government, operates under an annual appro- priations cycle. Even when this system works well, it tends to focus management attention on the needs for the next year's budget and divert focus from longer term planning. Because of the protracted budget debates of recent years, the fiscal year is often well under way before agencies' actual budgets are determined, and therefore little incentive exists to ad- dress long-term needs. The effect that the annual budget cycle has on short-term planning is compounded by the short tenure of top agency managers (discussed in a later section). Safety Inspections The Federal Aviation Act of 1958 charged the FAA with inspecting the operations of air carriers. The FAA conducts different kinds of inspections. Maintenance inspectors responsible for individual airlines periodically in- spect maintenance procedures (focusing on maintenance records) to ensure that airworthiness directives have been complied with (ASC 1988b). Main- tenance shops and aircraft also are inspected routinely. Flight standards inspectors review carrier staffing, shift assignments, and training, and inspectors with pilot training perform "check rides" to ensure that com- mercial pilots follow the FAA procedures. The FAA has been rebuilding its safety inspector work force since the mid-1980s after allowing it to diminish during the early years of dereg- ulation. The growth in the number of new carriers during the early years of deregulation and the growth in the number of mergers placed extensive new demands on the inspector work force to certify and inspect new or merged carriers. From about 1,500 inspectors in 1980, the number had increased to nearly 2,000 by 1988 (GAO 1988b, 36). The President's budget request for fiscal year 1991 calls for increasing the number to just over 3,000, which, some 7 years later, is the number recommended in an internal FAA report. As with other segments of the FAA's work force, shortages of inspectors exist in areas with high living costs. In addition, attrition is fairly high
254 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION throughout the work force, much of it caused by retirements. Roughly 30 percent of the current work force is likely to retire in the next decade. The FAA has been actively hiring inspectors for some time and currently has an inventory of nearly 1,000 qualified applicants; thus it appears that the FAA has a reasonable number of current and prospective workers in the pipeline. However, the FAA also has trouble competing with the private sector. In fiscal year 1988 the FAA estimated that the number of inspectors who found higher paying jobs in industry almost matched the number of retirees (GAO 1988b). FAA personnel experts believe that FAA salaries are not competitive with private industry, particularly in areas that have high costs of living. Even as the FAA stretches its maintenance personnel to meet existing maintenance requirements, it is substantially upgrading the requirements for inspecting aging aircraft to detect corrosion and fatigue. The recent new hires to bolster the safety inspection work force are designated to help the FAA ensure that air carriers meet the existing regulations. Many experts inside and outside the FAA believe the inspector work force will not be adequate in numbers or in technical training for these new demands (Aviation Week and Space Technology 1 990c). Whereas these concerns may be speculative at this stage, they cannot be dismissed out of hand. The FAA's inspection personnel have long been overstretched to meet the demands of a rapidly growing and changing industry. Although the number of personnel is increasing, the requirements are increasing as well, both for numbers and expertise. Shortages in Personnel with Special Skills The FAA's various regulatory functions require personnel with specialized skills. Pilots are needed to perform check rides in the most sophisticated aircraft. Engineers knowledgeable about the latest technology in materials and avionics are needed to certify the airworthiness of the most sophis- ticated new aircraft. Managers with scientific and technical training are needed to manage the acquisition of advanced technologies that are part of the FAA's capital improvement program. In all these areas involving high levels of expertise, the FAA has dif- ficulty attracting and retaining good personnel because of the relatively low entry pay for technical personnel in the Civil Service System. This is not a new problem. A 1980 study by a committee assembled by the National Academy of Sciences expressed concern that the FAA's inability to attract adequate technical competence for certifying the airworthiness
Constraints on the Performance of the FAA 255 of new aircraft was resulting in increasingly superficial technical review (NAS 1980). The FAA (1987) acknowledged difficulty in "recruiting and retaining well qualified aerospace engineers, manufacturing inspectors, and test pilots. . . . The current pay structure is not competitive with the aerospace industry." More recent studies have reached similar conclu- sions. For example, the Aviation Safety Commission found that starting pay for engineers hired by the FAA are well below those in the private sectorâas much as $10,000 less for those who have Ph.D's (ASC 1988b, 139). The problems are the most acute in areas that have high costs of living, which also happen to be areas in which a great deal of aviation activity occurs. Changing the FAA Culture After the controllers' strike and partly in response to controller frustrations with its management style the FAA formed a committee of experts, chaired by L. M. Jones, to review concerns about employee relations. Reports of the Jones Committee, issued in 1982 and 1984, pointed out a number of problems in communication and management style. Among other things, the committee found that the autocratic, militaristic management style typical of many FAA supervisors was not well suited to the participatory style desired by many employees, particularly controllers. The Jones Committee reports called for several changes in the FAA, including improved human resource management, a process for selecting managers with human relations skills, extensive use of organizational development techniques to improve communications, and greater em- ployee participation [as summarized by the Aviation Safety.Commission (1988b, 133)]. The FAA has responded to many of the recommendations in the Jones Committee reports. It now has, an Associate Administrator for Human Resource Management and an office responsible for organi- zational development. Training in communications and problem resolution is being provided to employees and managers. More emphasis is being placed on people management skills in selecting and training managers. The FAA has also surveyed its employees to determine staff morale. The employee surveys indicate that some of the management changes are being favorably received. On a number of questions regarding first- line supervisors, positive responses by employees increased by more than 5 percentage points between the 1984 and 1986 surveys (ASC 1988b, 134). Employee satisfaction with the FAA as an organization and with upper-level management, however, was poor in 1986, especially in the
256 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION en route centers. (Results from the 1988 survey were not available in the same format as previous surveys, which did not allow for comparisons over time.) The FAA is trying to use its survey results to improve employee-management relations throughout the agency. For example, on the basis of responses to the 1988 survey, the annual performance criteria of managers were modified to reflect employees' perceptions of managerial shortcomings. The FAA plans another employee survey for late 1991. As an indicator of the lack of confidence in FAA management, the air traffic controllers hired to replace the PATCO strikers subsequently voted (with 70 percent in favor) to be represented by a new union. This vote can also be interpreted as a reflection of problems with management that are inherent within a government agency. Managers have less flexibility for rewarding outstanding performance and, in the FAA's case, difficulty in making adjustments in the work force to respond to increased demand. The GAO has also been surveying the ATC work force about staffing levels, working conditions, system safety, and morale (GAO 1989). The 1988 survey found that most controllers and supervisors rated the overall safety of the system as "adequate," "good," or "excellent;" nonetheless they were concerned about a number of problems, including low morale and insufficient numbers of FPL controllers to handle the work load. In contrast with the results of the FAA's surveys, the GAO found little change in controller and supervisor responses on most questions when compared to their 1985 survey (GAO 1989). The GAO (1989, 8) noted that the FAA improvements in recruiting, training, and the pay demonstration project were steps in the right direction, but "these are long-term initiatives that will require time to bring about needed changes." Summary In recent years, and after considerable wrangling within the Reagan ad- ministration and between the administration and Congress, the FAA has been rebuilding its work force of controllers, maintenance technicians, safety inspectors, engineers, and test pilots. As a result of these new hires, the agency now has its largest staff since the late 1970s. The current work force is less experienced than in the past, however, and because of the growth in the aviation system, the demands placed on the FAA have been growing faster than its work force. Although systemwide staffing needs for air traffic controllers appear adequate, at least in terms of numbers, staffing problems do occur in parts of the country with high living costs, where some ATC centers and towers are understaffed. The new experi-
Constraints on the Performance of the FAA 257 mental program to offer wage differentials of up to 20 percent to attract more personnel to these locations that are difficult to staff may help, but early returns are mixed. Whether the increased support the FAA has found in responding to its personnel needs will be adequate without being accompanied by funda- mental institutional changes is brought into question by a variety of com- plex and interconnected issues reviewed in this section. The agency has been caught up in disputes between Congress and the Executive Branch over the number of controllers and other staff required and over the po- tential benefits of relying on more contracted services; these disputes resulted in personnel decisions being resolved on a political basis instead of a technical one. The FAA's staffing shortages in the past, the looming retirements of many of its most experienced technicians, its difficulty in attracting and retaining personnel with high-level technical and managerial expertise, and the lack of preparation for the changes in staffing that can be expected with the NAS Plan are all cause for concern. Substantial increases in the need for funding for personnel can be foreseen at the same time that the growing budget deficit will be placing enormous pressures to constrain spending. Whereas FAA managers apparently are making several efforts to respond to these issues, the question is whether they will be able to succeed, given the constraints within which they must operate. ACQUISITION OF ADVANCED TECHNOLOGY As discussed in Chapter 6, acquisition of the advanced ATC technology for the NAS Plan has been delayed and is not likely to be complete until after the turn of the century. Some of the delay is attributable to the scale and complexity of the NAS Plan, which rivals any other civilian high- technology development effort in the federal government (including those of the National Aeronautics and Space Administration) or in the private sector. Blame for some of the delay has also been attributed to inadequate management (partly attributable to the inability to recruit sufficient tech- nical managers) and cumbersome federal procurement regulations. Inadequate Planning and Management The FAA has been criticized by industry and the GAO for inadequate planning and management of its acquisition of the NAS Plan technologies. According to the GAO, the FAA underestimated the technical complexity
258 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION of some of the systems and made overly optimistic forecasts about delivery dates (GAO 1988a). The GAO has also argued that the FAA began pro- curing equipment before carefully defining its needs and made inadequate provisions for testing and evaluation in the procurement process (GAO 1988a, 1990a, 1990b). Had the FAA followed the procurement policy outlined in 0MB Circular A-109, which calls for continual reassessment of progress by top management and a "fly-before-buy" approach, many of the problems encountered could have been avoided, according to the GAO (Mead 1990b). Thus, from the GAO's perspective, the fundamental problems that the FAA has had in procurement stem from its failure to follow recommended government procurement policies and not from the policies and procedures themselves. The FAA's first Executive Director for Acquisition, John Burt, who filled a position created by Admiral Busey in 1990, shares the GAO's view that the FAA has not followed the 0MB Circular A-109 acquisition process (Burt 1990). Burt also acknowledged that in the past the respon- sibility for acquisition was diffused and lacked individual management responsibility and accountability, and that the FAA failed to staff-up ad- equately with managers competent in the acquisition of high-technology systems. Admiral Busey has instituted several reforms to improve the FAA's management of its acquisitions program: changes in internal pro- cedures to ensure that the A-109 process is followed, realignment of the FAA's groups responsible for systems development and acquisition to improve coordination, and creation of additional management positions with direct responsibility and accountability (Burt 1990). Complexity of High-Technology Acquisition Simple compliance with 0MB Circular A-109 procedures, however, may well not have avoided some of the problems the FAA has encountered with the NAS Plan, nor will it necessarily help the FAA respond to some of the immediate demands for capacity. In order to get support in Congress for an expensive acquisition program, the FAA and administration officials apparently felt compelled to put together a specific, comprehensive, high- technology package with a rapid development and implementation sched- ule that would also reduce the FAA's demand for personnel (a compelling argument in a period during which the FAA had just fired 11,000 air traffic controllers). By describing a fairly specific set of projects and by stating that the acquisition would improve the productivity of the FAA staff and improve the safety of the traveling public, the FAA was able to
Constraints on the Performance of the FAA 259 convince Congress to spend billions of dollars in an era of constrained resources. The initial NAS Plan depended, however, on several emerging tech- nologies that had not been fully proven and which no one had tried to put together into a single integrated system. One of the FAA's major con- tractors is primarily concerned with this systems-integration process and problems, whereas others are responsible for developing individual tech- nologies. As the GAO has argued, in some areas the FAA did not specify its needs very well in advance, nor did it anticipate the complexity of the systems engineering job that would be required for all these new tech- nologies to be developed, fit into a single integrated system, come on line as scheduled, and work with the high degree of reliability required for ATC. But, given the speed with which technology is changing in high- speed data transmission, satellite-based communications, and computer capability, no one has the prescience needed to specify a complete system 10 to 15 years in advance of delivery, especially one that can accommodate something as complex as ATC. Over time the FAA realized that the replacement of technology cannot be handled as a massive, one-shot effort, but must be an ongoing process. The current NAS Plan, for example, is now viewed as part of an ongoing capital investment program. The Aviation Safety Commission concluded that current federal pro- curement regulations are not well suited to agencies such as the FAA that depend on advanced technologies (ASC 1988a, 22). The implication of following all the steps required in these cumbersome regulations, which can take years to complete, is that by the time the FAA receives the technology it set out to acquire, the technology has been superseded by technologies developed in the interim. Before the original NAS Plan is completely in place, the FAA will be forced to adapt the technologies to demands being placed on the air traffic system that were not anticipated when the plan was first formulated (Fearn- sides 1988). Although the planners recognized that additional radar, com- munication, and computer capacity would be needed to respond to demand and provided for such technical capacity, they did not anticipate the sat- uration of airport and airspace capacities in the air traffic network. Con- sequently, they did not incorporate capabilities to help controllers manage this demand (Fearnsides 1988). This is an important distinction, the im- plications of which are difficult to appreciate in a brief overview, but the constraint on airspace capacity is the number of aircraft a controller can manage at one time, not the number the radars can track or the computers can display simultaneously. The NAS Plan technologies, once delivered, will be able to better identify and more precisely track more aircraft than
260 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION the typical controller can safely manage. In the past controller limitations resulting in reduced system capacity could be overcome by reducing the size of the sectors of airspace that controllers manage. Such action, how- ever, increases dramatically the amount and complexity of coordination among controllers, and there is a point beyond which individual work load cannot be reduced. Since the NAS Plan was promulgated in the early 1980s, new technologies have been developed that could provide decision support tools to enable controllers to better handle current and near-term capacity needs. Integrating these new technologies into the NAS Plan, however, is an extraordinarily difficult systems engineering task. Proce- dural innovations in system development, such as those recommended in the acquisition of complex military technologies by the Packard Com- mission report of 1986, could be applied within the overall NAS Plan acquisition program (Fearnsides 1988). For example, the use of rapid prototypes, especially as they enable the participation of controllers in the systems development process, provides an avenue for more precise def- inition of system requirements, thereby reducing development risk. Find- ing ways of expediting the introduction of new system functions into the field is difficult, however, especially in the context of current acquisition regulations. The amount of technical innovation required, though achiev- able, will be precluded by a faithful adherence to the multiple reviews and steps in the current procurement process. Hence the FAA's current emphasis on following 0MB Circular-l09 procedures is a remedy for poor decisions made in the past, but adherence to these procedures will delay substantially the adaptations needed to the NAS Plan to serve growing demand. MANAGEMENT Like most federal agencies, the FAA is an executive branch agency with a clearly defined organizational chart that implies straightforward lines of authority. As with other agencies, this clarity belies a considerable amount of oversight and control exercised by Congress and by federal officials in other agencies. The NAPA's study (1986) of the FAA concluded the following: Real control has been diffused into a complex, confusing, and often ka- leidoscopic array of authorities among the FAA leadership, the Office of the Secretary of DOT, the Office of Management and Budget, the Office of Personnel Management, the White House staff, and several Congressional committees.
Constraints on the Peiformance of the FAA 261 Whereas these diffused lines of authority and control are typical of federal agencies, the FAA, unlike most agencies, is in many respects an operating entity. Apart from its regulatory duties, which mirror those of many other agencies, the FAA is responsible for operating the ATC system, which is not unlike a public utility that must operate in an efficient, businesslike fashion to supply the demand placed on it. The commercial aviation system and the safety of the traveling public depend on the continual operation of the ATC system, without any significant interruption, for 24 hours a day, every single day. Like those of other federal agencies, top FAA managers (the Admin- istrator and Deputy Administrator) are political appointees, who serve, on average, about 2:5 years, which is slightly longer than the typical length of service for appointees at this level. Unlike other agencies, the FAA is a high-technology enterprise in the midst of several long-term effortsâmost important of which is the NAS Planâin which stable top leadership is vital. Micromanagement When the federal government's various transportation agencies were com- bined into the DOT in 1966, the FAA Administrator began reporting directly, on most matters, to the Secretary of Transportation. Although the FAA lost its independent status because of this move, it was believed at the time that the FAA's access to the White House and its position with the Bureau of the Budget (the forerunner of the 0MB) would be improved by being represented by a department head with cabinet status. Along with this change in its institutional status, however, the FAA's ATC system, which is essentially a technical operation, gained a new set of supervisors in the Office of the Secretary, many of whom are political appointees. Examples were cited previously of cases in which the FAA managers knew they needed additional personnel, but they could not gain the support of either the Office of the Secretary or 0MB. According to Allan McArtor, Administrator from 1987 to 1989, "The FAA is willing, but its hands are tied. . . . To try to advance a progressive agency with all that bureaucratic baggage is very difficult. You have a lot of bosses" (Fortune 1990). During the tenures of Secretaries Dole and Burnley, the Office of the Secretary moved beyond policy issues and became involved increasingly in administrative matters. For example, senior FAA officials (now retired), have cited anecdotes of the DOT Deputy Secretary intervening in decisions
262 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION about technical personnel matters that were two and three levels be- low that of the Administrator. Micromanagement during this period appears to have been the most pervasive since the FAA was incorporated into the DOT. Although micromanagement is usually portrayed as negative, it can also be viewed as a necessary component of administrative oversight. The FAA's expertise in ATC, for which relatively little expertise exists outside the FAA, has led the agency in the past to tend to discount recommen- dations for change that come from outside the agency. Secretary Skinner's recognition of the potential benefits of congestion pricing in his 1990 National Transportation Policy report, however, may be causing the agency to consider an option for managing demand that it has dismissed as un- workable in the past (DOT 1990). As must all agencies, the FAA operates within budget guidelines es- tablished by the 0MB, which currently are driven more by concern to limit the deficit than by the programmatic needs of individual agencies. Budgets for the FAA and other agencies are not based on a "bottoms up" summary of needs. Although the 0MB has long set budget targets for individual agencies, these targets formerly included more negotiation. In more recent years the budget has been developed after the agency is told how much it can have. In dealing with 0MB, McArtor noted, "You must suggest a budget that is within the 0MB budget, and as a good soldier you must say you can do everything you want within that budgetâwhen of course you know you cannot" (Fortune 1990). Most former FAA Administrators now favor giving the FAA greater autonomy in running its programs, partly because of the frustrations in dealing with the increased bureaucracy and the greater degree of control exercised by the Office of the Secretary and the 0MB over FAA affairs. Examples were also cited previously of how the FAA has been caught between the administration and Congress in disputes over personnel, con- tracting out, and the budget. Although the FAA's situation in this regard is typical of federal agencies and is not unique, it nonetheless leads to resolution of complex issues on political instead of technical grounds, which raises concern about how well-prepared the FAA will be for the demands of the future. Stable Leadership The typical, but nonetheless short, tenure of the FAA Administrators poses a litany of problems for effective management. It is difficult to appreciate just how complex and technical the FAA's missions are, and
Constraints on the Peiformance of the FAA 263 most Administrators need their first year or more just to learn the ropes. The FAA's top managers have come from outside the agency, often from the military, and enter the job without a full appreciation of the agency's extensive role in civil aviation. In contrast to the short tenure of key administrative personnel in the FAA, military leaders, who also have short tenures in top positions, have an entire career in the ranks they will ultimately lead. Because political appointments usually last fewer than 3 years, many appointees seek to make their personal, mark on issues that are tractable in the short term, and long-term reforms are not attempted. For example, the Aviation Safety Commission noted that recent Administrators have focused on trying to resolve problems in the controller work force because the issue has a high political profile. Meanwhile, troubling problems in other, less visible, but equally critical, areas of the FAA's work force remain unresolved (ASC 1988a, 22). Other management problems are posed by the short tenure of the Administrator. For example, the bureau- cracy knows that new Administrators are not likely to stay in office long enough to see new initiatives through; hence some reforms can be blocked by simple bureaucratic inertia (ASC 1988a, 22). Given the complexity of the NAS Plan and the extent to which the FAA relies on high technology to address capacity needs, the lack of sustained leadership to oversee this program is cause for concern. The increased reliance on technology makes the management job for the FAA more complex. This places even greater emphasis on management expertise. SUMMARY Despite the constraints within which it operates and the firing of most of its air traffic controllers in 1981, the FAA has been able to maintain a high level of safety since deregulation and has been able to accom- modate a substantial growth in demand for commercial aviation, albeit not without complaint about the slow pace. The demands placed on the FAA for ATC, maintenance inspections, and other safety functions, however, appear to be growing faster than the agency's ability to re- spond, which raises concern about how well the FAA will be able to perform in the future. The concern about the future performance of the FAA emerges from several unanswered questions: Given the projections of a continued budget deficit, will the FAA obtain the funding it needs for personnel and equipment?
264 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION With the increased future reliance of the FAA on advanced technology in the provision of ATC, will it be able to maintain the current high level of safety as its mission becomes more sophisticated and complex? Given that FAA salaries are not competitive with the private sector and a shortage of expertise in technical fields is projected in the years ahead, will the FAA be able to attract and retain the personnel it needs? Will the inspector work force be adequate to handle the new demands being placed on it for inspecting aging aircraft? Given the pending retirements of most of the electronics technician work force, how will the FAA keep its aged ATC technologies operating until the NAS Plan technologies are in place and operating smoothly? Will the increased reliance on contracted maintenance undermine the expertise needed on the FAA staff to ensure that the system performs adequately and safely? Will the FAA be able to make needed changes to the NAS Plan to provide for capacity demands despite the constraints of the federal pro- curement process? Will future FAA Administrators stay on the job long enough to pro- vide stable leadership of the FAA's many complex missions and will future DOT Secretaries become involved again in unproductive micro- management? The FAA managers who met with the study committee and staff are aware of these issues and can point to many efforts to respond. The committee, however, is concerned that institutional and political con- straints that lead to uncertainty over future funding levels, inadequate long-term planning for personnel, inability to attract the level of technical expertise needed, lack of innovation, and frequent changes in leadership will considerably hamper the ability of the FAA managers to respond to these concerns. Although the committee is confident that the FAA's air traffic service will maintain its commitment to safety, the increasing demand for ser- vice is likely to result in increased delay as controllers hold more flights on the ground until they can be merged safely into the traffic stream. Meeting this demand will not be easy. More effective performance of ATC in the future will not be provided by simply hiring more con- trollers, although shortages in a few high-cost cities have been difficult to resolve. As outlined in Chapter 6, the FAA must adopt a more systemwide approach to managing ATC. (ATC management in this context refers to control of traffic in the networkâby holding aircraft on the ground, spacing of aircraft in the air, or through various routings
Constraints on the Performance of the FAA 265 provided to aircraftâto avoid overstressing points in the network where traffic from alternate routes converge). Doing so requires both a re- orientation in the FAA's approach and a more rapid development of the research and the technology required to manage this task effectively and safely. This is one of the key points at which the constraints of being a government agency impede progress. The more stable, long- term leadership needed to manage this transition to a more complex, more technology-dependent mission is unlikely to be achieved with the frequent turnover of political appointees that has been typical in the past. The more rapid acquisition of advanced technologies required to adapt the NAS Plan to unanticipated capacity needs is unlikely to be accomplished in time to meet demand within the federal procurement process. The technical and managerial expertise to accomplish this more sophisticated and demanding mission may not be available to the FAA within the personnel and salary constraints of the civil service system, even after taking into account the recent efforts of Congress to allow for pay differentials in high-cost areas of the country. The funding needed for technology and personnel is likely to run afoul of the need to constrain federal spending in future years. Budgetary constraints may also impede the conduct of the FAA's safety missions. For example, because of limited staffing, the FAA's mainte- nance inspectors tend to focus their attention on maintenance records instead of inspecting actual maintenance. Although it is within the FAA's authority to perform more on-site inspections (and more unannounced inspections), the agency does not have sufficient personnel to make this a regular practice. The increased staffing levels for maintenance inspectors supported by the administration will help, but the demands are increasing faster than the staffing levels, particularly when the need to inspect aging aircraft is taken into account. Given the likely limits on federal expen- ditures for the foreseeable future, it is difficult to be assured that staffing levels and the expertise of the staff will be sufficient in the future. Another cause for concern in other areas of the FAA's safety oversight responsi- bilities is its inability, within the pay limits of the Civil Service System, to attract and retain test pilots, aeronautical engineers, maintenance in- spectors, and electronics technicians. As laudable as the current support for the FAA is in proposals by the administration and Congress for increased funding, the unanswered ques- tions about complex issues affecting the FAA's performance listed pre- viously make it difficult to be assured that these efforts will succeed. The major reports examining the FAA's difficulties over the last few years agree that reform is needed in budget, personnel, procurement, and man-
266 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINcE DEREGULATION agement (NAPA 1986; ASC 1988a, OTA 1988). The changes proposed in these studies to respond to these problems, however, have varied fairly widely, from as little as fixing the tenure of the Administrator to completely reforming the FAA into a more independent governmental enterprise. (The past studies and the options for reform are reviewed in Appendix B). Some reforms have been advocated that would involve relatively little change in the FAA's institutional status, which appears to be the direction pre- ferred by the current administration. Whether the FAA will be more effective in the years ahead, however, is brought into question by the effects the budget deficit will have on its staffing needs, the lack of flexibility in pay to attract needed technical and managerial personnel, cumbersome procurement procedures that will weaken the ability of the FAA to modify the NAS Plan to meet capacity needs, and the FAA's vulnerability to unproductive micromanagement. These problems suggest that institutional reform should be considered seriously and appropriate action taken; the pros and cons associated with the various options are discussed in Chapter 8. NOTES The administration's goal was to cover 85 percent of the total FAA budget from the Trust Fund because it estimated that 85 percent of the FAA's costs were attributable to the private sector. Although this goal was not met, the proportion of total costs covered by the Trust Fund increased from about 50 to 75 percent. These policy changes better allocate costs to the private sector as a whole, but considerable cross subsidies still exist within the private sector. For example, most of the revenue for the Airport and Airway Trust Fund is paid by passengers in commercial aviation, whereas general aviation users appear to underpay their share (CBO 1988). The use of FPL controllers as an index of ATC performance has some notable shortcomings. The FAA does not believe that it is a useful index because it has never made staffing plans or management decisions on the basis of the number of FPL controllers. In addition, the definition of FPL has changed over time, which makes comparisons of staffing levels difficult to make. The committee agrees that the FPL designation is not adequate as a measure of FAA's capability. The available literature on FAA's ability, however, uses the FPL designation, and in the absence of other measures this designation is used to convey, in a general sense, the debate over FAA staffing levels. The FAA has been able to hire controllers at a higher grade in high-cost-of-living areas, but apparently this has not provided sufficient inducement to fill vacant positions. With recent changes in pay for controllers, many controllers will be able to earn more than $150,000 a year when overtime pay is included.
Constraints on the Performance of the FAA 267 REFERENCES ABBREVIATIONS ASC Aviation Safety Commission CBO Congressional Budget Office DOT U.S. Department of Transportation FAA Federal Aviation Administration GAO General Accounting Office NAPA National Academy of Public Administration NAS National Academy of Sciences OTA Office of Technology Assessment ASC. 1988a. Final Report and Recommendations. Washington, D.C., April, 51 pp. ASC. 1988b. Volume II. Staff Working Papers. Washington, D.C., April, 139 pp. Aviation Week and Space Technology. 1990a. ATA Urges Busey to Overhaul New York Air Traffic Control. April 9, pp. 60-61. Aviation Week and Space Technology. 1990b. FAA Faces Difficulty Finding Inspec- tors. July 2, p. 72. Burt, J. 1990. FAA Acquisition Philosophy. Journal of ATC, Dec., pp. 26-28. Celio, J. 1990. Controller Perspective of AERA 2. The MITRE Corporation, Mc- Lean, Va. CBO. 1988. The Status of the Airport and Airway Trust Fund: A Special Study. Washington, D.C. DOT. 1990. Moving America: New Directions, New Opportunities. Feb. Executive Office of the President. 1990. Budget of the United States Government, Fiscal Year 1991. U.S. Government Printing Office, pp. A933-A939. Executive Office of the President. 1991. Budget of the United States Government, Fiscal Year 1992. U.S. Government Printing Office, Part 4, pp. 894-897. Fearnsides, J. 1988. Development, Not Just Acquisition; Operational Capability, Not Just Technical Capability. Presented at the AOCI 41st Annual Conference, Wash- ington, D.C. FAA. 1987. Aircraft Cerufication Regulatory Program Management Efficiency Study. Office of Airworthiness, U.S. Department of Transportation. FAA. 1989. Aviation Forecasts, Fiscal Years 1989-2000. Report FAA-APO-89-1. U.S. Department of Transportation, March. Fortune. 1990. The FAA's Loose Grip on Air Safety. Oct. 8, pp. 85-97. GAO. 1987a. FAA Staffing: Challenges in Managing Shortages in the Maintenance Work Force. GAO/RCED-87-137. Washington, D.C. GAO. I 987b. Air Traffic System: Pilot Program to Contract Out Maintenance at Selected Facilities. RCED-87-I04BR. Washington, D.C. GAO. 1988a. Air Traffic Control: Continued Improvements Needed in FAA's Man- agement of the NAS Plan. GAO/RCED-89-7. Washington, D.C. GAO. I 988b. FAA Staffing: Recruitment, Hiring, and Initial Training of Safety Related Personnel. RCED-88-189. Washington, D.C., Sept. GAO. 1989. Aviation Safety: Serious Problems Continue to Trouble the Air Traffic Control Work Force. RCED-89-1 12. Washington, D.C. GAO. 1 990a. Air Traffic Control: Status of FAA's Effort to Modernize the System. RCED-90- 146F5. Washington, D.C., April.
268 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION GAO. I 990b. Air Traffic Control: Ineffective Management Plagues $1.7-Billion Radar Program. IMTEC90-37. Washington, D.C., May. GAO. 199 I. Aviation Safety: Limited Success Rebuilding Staff and Finalizing Aging Aircraft P/an. GAO/RCED-91-1 19. April. Johnston, W., and A. Packer. 1987. Workforce 2000: Work and Workers for the 2/st Century. Hudson Institute, Indianapolis, md. Mead, K. 1 990a. Serious Shortcomings in FAA 'sTraining Program Must be Remedied. Testimony prepared for Subcommittee on Investigations and Oversight, House Com- mittee on Public Works and Transportation. 101st Congress, 2nd Session. Com- mittee Print 101-74. pp. 72-85. Mead, K. 1990b. issues Related to FAA's Modernization of the Air Traffic Control System. Testimony prepared for Subcommittee on Aviation, House Committee on Public Works and Transportation. 101st Congress, 2nd session. NAPA. 1986. The Air Traffic Control System: Management by a Government Cor- poration: A Study for the Air Transport Association of America. Washington, D.C. NAS. 1980. Improving Aircraft Safety: FAA Certtfication of Commercial Passenger Aircraft. National Research Council, Washington, D.C. OTA. 1988. Safe Skies for Tomorrow: Aviation Safety in a Competitive Environment. Congress of the United States, July. Oster, C., Jr, and C. Zorn. 1987b. Airline Deregulation: Is it still safe to fly? In Proc., Transportation Deregulation and Safety, Northwestern University, Evanston, Ill. U.S. Congress. House. 1989. Committee on Public Works and Transportation. Con- cerning the Status of the Air Traffic Control System. H. Rept. 101-12, 101st Congress, 1st Session, pp. 204-227. U.S. Congress. House. 1990. Report to Accompany HR 5229. Department of Trans- portation and Related Agencies Appropriations Bill, 1991. H. Rept. 101-584, 101st Congress, 2nd Session. Washington Post. 1990. Controller Retirements Worry Airlines: Senior Workers' Departures Sparked by Possible Shift in Pension Rules. Sept. 6, p. A25.
PART IV Summary
Conclusions and Recommendations G overnment economic regulation of private enterprise has grown during most of this century. Although such regulation was justified initially because it was believed to be in the public interest, subsequent studies by economists and political scientists found that regulation of many industries actually raised prices and limited choices for consumers. A burgeoning body of research reached similar conclusions, and by the early 1970s, proposals to deregulate the transportation, communications, and financial services industries were gaining momentum. A wave of regu- latory reform proposals gathered strength during the Ford administration, crested during the Carter administration, and flowed through the Reagan administration. Occasionally fierce opposition to these reforms by the affected industries was swept away by a flood of administrative actions and congressional legislation. This reform movement removed restrictions on entry, pricing, and competition in several industries but did not specify completely the appropriate governmental role to maintain competition and protect consumers; this role is still being determined in policy and practice. Regulatory reform in the aviation industry, one of the first industries to be deregulated, has been controversial. As the 1980s ended, authors of newspaper and magazine articles, expressing alarm about rising prices and delays as well as voicing concern about safety, referred to the deregu- lated iiidustry as the "frenzied skies." Some observers of the industry, noting increased airline bankruptcies, perceiving poorer service, and fear- ing decreased safety, believe that deregulation has been a failure. Several studies found quite the opposite, however, and reported lower prices, improved service, and no discernible effect on safety. The findings and recommendations of the study committee on air passenger service and safety since deregulation are summarized in this chapter. 271
272 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION CHANGES IN AIRLINE OPERATIONS AND FINANCIAL PERFORMANCE Deregulation transformed the management of commercial airlines. During the 40-year period in which commercial carriers were regulated by the Civil Aeronautics Board (CAB), managers were judged by their abilities to provide service and influence the regulatory process. The industry was very stable. Sixteen carriers were "grandfathered" into trunk lines in 1938; 11 remained in 1978 (this number includes Pan American, which had not been classified as a domestic trunk line because it was not allowed to compete in domestic markets). Profitability, however, was lackluster. The relatively modest profitability of airlines under regulation coupled with the expectation of increased competition with deregulation raised concern about the ability of the industry to attract sufficient capital to reinvest for long-term growth. Management Innovation With deregulation, airline managers began to be judged by the same criterion as managers in private enterprise: the provision of service at a price consumers are willing to pay. The 1980s ushered in a period of intense competition. During the early years of deregulation many new carriers started competing for service, and carriers theretofore limited to charter or intrastate service also joined the fray. After the confusing effects of the 1979 fuel shortage and subsequent price shocks, a severe recession, and the firing of much of the air traffic control (ATC) work force in 1981, intense fare wars broke out. Air fares fell dramatically. Although con- sumers may regret the loss of the remarkable bargain airfares of that period, many carriers were offering prices below costs, and many new entrants and some holdover airlines failed. By 1988, a decade after the beginning of formal deregulation, the five largest firms carried 74 percent of revenue passenger miles, up from 69 percent in 1978 and still growing. (As dis- cussed in the next section, however, competition among the remaining firms, though fewer in number than in 1984, remains intense in many markets.) Several features of airline operations that are now taken for granted, such as nationwide hub-and-spoke networks, complex discount fares, com- puter reservation systems (CRSs), and frequent flier programs, are man- agement innovations that were developed or more fully exploited during
Conclusions and Recommendations273 this competitive struggle. Other management innovations that are less visible to consumers include extra bonuses given to travel agents to in- fluence their recommendations of airlines to consumers and closer affil- iations between carriers that serve national or regional markets and small carriers that serve a few cities. These service innovations (combined with freedom of entry and exit) helped carriers better align service and price with cost, which has resulted in an airline industry more efficient than it was under regulation. Aircraft utilization and load factors have increased as a result of the rationalization of route structures, and many unproductive work rules have been elimi- nated. However, the industry is still in transition. For example, labor- management relations are still contentious at some airlines, as witnessed by the prolonged strike, bankruptcy, and ultimate failure of Eastern Air- lines in 1991; the past efforts of United Airlines' and to a lesser extent, Trans World Airlines' unions to gain control of the company from their current management. In addition, some major carriers are still working out complex and difficult mergers, and the debt taken on to finance some of these mergers has resulted in considerably more financial leverage than these airlines have had in the past. Financial Performance In recent years the net profit margin of the airline industry, which has never been particularly robust, has narrowed even further, thus raising concern about the ability of the industry to attract capital for reinvestment. In comparing financial performance between the pre- and postderegulation periods, the industry in some measures is not significantly weaker when account is taken of the run-up in fuel prices after 1979, the subsequent recession, and the air traffic controllers' strike of 1981. Nevertheless, on the whole, operating margins have weakened, real interest expense has grown, and returns on equity have declined. When compared with all manufacturing industries, the return to capital in the airline industry re- mains low and volatile. Difficulties with reliance on aggregate measures of financial performance such as these, however, limit the kinds of in- ferences that can or should be drawn from them. Three shortcomings should be noted: 1. Considerable variation exists within the industry, and the weakest carriers have skewed the trends downward.
274 WINOS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION The measures themselves have become increasingly difficult to in- terpret with the more sophisticated asset management practices that have developed. The important issue is the ability of individual airlines to attract sufficient capital, not how well aggregate measures of net profits compare with those of other industries. Variable Performance Part of the poor performance of the industry as a whole during the de- regulation era is attributable to the heavy losses of Eastern, Continental, and Pan American, all three of which had negative net profit margins from 1978 to 1988. (Eastern and Pan American also had negative profit margins in several years preceding deregulation.) Trans World Airlines has barely broken even since deregulation and also entered deregulation in a weak financial position. As of early 1991 Eastern had ceased operating, and Continental and Pan American, already weak before the fuel price shocks caused by the Persian Gulf crisis, sought protection from creditors in Chapter 11 bankruptcy proceedings. Trans World was threatening to seek Chapter 11 protection even as this report was being completed. Although the results for the industry as a whole are skewed downward by a few firms, it is useful to examine the financial performance of weak firms as well as strong ones because this sets the context for the pricing patterns in the industry. In recent years the weakest financial performers have also tended to be the pricing leaders, and the competitiveness of the industry, perhaps made more aggressive because of overcapacity, has meant that all carriers have attempted to match the discounts offered by the lowest-priced major competitors. As the weakest carriers either drop out of the industry or turn around their performance, the financial per- formance of the industry as a whole may improve. Limitations of Measures As airline managers have become more sophisticated in asset management, the financial indicators, accounting practices, and reporting systems de- veloped for the industry during the regulated period have become more difficult to interpret. A major concern about the industry, for example, is the increased financial leverage, measured with debt/equity ratios, and
Conclusions and Recommendations 275 how this debt burden might make carriers increasingly vulnerable during an economic downturn. The increased debt, and the revenues required to finance it, reduces capital available for reinvestment and contributes to lower net income margins. Whereas the bankruptcy of Continental, one of the most highly leveraged carriers, gives credence to this concern, it is increasingly difficult both to measure the extent of leverage and to determine, prima facie, whether it is beneficial or harmful to consumers. Many airline assets, particularly aircraft, have been obtained through both capital and operating leases since deregulation. The provisions in many of the operating leases, which allow carriers to return the aircraft on relatively short notice, have helped managements reduce fixed costs and thereby better shield returns from the effects of a downturn in demand. In addition, other airline assets that have genuine market value, such as long-term leases at key airports, slots, and aircraft delivery positions, are not considered assets in traditional financial measures. Because standard financial measures do not take into account increased management flex- ibility and do not include the value of certain intangible but real airline assets in measuring equity, the utility of trend data on debt/equity ratios is reduced. Although excessive. leverage is not desirable, leverage itself should not be perceived as negative. Leverage can focus management attention on the need to control costs. At the same time, excessive leverage can increase pressure to ignore the long term and may encourage initiation of or ac- quiescence in oligopolistic pricing behavior. Excessive leverage can also force sales of assets from leveraged to financially strong carriers, which lessens competition and leaves the government with a difficult dilemma. Ability To Attract Capital Throughout its history the airline industry has performed poorly on many financial measures; nevertheless, capital has continued to flow into the industry. Returns on equity and investment are key measures of a firm's ability to attract capital; these measures have improved slightly for the strong carriers since deregulation but have declined sharply for the weak carriers. Airline earnings are heavily influenced by the business cycle, which increases risk and affects the cost of capital. During much of the 1980s the basic business risk of the airlines (as measured by their asset betas) became more like that of the economy as a whole. After the mid-1980s, and especially after 1988, airline betas increased well above those of
276 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION Standard and Poor's 500 Industrials, apparently because of concern about the increased debt taken on to finance a major aircraft reequipment cycle and concern about increased debt to finance mergers. These market per- ceptions of increased risk, followed in late 1990 by a run-up in fuel prices and a downturn in the economy, have heightened concern about the fi- nancial posture of the industry, suggesting that more than just the weakest carriers will have trouble attracting investment. It is impossible to predict with certainty how the ongoing industry shakeout will affect competition and balance sheets. The loss of one or two carriers, for example, may reduce excess capacity in the industry and improve cash flow and profits for the remaining carriers. The loss of too many carriers, however, could reduce the amount of competition below a level necessary to discipline pricing. Although having three carriers in an individual market likely ensures adequate competition to discipline pricing, having only three carriers providing nationwide service would probably not be adequate. If as many as five or six major carriers survive the pending shakeout, the committee believes that pricing will remain competitive. The number of carriers is important to competition, but no less im- portant is the ability of the remaining carriers, both major and niche carriers, to compete with one another in the maximum number of mar- kets. The experience of consumers during deregulation is reviewed in the next section; impediments to competition are reviewed in the sub- sequent section. PASSENGER FARES AND AIRLINE SERVICE Average fares have declined in real terms since deregulation, and total passenger trips have increased sharply. Although it is difficult to predict how prices and service would have unfolded had CAB regulation contin- ued, some insight into likely fare trends can be gained by making ad- justments to the fare formula that the CAB developed in the late 1970s for monitoring fare changes. This fare formula, known as the Standard Industry Fare Level (SIFL), was based on the predominant unrestricted coach fare in 1977, and after about 1978 the formula was used by the CAB to set allowable airline fares. (Although the SIFL is no longer used in establishing allowable fares, the U.S. Department of Transportation (DOT) continues to estimate it.) The SIFL has increased more than 75 percent since 1977, far faster than the average yield, which has increased about 50 percent (Figure 3-3). The carriers have been able to absorb some
Conclusions and Reconunendations 277 of the increase in cost through increased productivity and, perhaps to a lesser extent, through reduced profits. The SIFL is useful as an indicator of how yields might have changed under continued regulation. The SIFL itself, however, is influenced by the moderation in some costs, which can be attributed directly to dereg- ulation. Hence the trend in the SIFL understates the actual fare levels that would have prevailed under continued regulation; the SIFL thus provides a conservative estimate of consumer benefits since deregulation. In calculating the real yield since deregulation, the SIFL gives a better measure of cost inflation in examinations of aggregate price trends in the airline industry than does the gross national product (GNP) implicit price deflator because it accounts for the share of cost attributable to fuel, aircraft, and labor, and how the trends in these costs affect the overall cost of airline operation. Fuel costs, in particular, have a much greater influence on total input costs to airlines than to the economy as a whole. In contrast with the SIFL, the GNP price deflator can be used as a measure of how yields compare with inflation as related to consumer goods and services generally. As shown in Figure 3-3, the GNP price deflator increased more slowly than the SIFL during the early years of deregulation. This is to be expected because the fuel price shocks of 1979 to 1980 dramatically increased industry costs. The subsequent moderation in fuel costs is reflected in the declines in the SIFL between 1982 and 1987. By 1989 the GNP deflator had increased slightly more than the SIFL; it increased 88 percent from 1977 to 1989, whereas the SIFL in- creased 75 percent. In the analysis of fare trends that follows, changes in yields are calculated on the basis of the GNP deflator. As can be seen from Figure 3-3, this results in a somewhat greater decline when the average real yield is estimated than would result from using the SIFL. Given that the difference in the two indexes is not great in recent years and that growth in the SIFL has been moderated by deregulation, any overstatement of the benefit in real yields when using the GNP deflator is likely to be small, if not negligible. The average yield, when adjusted by the GNP deflator, fell 20 percent between 1977 and 1989. Travelers making one-way trips of 1,000 miles or more have been the primary beneficiaries (Figure 3-8). When both market distance and traffic density are considered, it is evident that average real yields have fallen on the order of 10 to 25 percent for trips of 1,000 miles or more (Table 3-2). In the average short-haul markets, 500 miles or less, average yields in all but the largest markets have increased far faster than the rate of inflation. In the most heavily traveled short-haul markets (800 or more passengers per day) yields have fallen 3 percent,
278 WINDs OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION but for other short-haul markets, yields have increased about 20 to 30 percent. Most passengers have benefited because the densest short-haul markets alone account for 28 percent of all trips, whereas the short-haul markets that show substantial increases account for only about 3 percent of total trips (Table 3-2). The average market with a distance of 500 to 750 miles has experienced an increase in yields of about 15 percent more than inflation; travelers in these markets account for about 14 percent of all trips. For trips of 750 to 1,000 miles, average yields have increased about the same as inflation. Altogether, 79 percent of trips occurred in markets that experienced real yield declines since deregulation. As described in the next two sections, the amount of competition in city-pair markets appears to be an important determinant of the trends in average yields, but some of the trends in yields can be interpreted as a realignment of revenues with costs. During regulation the CAB fare policy explicitly rewarded carriers for the longest-distance flights by allowing the most profitable fares. Shorter-distance flights serving small commu- nities, however, were kept at or below cost by the CAB fare policy, which resulted in an internal cross subsidy. As carrier pricing was rationalized with costs, shorter-distance fares rose, longer-distance fares declined. Benefits of Competition The overall level of competition in the airline industry has increased substantially since 1978. At that time only 20 percent of city-pair markets were served by three or more competitors compared with 40 percent by 1988 (competitors are defined as those having 10 percent or more of the market). The markets with three or more competitors in 1988 served about 65 percent of passengers. Competition was even greater in the mid-1980s when more than half of city-pair markets had three or more competitors. The level of competition in the mid-l980s may not have been sustainable: many carriers failed or were acquired by or merged with other carriers. Since the mid-1980s the industry has become more concentrated and the overall level of competition has decreased, but as noted above, the trend in yields has remained below the trend in costs and most passengers are traveling in multi-carrier markets. The amount of competition in city-pair markets is a major determinant of the trend in average yields. For example, the most heavily traveled short-haul markets have three or more competitors, and contrary to other short-haul markets, the average yield in these markets has decreased since
Conclusions and Recommendations 279 deregulation. Most other short-haul markets, however, are served by only one or two carriers, and yields in these markets have tended to increase the most. For trips of 1,000 miles or more, for which the average markets have experienced the largest decrease in yields, most city-pair markets have three or more competitors. Hub-and-Spoke Networks and Competition The enhancement of hub-and-spoke networks following deregulation helps explain why competition in markets of roughly 800 miles or more has increased substantially, whereas competition in most shorter-distance mar- kets has not. Major carriers have used the service efficiencies of hub-and- spoke systems to build extensive networks, many of which extend nationwide, to meet consumer preferences for single-carrier service and to attract passengers with their extensive route systems. When a traveler wishes to fly from coast to coast, for example, he or she can choose among several airlines that offer one-stop service fosa_trip that could entail a connection at any of a number of hub airports. As hub systems have increased in number and extent, consumers have gained more choices for trips long enough to require a connection at a hub, especially when their journeys are 800 miles or longer. As many small airports have become spokes for more than one airline, passengers from many smaller com- munities have gained choices among carriers for longer-distance trips. For air service directly to and from a city with a hub airport, however, travelers tend to have fewer choices among carriers, particularly when one airline provides most of the service to the hub city. In general, for city-pair markets in which one of the cities is a concentrated-hub market, far less competition exists than in other markets. Altogether, among the 795 city pairs sampled for this study, for the city pairs in which one city is a concentrated-hub market, 43 percent are single-carrier markets com- pared with 13 percent of nonhub city pairs, and average yields are also higher in concentrated-hub city-pairs than in all other markets (Table 3- 7). As discussed next, however, concentration at hub markets does not fully explain the higher yield in these markets. Some of the spoke cities in these markets are not large enough to sustain service by more than one carrier, and, combined with their short-haul character, they may cost more to serve than do larger markets. Some of these markets receive nonstop service to major hub airports as a result of the extension of hub-and-spoke networks, which airlines would not be able to justify without the efficiencies of hubbing. Many nonhub city pairs
280 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION connect markets in coastal cities that are likely to have a larger share of discretionary tourist travel and hence more discounted air fares, than the typical hub city pair. In addition, hub city pairs have considerably more frequent nonstop service than nonhub city pairs. Thus travelers in these markets are offered a different, and more frequent, level of service than travelers in nonhub city pairs. Business travelers value the increased sched- ule frequency that hub and spoke networks provide and appear to be willing to pay a premium for more frequent service. Service and market dimensions help explain part, but not all, of the higher yields earned by carriers that dominate their spoke markets. Vir- tually all studies of pricing at hub airports have shown that a hub "pre- mium" exists, and estimates of prices for trips that involve a concentrated hub range in round terms from about 10 to 30 percent higher than trips that do not involve a concentrated hub. The hub premium, however, is the result of a complex set of influences, and isolating the contributing role of these influences was beyond the scope of this study. Most scholars believe that some component of the hub premium is accounted for by market power. The committee recognizes that the magnitude of the hub premium attributable to market power is difficult to measure and is there- fore debatable. However, given the reduced amount of competition in concentrated-hub markets, the carriers dominating these markets are in a position to exercise some degree of market power, and this position could be enhanced by continued concentration in the airline industry. Not all single-carrier markets are without some sort of competition, however, because for the shortest trips (less than 250 miles), many travelers choose to travel by surface transportation. Passengers traveling in markets dominated by a single hub carrier for trips of more than 250 miles account for a relatively small share of total trips, approximately 7 percent. Rural and Small Communities One of the explicit goals of deregulation was to allow carriers to move toward a market-oriented pricing and route system. In the regulated era, service to small communities was ensured by a combination of direct public subsidies and by cross subsidies within airline operations. In other words, the profits from long-haul markets were used to offset the higher cost of providing service to small communities with relatively little de- mand. Allowing the market, instead of subsidies, to drive carrier behavior was sure to affect the level and price of service to the smallest communities.
Conclusions and Recommendations 281 The smallest communities, though many in number, account for about 4 percent of total passenger enpianements. After deregulation, the smallest communities that had been receiving scheduled commercial traffic did tend to lose service provided by jet carriers, but this service was replaced typically by fast-growing commuter airlines. Carriers were already withdrawing service from the smallest com- munities before deregulation, but the freedom of entry and exit provided during deregulation probably hastened this trend. A considerable degree of variation exists among the smallest communities, however. To estimate the effects on the smallest communities, the experience of the airports that had been receiving subsidized service before deregulation was ex- amined because these communities were believed to be most in jeopardy of losing service. Service For a sample of 100 cities that received subsidized jet service under the Section 406 subsidy program, almost all of which became eligible for subsidies to ensure commuter service under the Essential Air Service (EAS) program, the median city experienced a 29 percent increase in service (daily departures). (EAS-eligible communities in Alaska, Hawaii, and Puerto Rico are not included in this sample because of their unique characteristics.) This increase in service is largely due to the greater flight frequency provided by commuter carriers operating smaller, less costly aircraft than the previously subsidized jet operators. About one-third of communities did receive somewhat less service after deregulation. The growth in service at other airports within driving distance, however, bal- ances most of these losses. If the airports within 100 miles of the subsidized airports are included, the median area has experienced an 85 percent increase in service, and only 9 percent of these areas received less service in 1988 than in 1977. Even as service was improved, federal subsidies to ensure air service declined from $72 million in 1978 to $28.4 million in 1988. (Even more money was probably saved by the carriers and their passengers because internal cross subsidies were eliminated.) Part of this decline in the public subsidy is attributable to the more efficient form of subsidy provided by the EAS program. Under the previous subsidy system, the Section 406 program, subsidies were provided virtually on a cost-reimbursement basis, without clear criteria for the level of service to be provided. In contrast, carriers wishing to provide EAS service must bid competitively for the
282 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION service. The EAS program also sets specific criteria for service, including more direct service to nearby hubs, whereas under the previous subsidy system carriers were required to serve the smallest communities but had considerable latitude in how they would be connected to their route system. As a result many small communities received service, but several small communities were often linked to a linear route. Trips on such routes required multiple stops before a major city was reached. The current program has clearer requirements for providing service to major hubs, which improves the timeliness and quality of service. Fares Less is known regarding the pattern of fares in small communities. (Fares were expected to increase after the phasing out of cross subsidies and reduced direct subsidies.) The data on fares for travelers from the smallest rural communities are sparse; commuter carriers are not required to par- ticipate in the DOT's quarterly sample of airfares, and passenger traffic in most small-community markets served by carriers that do report is light. For these reasons a representative sample of fares charged to passengers in the smallest markets traveling between these small cities and a hub airport is not available. For travelers making connections at a hub and traveling to another major city, more data are available and indicate that travelers in these city-pair markets pay roughly the same fares as travelers in other markets of similar distance; these travelers constitute the great majority of enplanements from small communities. BARRIERS TO COMPETITION Although the concentration of the industry since the mid-1980s has not overly reduced competition, at some point further concentration could lead to diminished competition below a level adequate to discipline pricing. Some characteristics of the industry now reduce or limit competition and the prospects for new entry, and deserve close scrutiny. Risk of Entry Some proposals for deregulation were accompanied by the belief that carriers would have relatively free access to markets because of the mo-
Conclusions and Recommendations 283 bility of airlines' chief assetsâaircraft. Carriers dominating individual markets would not charge monopoly fares, according to this theory, be- cause of the ease with which a competitor could enter at reduced prices and contest the incumbent carrier (hence the name contestability theory). Thus the mere threat of entry was expected to discipline pricing. Con- testability theory requires low costs of entry and exit. Access to many markets, however, is far from costless, partly because of the difficulty of obtaining terminal space at many hubs, but mostly because of the risk associated with competing with a major airline at one of its hubs. A competitor that wishes to challenge another carrier at its hub is faced with a considerable financial outlay. The cost of providing a competitive level of service at a hub is substantial; it requires outlays for advertising, personnel, and aircraft operations during start-up when the competitor attempts to win business away from the hub carrier. The risk of being unable to recover these outlays is the largest single deterrent to entry at hub airports. Indeed, instead of engaging in head-to-head competition at another carrier's hub, most carriers have opted for creating their own hubs at formerly underused airports. Hence it may not be unusual for a .dominated-hub airport, such as St. Louis-Lambert International, to have excess gates. Providing sufficient capacity to allow additional spoke op- erations may not significantly reduce the competitive advantages of the hubbing carrier but will provide a marginal improvement in competition by adding more choices for passengers, especially leisure travelers. Airport Terminal Capacity Limits In addition to underestimating the nonrecoverable costs of entry and exit, contestability theory underestimated the difficulty of obtaining favorable gates and terminal space at strategic hub and nonhub airports. Part of the difficulty traces to the way that airports are financed, the methods for which were developed before deregulation. Although every airport has a somewhat different arrangement, most are financed with long-term bonds in which the revenues earned from the airlines serving the airport are used to guarantee to the bondholders adequate revenues to cover the operation of the airport. As incentives to make these long-term commitments, car- riers were given leases providing exclusive use of some facilities, usually gates and first-class lounges or frequent traveler membership clubs. In some cases the airlines have veto power over a capital expansion that would increase their costs if a majority of the airlines signatory to the lease refuses to support the expansion. Carriers that entered deregulation
284 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION with a substantial presence in airports that were important traffic generators or that would become hubs and carriers that gained control of unused airport capacity early in the deregulated period gained a competitive ad- vantage that was often, in effect, a residual effect of the regulated period. Air carriers that have excess gates are in a position to charge lease costs for them that reduces the marketing advantage of low-cost carriers. Prudent fiscal management by air carriers and airport managers will tend to keep excess capacity at airports to a minimum because airlines do not wish their passengers to pay for facilities they will not use. For hub airports, once a carrier builds its hub, maintaining sufficient idle capacity to support a second carrier that wishes to build a hub will not ordinarily be practical (although this is not the primary deterrent to establishing a competing hub operation). At many major airport terminals, few gates are available on short notice; entry on almost any scale requires either subleasing at a potentially disadvantageous rate or financing new construction. Incumbent airlines cannot stop a competitor wishing to enter a market from financing new construction for terminal space, and a great deal of such entry has occurred throughout the deregulated era. New construction projects, however, typically require 2 years from the planning to the completion stage, and the new entrant will be paying for the space at a higher cost than the incumbents, who typically have leases with more favorable terms (because they were financed in the past with inflation- eroded dollars). Though creating a disadvantage for entry, this phenom- enon is not unique to airline competition; it is common in durable goods industries. The length of time required to enter and the higher cost borne by the new entrant, however, makes it more difficult for potential com- petitors to enter markets. Relatively free from the threat of new entry for this period, the incumbent carriers are in a better position to exercise market power in some markets. Slot-Controlled Airports Since 1969 carriers wishing to serve four major airports, Chicago O'Hare, New York Kennedy and LaGuardia, and Washington National, have had to have permission to take off or land during a specified time (referred to as slots). The FAA imposed slot control in an attempt to reduce severe congestion and delay, particularly at O'Hare. Initially the slots were al- located by airline scheduling committees. In 1986 the FAA began to allow carriers holding slots to sell or lease them, but did not require that sales
Conclusions and Recorn,nendaiions 285 be made by auction. As a result, the slots have been traded privately, often without the knowledge of some potential bidders, and, more recently, have been leased for short terms to smaller carriers that are not in direct competition with the slot holder. The slot allocation system has not served competition well. Although the slots at the New York airports and at Washington National are held by a fairly large number of airlines, American and United have increased their share of the slots at O'Hare from 65 percent in 1986 to more than 75 percent, further strengthening their position at a key hub. More im- portant, however, new entrants have had difficulty acquiring slots through purchase. If the FAA had not occasionally reallocated unused slots for new entrants, new entrants would probably have been unable to gain the amount of access they currently have to these important markets. In 1990 Congress required the FAA to initiate a rulemaking proceeding to recon- sider existing slot allocation methods. In this regard, the committee be- lieves that rationing scarce resources with a price mechanism would be more efficient than any administrative scheme (see recommendations fol- lowing the section on airport capacity). Airline Marketing Strategies As discount fares have proliferated and as hub-and-spoke networks have expanded to provide many alternative routes for most trips, the complexity of air travel has increased. Since deregulation, therefore, travel agents have assumed a much larger role in assisting consumers in the search for the flight and fare that will best suit their needs. In 1978 agents booked only 38 percent of all tickets as compared with 80 percent in 1988. CRSs Advances in CRSs have helped make the transition possible to the more complex array of information about fares and flights that carriers offer to attract passengers. CRSs handle millions of fare changes and book- ings each day, and 92 percent of all domestic bookings are made through a CRS. Four major systems operate in the United States, and each CRS lists all the flights and fares of all but one domestic airline (Southwest Airlines). The two largest systems (Sabre, which is owned by the com- pany that owns American Airlines, and Apollo, the majority of which
286 WINDS OF CHANGE: DoMEsTIc AIR TRANSPORT SINCE DEREGULATION is owned by the company that owns United Airlines) handle 70 percent of all bookings. Despite the technological advances they represent and the efficiency gains they have made possible, CRSs are controversial. The initial debate focused on the way in which the airline that owns the CRS (referred to as the host) was able to bias the screen display provided to travel agents in favor of its own flights. Most travel-agent bookings are taken from the flights presented on the first screen, many from the first few lines, and hence any bias could distort the information provided consumers. Federal regulations appear to have greatly reduced screen bias. Despite the success of federal regulation with respect to screen bias, CRSs provide the airlines that own them with considerable competitive advantages: They allow the host to earn substantial, some would say exces- sive, booking fees (in most cases twice as high as average cost) from its competitors; It appears that they may be used by travel agents to favor the host in booking travel and thus earn substantial profits for the airlines owning them; and They give the host an advantage in managing the levels of discounts it must offer and the number of seats to be sold on discount. The cost of developing a CRS is such that only the largest airlines, or consortia of airlines, can afford to do so and is effectively a barrier to entry for airlines that might otherwise wish to enter the market. Nonairline potential investors face the additional entry disadvantage that they cannot easily benefit from airline incremental revenue generated by CRSs. Travel Agent Incentives Although most business travelers who fly with some frequency tend to have a preference for an airline, some business travelers, perhaps one- fourth, and many leisure travelers, perhaps half, apparently do not have a strong preference for a particular airline. Hence travel agents are in a position to sway consumer choices. Airlines use a variety of strategies to influence agents to sway passengers' flight selections. Carriers that own a CRS often provide favorable rates on a multiyear lease of the equipment and software needed to operate the CRS and provide training at low or no cost. Almost all carriers try to influence agents by providing bonuses
Conclusions and Recommendations 287 on top of the standard commission when they book a certain number or share of passengers on the carrier. The bonuses are typically about 2 percent of ticket revenue; the average agent commission for booking a ticket is 8 to 10 percent. These bonuses, referred to as travel-agent com- mission overrides, have influenced agent booking patterns. Installation of one carrier's CRS in an agency does not preclude another carrier from offering overrides to travel agents, and in some markets the CRS-owning airline is not the airline with the largest volume with a travel agency. Any carrier can offer overrides. Carriers that own CRSs and offer overrides, however, are at a competitive advantage. Carriers that have provided a CRS to an agent and who offer commission overrides appear, in the DOT's conservative estimates, to earn at least 15 percent more bookings from the agent than they otherwise would. Estimates of the incremental revenues earned made by the vendors themselves typically are far larger than 15 percent. Few consumers are aware that most travel agents have these special relationships with carriers. Travel agents state that their primary mission is to find the best deal for their customers, that they depend on repeat customers to stay in business, and that many fares for the same trip are equivalent. Hence their recommendation of a preferred carrier does not necessarily increase consumers' costs. An agent wishing to earn an over- ride, however, may not be as assiduous in finding the lowest discount for a price-conscious customer when a discount fare (albeit not as large a discount) is available from a preferred carrier who is offering commission overrides. This works to the disadvantage of low-cost, new-entrant airlines that do not own CRSs and attempt to compete by providing the lowest fares. Frequent Flier Programs Frequent flier programs have become widespread and popular with many frequent travelers, many of whom travel for business. Frequent flier awards are designed to secure the travel of individuals, usually business travelers, who are less willing to take advantage of the restrictions applied to the deepest discount fares (which usually require a Saturday stay over). These programs are designed to influence consumer choices, and they do. Be- cause they provide awards to the traveler instead of to the company paying for the ticket, travelers may choose airlines on the basis of their frequent flier program and not on the basis of the fares their employers pay. This increases the cost to the business paying the airfare. Frequent flier pro-
288 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION grams give major airlines with large national networks an advantage over smaller competitors or new entrants. Major carriers structure rewards on a nonlinear basis to induce travelers to concentrate their flights on a single airline. Their large networks make it easier for the traveler to stay in one program and earn rewards. Small carriers, without large networks, cannot offer multiple opportunities to earn mileage or attractive bonus trips to resorts or international locations that are not part of their route system. (Although cooperative arrangements could, in theory, develop between small airlines, they would likely be cumbersome and short-lived.) As a result, the frequent business traveler, a major source of airline revenue, can be lured away from lower-cost flights by the prospects of earning bonus trips or upgrades to first class. Any new entrant faces the task of luring business travelers who do not pay the cost of their tickets away from airlines with which they have a substantial frequent flier account balance. Code Sharing Deregulation accelerated the trend of jet service withdrawal from small markets. Small regional and commuter carriers (those flying aircraft with fewer than 60 seats) replaced jet carriers in most markets and began providing feeder traffic into the major carriers' hubs. During the mid- 1980s the majors recognized the value of this traffic and, to better inform travel agents of coordinated schedules and joint fares, began offering to share their CRS code with commuters that would coordinate their schedules with the larger airlines. The larger carriers also received other benefits. Most commuters were offering service in aircraft that are much less ex- pensive to operate. As a result, the feeder traffic could be served by a commuter at much lower cost than by the major carrier itself. Relationships to coordinate schedules, baggage handling, and fares be- tween large airlines and smaller ones predate deregulation, but with the proliferation of CRSs the code-sharing arrangement became an efficient signal that such relationships exist. The effects proved substantial, re- sulting in considerable disadvantages to commuter airlines that did not have a code-sharing partner. Currently the top 50 commuter airlines carry 92 percent of commuter traffic, and almost all have a code-sharing ar- rangement. Twenty of them are owned in whole or in part by a larger airline. Cooperative agreements between large and small carriers predate de- regulation and the development of CRSs and offer considerable benefits
Conclusions and Reco,nrnendaiions 289 to consumers. The coordination of schedules and baggage handling and the extension of discounts through joint fares meet consumer preferences for single-carrier service and help integrate travelers from small com- munities into the national system. Code sharing appears to offer other incidental benefits to consumers, because the larger carrier may help finance new aircraft, share terminal space, and impose minimum service requirements on their code-sharing affiliates. Code sharing also has implications for airline competition and prices charged to travelers from small communities. The CRS algorithms give preference to code sharing over interline connections in displaying con- nections, even in cases in which the interline service better serves pas- senger cost and schedule needs. Agents have to work harder to find these connections if they do not entail a code-sharing agreement, and after viewing two or three screens, or in the interest of earning a commission override, may well conclude that the code-sharing flight is best. Code-sharing arrangements do not always inhibit competition for trips beyond a hub because many small cities are served by commuters offering service to alternative hubs, but for trips to a specific hub, code sharing may well inhibit competition. The result is at the least to weaken the ability of other commuters to compete in local markets with code-sharing commuters and at the worst perhaps to constitute an absolute barrier to entry in some local commuter markets. The relationship also gives the major airline more control over feeder traffic, thereby strengthening its hub against competition from other carriers. Summary Being able to exploit all these competitive advantages is not a necessary condition for survival in nationwide competition. For example, some new entrants, which do not have CRSs, have survived and even prospered in some years by finding niche markets. The existence of these barriers, however, appears to partly explain how the airlines that existed before deregulation, with their higher cost structures, managed to compete suc- cessfully with lower-cost new entrants. (These "holdover" carriers also lowered their costs somewhat by negotiating more productive work rules, lower wages, and otherwise rationalizing their service.) Public policy changes to reduce all these barriers to competition without posing more costs than benefits, however, are not readily apparent. Efforts to reduce hub dominance, for example, could reduce the benefits of hub- and-spoke systems without guaranteeing increased competition and lower
290 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION fares. As indicated in the committee's recommendations (presented be- low), some incremental adjustments to policy governing CRSs and travel agents would probably reduce their negative effects on competition. Mergers and Acquisitions The number of mergers and acquisitions permitted during deregulation, especially after 1984 when the DOT gained antitrust authority, allowed the industry to return to the level of concentration that existed in 1978. Not all these mergers have proven successful, however, and some of the largest carriers created during this period lost market share. Some of the mergers permitted since 1978 may have actually enhanced competition between hub networks, specifically those that were "end-to-end" in char- acter. Such mergers allowed carriers serving different regions to combine into firms large enough to compete with the other major carriers in longer distance markets. The oligopoly that has reemerged through these mergers and acquisitions is still quite competitive in the markets in which most passengers are traveling, but continued concentration is likely to occur as a result of the financial weaknesses of some firms. Hence the U.S. Department of Justice (DOJ), which has had antitrust authority since 1989, faces some difficult choices. For example, mergers or acquisitions may be proposed that would increase competition in nationwide markets but may result in a carrier more able to dominate a hub airport. Recommendations To Enhance Competition Although the aviation industry has not evolved as some proponents of deregulation predicted, the industry is more competitive and efficient than before, and the majority of consumers are benefiting from improved ser- vice and lower fares. Barriers to entry have developed, and because of the lackluster profitability of the industry as a whole and the weak position of some firms, further consolidation of the industry appears likely. As discussed in the conclusion to Chapter 4, few measures to eliminate po- tential barriers to entry are evident, however, that would not impose more costs than benefits. In the judgment of the committee, impleiiientation of the following recommendations would improve the performance of the marketplace, with attendant benefits for consumers, without undermining the improvements wrought by deregulation.
Conclusions and Recommendations 291 Antitrust Policy A more forthright federal antitrust policy is required than has been prac- ticed during much of the 1980s. The DOJ has been more active on antitrust issues than was the DOT. The study committee supports the more activist role adopted by the DOJ, but recognizes that further mergers or asset acquisitions leading to greater industry concentration are likely. Given the prospects for further concentration, the committee recommends that the DOJ oppose mergers or acquisitions in which the carriers offer substantial parallel service in city-pair markets or share a hub airport. However, the DOJ should not necessarily oppose mergers or asset acquisitions of car- riers with complementary or end-to-end routes. Such mergers or asset acquisitions often may not harm competition. Given the importance of having at least three effective competitors in city-pair markets involving a connection at a hub, the maintenance of this level of choice for consumers should be used as a test for the adequacy of competition in "over hub" traffic when merger and acquisition proposals as considered, including acquisitions of individual assets, such as gates. CRSs CRSs offer considerable competitive advantages to the carriers that own them. These advantages would be reduced by improving the competition between the existing CRSs. Although carriers compete vigorously with one another in the selling of these systems, once an agent is under a lease, switching systems is not practical for the duration of the lease. Lease clauses prohibit the use of the leased equipment to access other CRSs, and few agents can afford to operate more than one CRS. The committee recommends that travel agents be allowed to use their own equipment or desk equipment leased from the host carrier to access multiple CRSs. This would require a DOT regulation to prohibit contract or lease terms that restrict the ability of travel agents to use equipment that is connected to one CRS to switch freely among CRSs, along with continuing the prohi- bition against display and function bias and extending it beyond CRS owners to any software used in the interface allowing multiple access. Implementation of this recommendation would improve service to the traveling public. The advantages that CRS-owning carriers have over their competitors would be reduced by this "multiaccess" provision because agents could readily access alternative systems to compare simultaneously in an integrated display the best offerings for their clients and to determine
292 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION the latest fares and seat availability (assuming that the existing prohibition on bias would be extended to others as it became possible to access CRSs through other software). Airlines that do not own a CRS could encourage travel agents to book their flights through CRSs that charge them lower booking fees. The system would be most competitive if the booking fees were charged to travel agents, who could switch between systems, instead of to airlines that for competitive reasons cannot afford to opt out of participating in individual CRSs. Consumer Information Travel agents are in a powerful position to influence consumer choices, and consumers are typically unaware of the financial relationships that exist between many travel agents and airlines. At present, when an agent tells a customer that a certain flight most closely matches the customer's preferences, the customer has no way of knowing whether the agent's recommendation is influenced by a desire to earn an extra commission. The committee recommends improving consumer information by requiring agents to disclose the incentive commissions (commission overrides) that they receive from carriers. This could be accomplished by a DOT regu- lation that has the effect of making consumers aware of the existence of override programs and the identity of the carriers favored by the agent. COMMERCIAL AVIATION SAFETY Commercial aviation safety has steadily improved throughout the post- World War II era (Chapter 5). The fatal accident rate has declined since deregulation, and, during two years since 1978 (1980 and 1984) no fatal accidents occurred in regularly scheduled passenger service provided by large airlines that offer jet service. In addition, in 1986 only one person was killed in accidents involving jet carriers. The three safest years in the history of scheduled passenger service provided by large airlines occurred after deregulation. The changes ushered into the industry with deregulation could con- ceivably have increased risk for three reasons: (a) decreased emphasis on maintenance caused by increased pressure on costs, (b) increased reliance on flight crews with limited experience, resulting from increased entry and the booming demand for air travel, and (c) the inexperience of new-
Conclusions and Recommendations 293 entrant airlines. Studies in which various dimensions of these concerns were examined, however, have not established any statistically significant relationship between carrier actions as a result of these concerns and accident and fatal-accident rates. Concern has been expressed by many that the economic pressure on carriers brought about by deregulation may have increased risk, even if not yet manifested in increased actual accidents. Measures of risk have therefore been proposed, both to examine the deregulation record and to measure current trends. Reporting biases in nonaccident safety indicators such as near-midair collisions, however, prevent them from being used to evaluate changes in risk during deregulation. The FAA has begun an effort to develop safety indicators and improve the quality of existing data, which is likely to be a difficult and lengthy process. Carrier use of digital flight data recorders (DFDRs) appears to be a more direct and systematic way of using data to improve safety. Such data could be used to measure and evaluate air crew performance, a critical element of flight safety. DFDR programs are already in use by air carriers throughout the world. Adoption of such programs in the United States has been stymied by labor-management disagreements, but both sides appear increasingly to favor their adoption. The FAA has begun to work with both sides to establish arrangements that would permit DFDR pro- grams to be used in the United States. Their development may require Congress to protect the data to ensure confidentiality and the FAA to change federal regulations to ensure that the data would not be used for enforcement. For communities to which deregulation brought a change in air service from jet to turboprop service, there may well have been some increase in risk during the early years of deregulation. The withdrawal of jet service from small communities had been occurring for many years before 1978, but deregulation probably accelerated that trend. Fortunately, the differ- ence in risk between jet and turboprop service has narrowed considerably over time, partly because of more stringent safety regulations applied to nonjet equipment in 1978. By the 1990s, the fatal accident rate of com- muter carriers offering regularly scheduled service was nearly the same as that of the carriers offering regularly scheduled service in jets (0.07 fatal accident per 100,000 departures versus 0.08 fatal accident per 100,000 departures). The aging of the aircraft fleet, which was already under way before the late 1970s, is only indirectly related to deregulation. The reliance on older aircraft by new entrants and the boom in air travel during the 1980s may have created more demand for aging aircraft than would otherwise have
294 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION been the case. The stringent maintenance and inspection schedules for aging aircraft suggested by an industry task force and incorporated by the FAA into regulation, however, appear to reduce concerns about the risk of catastrophic accidents caused by the failure of aged airframes. These more rigorous maintenance schedules will increase the cost to air carriers as well as the demands on the FAA's safety inspection work forceâwhich has not kept up with industry growth over the last decade. The FAA's special in-depth inspection programs for several carriers during the mid- to late 1980s revealed that its routine inspection program (which relies on inspections of paperwork and airline self-regulation to ensure safety and not hands-on inspection of aircraft and maintenance practices), is not sufficient to ensure carrier compliance with maintenance regulations. For example, Eastern Airlines in early 1991 admitted that maintenance officials falsified maintenance records in 1987 and 1988. These falsified maintenance records were detected after a special in-depth inspection, but not by the FAA's routine inspection procedures. This suggests that the FAA's staffing and procedures must be made adequate to provide more hands-on inspection. Although the safety record in commercial aviation has improved since deregulation, continued vigorous efforts are required in carrier oversight and safety regulation to ensure that safety is not compromised. The com- mittee recommends that FAA staffing and inspection procedures be made sufficient to ensure that maintenance is adequate to ensure safety. In addition, the FAA should continue to reduce dfferences in equipment standards and operating requirements between national and regional (commuter) carriers. AIRPORT AND AIRWAY CAPACITY The accelerated growth in demand for air travel since deregulation has begun bumping up against capacity limits at some airports and the airspace around them. Deregulation undoubtedly accelerated traffic growth and the shift to hub-and-spoke networks, which places more peak demand on airport capacity at hubs. Deregulation also gave airline managers more flexibility in applying their assets to meet increasing consumer demand for air travel, and in response to congestion at hubs and at other major airports, airline managers have used this flexibility to develop underused airports into small and medium-sized hubs. As consumer demand has grown since deregulation, the public sector has not provided an adequate
Conclusions and Recommendations 295 supply of airports and controlled airspace. As a result, congestion and delay have increased. Average delay per flight since 1979 has increased more than one-third. The absolute amount of delay per flight is still modest: about 6 minutes per flight in 1979 and somewhat more than 8 minutes per flight in more recent years. During inclement weather, however, delay is much more common, and some flights are delayed considerably more than the average, some more than an hour. Two distinct but interrelated capacity problems currently face the aviation system: an airport capacity problem that occurs at a few major airports during peak periods and an airway or airspace problem that has systemwide characteristics. Airport Congestion Congestion, as measured by 20 delays that exceed 15 minutes per 1,000 operations, occurs at relatively few major airports: Chicago (O'Hare), Atlanta, Dallas-Fort Worth, Denver, the New York City area's three major airports (LaGuardia, Kennedy, and Newark), Boston, and San Francisco. Of these, however, the delays at five of them are by far the worst. At the three major New York City area airports, Chicago O'Hare, and San Fran- cisco, more than 40 flights per 1,000 operations are delayed for 15 minutes or more. These five airports together account for 17 percent of annual enpianements nationally. More Efficient Use of Underused Airports Whereas capacity expansionsânew or greatly expanded airportsâcan be justified in some instances, numerous capacity-increasing options exist that are less expensive than new construction and, given the difficulty of adding runways and airports, will have to be relied on in coming years to meet capacity needs. For example, congestion at the airports mentioned above, as well as that at almost all of those projected by the FAA to be congested within this decade, could be relieved by relying more heavily on existing capacity at nearby reliever airports or underused airports in nearby cities. Islip and Newburgh, New York, for example, have under- used airports that are within 50 miles of one of the three major New York City area airports. Similarly, for most congested hubs and those likely to be congested within the next decade, an alternative, underused airport
296 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION already exists that has adequate runways and the potential to relieve some of the connecting traffic. With the increased discretion provided to carriers to enter and exit markets since deregulation, air carriers have already developed several small hubs, partly to bypass congested hubs. For ex- ample, Piedmont developed Charlotte and American developed Raleigh- Durham as alternatives to Atlanta's congested airport. More Efficient Deployment of Aircraft Air carriers can also be expected to respond to congestion by deploying larger aircraft into congested markets. The preponderance of aircraft de- ployed during the first decade of deregulation were short-range (narrow- bodied) aircraft with seating capacities of up to 170; most had 120 seats or fewer. Use of larger aircraft in domestic markets would allow more passengers access to airports with sufficient demand without a proportional increase in aircraft operations. As total demand grows in the future, carriers can also use smaller long-range aircraft to provide direct flights that bypass congested hubs. Matching Demand and Supply with Market Mechanisms Because much of the congestion at major airports occurs at peak periods, peak-period pricing of runways could be used to allocate scarce capacity more efficiently. The actual fee imposed would depend on the value placed on the landing by the users with the least value for landing during the peak. As the price increased above the level these users would be willing to pay, they would be the first to shift to avoid paying the cost. Even so, the fees charged would be significantly higher than at present and could earn airports substantial revenues. The most recent experiment with in- creased landing fees at Boston Logan airport, though not true peak pricing, shifted general aviation traffic out of the peak, improved the on-time performance of the carriers using the airport, and reduced passenger delay. Higher fees for peak-hour use of airport runways would shift some traffic to the off-peak, thereby reducing some congestion and delay. They also would shift some users to reliever airports and add a new, potentially large source of revenues for airport operators. Congestion pricing would also give more incentive to develop underused airports and promote re- liance on larger aircraft to allow more passenger flow without increasing congestion.
Conclusions and Recommendations 297 Peak-period pricing, however, does entail some problems. Most no- tably, it would provide airports, already in a monopoly position, with the ability to earn substantial revenues and, without other provisions for use of the funds earned, would not necessarily lead to additional investments in capacity. This problem could be resolved by having the funds earned collected by an off-budget central authority of some kind, which in turn would make grants to airports for investments to increase capacity. Each airport at which congestion pricing was used would be able to apply for the funds to increase capacity by any effective means (such as building new runways, buying new equipment, or soundproofing nearby homes). The airports earning the revenue through pricing would have an incentive to bring the funds back to their airport and would thus be encouraged to develop effective proposals for expanding capacity. Airway Capacity National Airspace System Plan The FAA's most ambitious response to the increased demand experienced since deregulation is its plan to replace and upgrade the ATC system. The National Airspace System Plan (NAS Plan) was originally promulgated in 1981 as a 10-year, $12 billion program to replace outmoded technol- ogyâliterally to replace vacuum-tube electronics and in some cases sur- plus World War II radars with state-of-the-art, solid-state systems. During the past decade the NAS Plan has evolved into an ongoing multiyear capital improvement program currently estimated to cost $27 billion and with completion of all the elements in the current program not expected until 2005. The FAA has been widely criticized for not planning its capital needs well and not managing the program effectively, but the increase in total cost and longer implementation schedule also reflect the FAA's awareness that the original NAS Plan would not provide sufficient additional capacity to meet growing demand, nor would it take advantage of technology developments since the plan was envisioned. The program, as currently structured, would substantially upgrade the computers, radars, weather sensors, information systems, and communication links between ATC and aircraft that make up the ATC system. Almost all of the original NAS Plan is now under contract, but some critical components of the plan, such as the Advanced Automation System, have experienced substantial
298 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION delays. Full implementation and costs have escalated, partly because of cumbersome federal procurement regulations and insufficient numbers of managers in the FAA with the technical expertise needed to oversee the contracts, but delays are also attributable to the increased technical de- mands being placed on the system. The various components are now not expected to be fully in place until between 1993 and 2004. If the system works out as anticipated, the FAA projects a roughly 30 percent gain in airspace capacity. Both increased reliance on automation and the devel- opment of more sophisticated (and, it is hoped, more reliable) technology are expected to increase capacity. This change to a more technologically complex mission requires more intricate maintenance because of the in- creased reliance on software; this has direct implications for the FAA's electronic technician work force. Given the complexity of this massive advanced technology acquisition program, uncertainties in federal funding, uncertainties of some of the unresolved technical issues, and the FAA's track record, further delays and cost increases would not be surprising. As the FAA's capital improvement plan has evolved during the last decade, the airspace around some major airports has become increasingly saturated, causing ripple effects throughout much of the ATC system. The simulation models that the FAA has under development indicate that the national airspace system, as currently configured, is rapidly approaching capacity in critical sectors of the airspace. Capacity Management Task In the past, the FAA could rely on local solutions to airspace congestion, but with the growing saturation of the system, changes in procedures at one point may have the unintended consequences of moving the bottleneck (or delays and noise) to some other part of the system, with unanticipated consequences. The change in perspective for the FAA, from relying on controllers to solve local problems to facilitating flows in a national net- work, represents a major shift in the agency's operation; the reorientation within the FAA must be continued. Simulation models and other related research techniques generally would help put air traffic operations on an analytical, systems-performance basis. Models to handle the complexity of the network issues now faced by the FAA, however, are in their first generation, and refinements to an acceptable level of accuracy would require a continued research effort on the part of the FAA. Improved understanding of the systemwide needs for airspace capacity, yielded by these models, could also drive the FAA's plans for capital invest-
Conclusions and Recommendations 299 ments, operations and maintenance, training and procedures, and tech- nology acquisition. Recommendations for Expansion and Better Use of Airport and Airway Capacity Congestion Pricing Congestion pricing of runways could reduce congestion, encourage the use of underused airports, and provide additional revenues for enhancing capacity. The DOT should permit and encourage airports to experiment with congestion pricing and invite evaluation of the effectiveness of these efforts by independent researchers. In the development of these proposals, the DOT should consider how to avoid the potential exercise of monopoly power by airports of airlines and their customers and how the revenues earned by congestion pricing will be used to provide needed additional capacity. Research The FAA should emphasize research on simulation modeling of airport and airspace capacity and related research. Greater use of such techniques would lead to the establishment of performance measures that would help the FAA make better use of existing airport and airspace capacity. The FAA should support and learn to use such research to manage capacity more effectively. CONSTRAINTS ON THE PERFORMANCE OF THE FAA Deregulation ushered in an era during which the government partly with- drew from its past interventions in the marketplace. Despite the govern- ment's withdrawal from controlling prices and entry, however, the FAA continues to have a prominent role as the exclusive provider of ATC and as the regulator of safety. Actually both of the FAA's roles involve safety because its provision of ATC is conditioned by stringent standards to ensure safe separation of aircraft. The FAA's ability to regulate the system
300 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION and provide a safe level of ATC has been outpaced by the demand for air travel since deregulation, and despite renewed support in the President's budget proposals in recent years, a number of questions raise concerns about how well the agency will meet the challenges of the future. Much of the FAA's struggle in the past has been over hiring and training replacements for the several thousand air traffic controllers fired in 1981, a process that is only now nearing completion. The FAA has been able to maintain a high level of safety during this period by deliberately re- stricting traffic and by requiring overtime, and as a result of the dedication and hard work of its air traffic controllers and maintenance technicians. Totally apart from the past contentious disputes between the Reagan ad- ministration and Congress over how many controllers are needed, the entire process of recruiting, identifying, and training air traffic controllers is time-consuming. Each trainee requires 3 to 5 years of training to reach full controller status, and many trainees wash out somewhere in the pro- cess. Although it appears that the FAA now has an adequate controller work force for the system as a whole, it continues to have problems maintaining desired staffing levels in high-cost areas of the country. Although most of the congressional and media attention has focused on the difficult task of replacing and training several thousand controllers, the FAA also has struggled with other elements of its work force in safety- critical jobs. In the mid-1980s the number of technicians responsible for maintaining ATC technology was allowed to decline in anticipation of the delivery of low-maintenance technologies that are part of the NAS Plan. As a result of the projected delays in the NAS Plan and the retirement eligibility of 60 percent of the current work force of maintenance tech- nicians by the year 2000, the FAA may have fewer experienced main- tenance technicians than will be needed. This may present a major problem as both the systems and the technicians age and as the most experienced technicians retire. Partly because of budgetary pressure and ceilings on the federal work force during the 1980s, the FAA's safety inspector work force declined in the early 1980s even as the number of new carriers and aircraft increased. Even though the FAA has been increasing its inspector work force, the agency's inspection procedures still rely on accurate reporting of main- tenance by the air carriers. The committee's recommendation on safety, given above, is designed to reduce the agency's reliance on accurate reporting by the airlines but will require additional FAA staffing. In order to ensure the airworthiness of aircraft and compliance with safe operating procedures, the FAA requires pilots and aeronautical en- gineers with special skills. The agency has long had a problem attracting
Conclusions and Recom,nendaiions 301 and retaining adequate expertise in these areas because government salaries are not competitive with those of the private sector. The FAA, despite considerable challenges, has managed to maintain a safe level of ATC since deregulation and, after years of political and administrative disputes, is rebuilding its work force of controllers, main- tenance technicians, safety inspectors, engineers, and test pilots. As a result, the agency now has a larger staff than in 1978. Whether this work force will enable it to meet growing capacity demands in the future without compromising safety is brought into question by (a) the institutional con- straints within which the FAA must operate and (b) the degree to which these constraints will keep the agency from addressing some complex questions. For example, the FAA field staff evidence an admirable commitment to safety, but the agency as a whole is often slow to respond to emerging safety concerns. For example, the National Transportation Safety Board raised concern about the threat of increased runway incursions in a 1986 report. In 1988 the Aviation Safety Commission observed that the FAA's own data on the risk of runway accidents indicated a growing problem. The commission urged a review of runway signage, directional indicators, and taxiway and intersection markings. Simply standardizing these runway features would reduce the risk of pilot error and would do so at a low cost. In December 1990 and February 1991 accidents on runways at Detroit and Los Angeles cost 42 lives. A week after the Los Angeles accident, the FAA announced the development of a plan that would, among other things, give high priority to standardizing runway signs and markings. The Aviation Safety Commission recommended such action in 1988. Institutional Constraints The FAA has been subject to much criticism in recent years because of persistent staff shortages affecting air traffic and safety inspections and by delays in the implementation of the NAS Plan. Many, if not most, of the FAA's problems, however, appear to originate with the structural and statutory constraints within which the agency operates, which makes dif- ficult the management of a complex organization with many operational responsibilities, notably running the ATC system. For example, the prob- lems with the FAA's work force and the issues the agency confronts are, and have been, well known by many of its professional managers. The FAA has simply not had the discretion or the resources (with the exception of the NAS Plan) to make the most sensible choices. Although these
302 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION problems can be summarized as falling into four areasâfunding, person- nel, procurement, and managementâthey are all tightly interrelated. Funding Despite the federal deficit, the FAA's annual appropriations tend to in- crease each year, but when these funds have been adjusted for inflation and compared with the increased demand for FAA service, the increased funding has not been sufficient. Efforts to mask the size of the federal budget deficit and disagreements between Congress and the executive branch over how much of the FAA's costs the Airport and Airway Trust Fund should cover have impeded funding in the past. Most available new funds have been invested in the NAS Plan, whereas other accounts have remained relatively flat (in constant dollars) or have declined. Given con- tinued federal deficits projected for future years and the increased resources needed for the NAS Plan, the FAA is likely to continue to receive in- adequate funding. Personnel The FAA's pending shortages in electronic technicians appear to result from the funding and staffing restrictions imposed in the past, but other FAA staffing problems trace to the rigidities of the Civil Service System. The FAA has had problems attracting and retaining personnel with the highest level of skill, both technical and managerial, because the salaries it can offer are not competitive with those offered by the private sector. Recent pay reform legislation enacted by Congress would help in this regard, but in the committee's judgment it will not go far enough to solve the FAA's problems. Procurement Acquisition of advanced technologies by federal agencies is a lengthy and cumbersome process, and a program as. ambitious as the NAS Plan is likely to have been delayed even if well managed. Part of the FAA's problem with the management of the NAS Plan results from its past inability to increase the number of managers with adequate technical expertise to oversee the program. The current Administrator has made
Conclusions and Recommendations 303 several reforms to improve the process and has received support for in- creasing the number of managers needed; the committee questions, how- ever, whether the FAA can attract and retain adequate expertise given the private-sector salaries that such individuals can earn. In addition to concern about staffing, adaptations to the NAS Plan that could be made to provide more capacity are hamstrung by the rigidities of the federal procurement process. In recent years several high-level studies (such as the Packard Commission study) have concluded that a massive restructuring of the federal acquisition process is urgently needed. Management Status and Tenure of the Administrator The FAA Administrator is appointed by the President and confirmed by the Senate, and reports directly to the Secretary of Transportation. Al- though this a reasonable arrangement for establishing lines of authority, in practice the FAA Administrator has typically served a short termâ usually 2 to 3 years. In contrast, chief executive officers of Fortune 500 firms serve an average tenure of 8.5 years. Within the federal government, administrators who are appointed to fixed terms (which range from 5 to 12 years) serve nearly the full length of the term. Past FAA Administrators have cited frustrations that stem from the FAA's need to run an operational agency with an additional layer of bureaucracy on top, which has perhaps contributed to the short tenure. Given the increased complexity of the FAA's mission and need for stable leadership during a period of consid- erable change, the record of short turnover by Administrators is cause for concern. Micromanagement Intervention into the management of the FAA's operational responsibilities was infrequent from 1966, the year the FAA was integrated into the newly created DOT, until the 1980s. The mushrooming federal budget deficit and an administrative goal to reduce the size of the federal government apparently created an environment in which the Office of Management and Budget and the Office of the Secretary of Transportation began im-
304 WINDs OF CHANGE: DONIEsTIc AIR TRANSPORT SINCE DEREGULATION posing cutbacks on FAA staffing and funding that weakened the FAA's abilities to carry out its mission. Unanswered Questions for the Future The concern about the future performance of the FAA in ATC and safety oversight of carriers and aircraft emerges from several unanswered ques- tions, many of which will be difficult to address because of the FAA's structural impediments: Given the projections of a continued budget deficit for many more years, will the FAA obtain the funding it needs for personnel and equipment? With the increased future reliance of the FAA on advanced technology in the provision of ATC, will it be able to maintain the current high level of safety as its mission becomes more sophisticated and complex? Given that FAA salaries in key segments of its work force are not competitive with those of the private sector and a shortage of expertise in technical fields is projected in the years ahead, will the FAA be able to attract and retain the personnel it needs? Will the inspector work force be adequate to satisfy the new demands being placed on it for inspecting aging aircraft? Given the pending retirements of most of the electronics technicians, how will the FAA keep its aged ATC technologies operating until the NAS Plan technologies are in place and operating smoothly? Will the increased reliance on having ATC maintenance "contracted out" reduce the expertise needed by the FAA staff to ensure that main- tenance is performed properly and that the system performs adequately and safely? Will the FAA be able to make needed changes to the NAS Plan to provide for capacity demands within the constraints of the federal pro- curement process? Will future FAA Administrators stay on the job long enough to provide stable leadership of the FAA's increasingly complex mission, and will future DOT Secretaries become involved in unproductive micromanagement? The FAA managers who met with the study committee and staff are aware of these issues and can point to many efforts to respond. The committee, however, is concerned that institutional and political con- straints leading to uncertainty over future funding, inadequate ability to plan for long-term personnel needs, inability to attract needed expertise,
Conclusions and Recommendations 305 and frequent changes in leadership will considerably hamper the ability of the FAA to respond to these concerns. Options for Reforming the FAA Three basic choices are available for institutional reform, and the options range from fully public to private models: Return the FAA to its former independent status. (a) Assign the ATC functions of the FAA to a public corporation (a governmental enterprise) while the rest of the FAA's functions remain in the DOT, or (b) convert the entire FAA to a public corporation that is responsible only to the President and Congress or to the DOT Secretary. (a) Assign the ATC functions of the FAA to a congressionally chartered private corporation, or (b) assign the ATC functions to a congres- sionally chartered private corporation and imbed within this corporation a legally independent, governmental unit that would be responsible for setting and enforcing safety standards and for performing many of FAA's safety oversight and regulation other than those directly related to ATC. The first option, that of making the FAA independent, may afford some improvements in agency management but would not address needed fun- damental reforms in pay, procurement, budgeting, and financing. Signif- icant progress in this regard would be made by converting the Airport and Airway Trust Fund into an enterprise fund, by allowing the FAA to charge users for its services, and by giving the agency's managers greater dis- cretion over personnel. In addition, the reforms in procurement suggested by the Packard Commission and others could be adopted, if the FAA were freed from restrictive federal procurement legislation. Options 2 and 3 incorporate these proposed reforms. Separating the FAA's air traffic service from the rest of the agency by assigning them to either a public or private corporation, as suggested by Option 2(a) and 3(a) would improve the efficiency of the ATC service. Separating the ATC functions of the FAA from its other regulatory func- tions, however, could reduce the efficacy that exists when all regulatory and operations functions are housed in one agency. Although it has not been proven that separating the different functions of the FAA would compromise safety, many observers familiar with the FAA have argued that it would. Options 2(b) and 3(b) propose to retain this linkage either
306 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION by keeping the separate functions of the FAA in one entity [Option 2(b)] or in close proximity [Option 3(b)]. Option 2(b), the public corporation option involving all of the FAA, is attractive because it would appear to provide more managerial discretion for the FAA in dealing with problems in funding, personnel, and pro- curement. This option would include the kinds of public oversight that are exercised over most federal corporations. Divorcing the corporation entirely from the DOT, however, would risk some loss in the ability to develop a comprehensive national transportation policy. It would also require the establishment of a process or an agency to review the rates charged to its customersâanalogous to the Postal Rate Commission. To help address these shortcomings, proposed in Option 2(b) is the estab- lishment of a policy and reporting link to the Secretary of Transportation, whose authorities would not be delegable to lower-level officials. Whereas the private entity suggested in Option 3(b) would have the greatest flexibility in developing management and pay incentives that should improve the performance of ATC, many would be uncomfortable with completely privatizing the Air Traffic Control Service. The proposal to imbed a public safety oversight function within this private entity might help ease this concern. Options 2(b) and 3(b) would give the FAA the discretion and authority needed to address its most pressing operational problems. Congressional oversight would be retained, but the forms of accountability would differ. Subsequent study of the possible reorganization of the FAA requires more operational analyses of future demands and likely performance than is now available. At least five types of information are desirable: (a) a reasonably detailed summary of the current (and past) level of activities in the several functions carried out by the FAA; (b) estimated changes in demands as well as any operational modifications intended to cope with them; (c) a characterization of the externally enabling and constraining measures levied by federal law, regulation, the Office of Management and Budget, the Office of the DOT Secretary, and Congress; (d) a review of the experience of civil governmental agencies that are of a more or less analogical character to the FAA and a delineation of how well existing corporate forms within the federal government as well as ATC corporations in other nations exercise the greater discretion they have; and (e) a com- parison of how public accountability would be exercised in options 2 (b) and 3 (b). Without more study it is unclear which option would provide for superior overall performance. The focus of options 2(b) and 3(b) is on improving the effectiveness of the FAA's headquarters operation and general staffing (both expertise
Conclusions and Recommendations 307 and total number), which the committee believes deserve most emphasis. Little information is available, however, on whether either option would in any way place at risk the operational integrity of the agency. Although the FAA's performance can and should be faulted in some areas, the performance of the FAA's staff in the field indicates an admirable com- mitment to safety; the effect that any reorganization might have on this commitment deserves careful consideration. The committee is concerned about the ability of the FAA in its current form to meet future challenges posed by continued air transport growth. This concern extends beyond the efficient provision of airspace capacity; the committee is not assured that the FAA in its current form, despite the best intentions of its managers and staff, will be able to continue to maintain the high level of safety in the aviation system that the American public enjoys and has come to expect. Of the options listed above, 2(b) and 3(b) would provide the authority and discretion needed to improve operational performance without severing links between regulatory and operational functions, which may compromise safety. The committee rec- ommends a publicly mandated study of change in the organization of the FAA by an independent group or organization that focuses on the relative merits and drawbacks of options 2(b) and 3(b), with a report to the President and Congress within 2 years after the study gets under way.
Appendix A Measuring Financial Risk in the Airline Industry John S. Strong College of William and Mary Financial markets provide one means of evaluating changes in the relative risk of the airline industry during the past two decades. A widely used measure of sensitivity to the business cycle is a company's equity beta, which measures how volatile a company's stock is compared with the overall stock market. Equity betas are estimated by regressing a company's stock returns on those of the market portfolio as a whole. A value of 1.00 indicates a level of equity risk equivalent to that of the market; a value greater than 1.00 indicates higher risk, and less than 1.00 indicates less risk. If companies were all-equity financed, the risk borne by stockholders would be equivalent to the riskiness of the company's assets as currently employed. Because virtually all companies have some debt, stockholders also bear the financial risk of leverage. An attempt can be made to separate these risks to some degree by recognizing that an airline's asset risk (the risk of its underlying business) is shared by its debtholders and its share- holders. Using beta as the measure of relative risk, this is equivalent to the following: Asset beta = debt beta (percentage of debt in total capital) + equity beta (percentage of equity in total capital) In short, equity betas can be used to examine what has happened to the - relative riskiness of airline stocks and asset betas as one measure of the changes in basic business risk in the airline industry. For comparison purposes, the average asset beta for companies in the Standard and Poor's Im
Appendir A.⢠Measuring Financial Risk 309 500 index during the 1980s was roughly 0.75, having risen from approx- imately 0.64 in the 1970s. ESTIMATING TECHNIQUE The estimation of equity betas and asset betas done here is consistent with standard asset pricing theories of finance. The equity betas were estimated by regressing 60 months of stock returns on returns on a 1-year Treasury bill and those of the Standard and Poor's 500 stock index (the market proxy). This results in the following regression: R1 - R1 = a + B, (R,,, - R1) + error term (1) where Ri the return on company i's stock in a given month, R1 = the Treasury return, a = constant, R, = the return on the stock market overall, and B1 = the company's equity beta. Because the market portfolio is perfectly correlated with itself, it will have an equity beta of 1.00. These equity betas incorporate not only the business risk inherent in the company's assets as currently employed, but also the effects of fi- nancial risk borne by stockholders. This financial risk is a result of the use of debt financing. Just as the claims on a company's assets or profits are divided between debtholders and stockholders, the risks can be treated in similar fashion: Asset or business risk = debt risk + equity risk (2) This can be estimated by decomposing the equity risk (measured by equity betas) into component parts: Asset beta = debt beta * {(l - t)D/[(1 - t)D + E]} (3) + equity beta * {E/[(1 - t)D + E]} where D = debt on the balance sheet, E = equity on the balance sheet, and t = the company's tax rate.
310 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION (Because interest on debt is tax-deductible the after-tax cost of debt is relevant.) This is equivalent to saying that the business risk of a firm is composed of the riskiness of the debt weighted by the percentage of debt and the riskiness of equity weighted by the percentage of equity. Most textbooks and analysts assume that the beta of debt is zero because debtholders are assumed to bear much less risk than stockholders. Al- though this is true in general, it need not be true for individual companies, so it should be considered. Also, the focus here is to measure how risky the debt, equity, or assets are relative to the economy as a whole. The increased volatility of interest rates in the 1980s made debt more risky than in the past; therefore it is especially important to incorporate debt risks now. This was done in this analysis by regressing the returns on each airline's publicly traded debt on the Standard and Poor's 500 Index to provide a measure of the sensitivity of debt returns to the business cycle. Airline debt betas ranged from 0.14 to 0.38, with an average of 0.21. For comparison, the average public debt issue traded on the New York Stock Exchange had a debt beta of 0.12 in the 1980s. These estimated debt and equity betas were inserted into Equation 3 to calculate asset betas. In that equation, the debt term includes all long: term debt, including capital leases. The equity is taken from the balance sheet. The tax rate is the effective tax rate from the company's income statement. In almost all cases, this tax rate is below the statutory rate. The data cover all airlines whose equity was publicly traded on the New York or American Stock Exchanges for the preceding 60-month period. The security returns are from COMPUSTAT, Moody's Bond Guide, and the Center for Research in Security Prices at the University of Chicago. The financial statement data were collected from Moody's Transportation Manual for each year. In summary, capital asset pricing theory is used to find equity betas, which are combined with debt betas and financial statement information to construct estimates of asset betas for each year. CONCLUSIONS Equity and asset betas estimated annually for the 1973-1989 period are presented in Table A-l. Each year's estimate is based on the returns of the previous 60 months, consistent with standard practice.
Appendix A: Measuring Financial Risk 311 TABLE A-I EQUITY AND ASSET BETAS OF THE U.S. AIRLINE INDUSTRY, 1973-1989 Year Equity Beta Asset Beta U.S. Airline Industry 1973 1.23 0.86 1974 1.28 0.84 1975 1.34 0.97 1976 1.14 0.78 1977 1.12 0.72 1978 0.96 0.64 1979 0.88 0.71 1980 1.04 0.78 1981 1.01 0.73 1982 1.08 0.64 1983 1.20 0.71 1984 1.17 0.79 1985 ç 1.18 0.69 1986 1.12 0.79 1987 1.10 0.75 1988 1.37 0.89 1989 1.35 0.92 Standard and Poor's 500 Industrials 1973-1980 1.00 0.64 1981-1989 1.00 0.75 Nom: Equity and asset betas shown are estimated on a carrier basis. The numbers shown are industry averages weighted by total assets in that year. Equity betas were estimated using the security returns of a 60-month period, whereas asset betas were calculated by removing the effects of financial leverage on the equity betas. For additional details see the section on estimating techniques. The main conclusions are as follows: During the last few years of regulation the.riskiness of airline stocks declined, but remained higher than that for the stock market as a whole. The low default risk of debt and the more stable interest rate environment meant that the risk from leverage borne by airline stockholders was less than average. During regulation the basic business risk as measured by asset betas was higher than in the economy overall, although this too declined. In the early to mid-1980s, airline asset betas remained fairly sta- ble, whereas those of the overall economy rose. This trend suggests
312 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION that the relative business risk of the industry became more like that of the economy as a whole. However, equity betas rose, compared with the late 1970s, as a result of the increasing amount of debt in the airline industry. Thus, during this period, the riskiness of airline stocks rose, not so much because of increased business risk but rather because of increased debt risk. 4. In 1988 and 1989, however, both equity risk and business risk rose sharply. As large capital expenditure programs began to take effect and new planes were delivered, the operating leverage (fixed capital costs) of the carriers increased. Because much of this fleet was debt- financed, the financial risk rose as a result of both the increased amount and the increased riskiness of the debt. This effect was magnified further by additional debt incurred in the wake of mergers and acquisitions during 1988 and 1989.
Appendix B Organizational Options for the Federal Aviation Administration PREFACE To assist the committee in evaluating previous proposals for reform of the Federal Aviation Administration (FAA), as well as previous studies on this topic, Herbert N. Jasper was commissioned to prepare a paper, which is contained in this appendix. The committee agrees with Mr. Jasper's summary of the issues and problems facing the FAA, and with much of his analysis, but some members did not share his conclusions. The committee believes that more study is required of the organizational options available for the FAA before conclusions can be drawn and final recommendations made. The committee believes that this paper represents a substantial contribution to the literature addressing the reforms needed at the FAA and that any subsequent group charged with a study of the FAA will find its understanding much advanced by it. 313
314 WINOS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION ANNOTATED TABLE OF CONTENTS The FAA's Mission and Status, 316 States where responsibility for provisions of the Federal Aviation Act of 1958 now resides. Organizational History, 318 Gives reasons for each reorganization since 1926. Constitutional, Political and Economic Environment, 319 Describes framework within which the FAA operates. Problems Confronting the FAA, 321 Discusses Administrator's short tenure; micromanagement; personnel, budget and acquisitions constraints; performance; culture; responsive- ness; rulemaking; and role conflicts. Alleviation of Problems Without Reorganization, 334 Assesses possibilities for improvements in each major problem area without legislation, or without a new institutional form. Greater Flexibility Vested in Other Federal Agencies, 338 Discusses agencies' flexibilities and refers to Exhibit 1, Table B-i, which cross-references many of the provisions discussed in the exhibit, and Table B-2, which compares the various institutional forms in re- lation to probable effectiveness of desired reforms; draws lessons from Exhibit 1 relative to reforms which would likely accompany a given institutional form. Recent Proposals to Reorganize the FAA, 341 Characterizes the reasons why observers have argued for the status quo, or for four other institutional forms; gives advantages and disadvantages of each, and a summary assessment of each, focusing primarily "on the merits." Observations On Political Feasibility, 357 Comments on what is likely to be required to secure passage of legislation.
Appendix B: Organizational Options for the FAA315 Some Possible Configurations for an FAA Corporation, 359 Based on both desirability and feasibility, outlines what proposed fea- tures might result in legislation to produce a corporation which could function far more effectively than does the the FAA. Summary, 363 Recaps reasons for choice of institutional form in relation to principal problems to be overcome. Exhibit 1, 366 Provides all significant cases in which freedom from the constraints under review has been given to agencies, organized by type; also cites all known cases of fixed terms for heads of single-headed agencies. References, 382 Lists principal sources for identification of the FAA's problems, prec- edents for changes considered, and proposed institutional remedies.
316 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION Organizational Options for the Federal Aviation Administration Herbert N. Jasper As described in this appendix, the FAA and its predecessors since 1938 have been in a number of organizational locations. The FAA was included in the Department of Transportation (DOT) when that department was established in 1966. For the next 15 years or so, few questions were raised about the FAA's organizational locus, its reporting channels, or its ability to function effectively. However, in response to growing concerns about aviation service and safety issues since the early 1980s, a number of studies have been conducted that have addressed the question of the organizational status of the FAA. In 1989 the Transportation Research Board of the National Research Council appointed a committee to study air passenger service and safety since deregulation. The committee determined that it would be helpful to have a review of the various organizational proposals that had been advanced in various quarters, including in Congress, with the objective of assessing which option or options would be most likely to enhance the FAA's ability to discharge its safety and service functions. This review of organizational options for the FAA was not to entail original research. Rather, it was to identify relevant problems confronting the FAA that had been discussed in previous studies or congressional testimony, assess the degree to which there was consensus about those problems, evaluate the leading reorganization proposals with respect to their predicted impact on those problems, and suggest one or more options that appear to be first, desirable, and second, feasible. Most of the problems discussed in this appendix therefore, are those presented in other studies. MISSION AND STATUS The FAA is one of nine operating agencies in the DOT. The others are the U.S. Coast Guard; the highway, railroad, traffic safety, urban mass transportation, maritime, and research and special programs administra- tions; and the St. Lawrence Seaway Development Corporation.
Appendix B: Organizational Options for the FAA 317 The FAA's authorities and responsibilities are derived from the Federal Aviation Act of 1958. The act creating the DOT in 1966 transferred most of those functions to the DOT Secretary, except for promotion functions (see below) and the safety functions, which it sought to place directly in the FAA (but see section on Agencies Within Cabinet Departments). The Secretary's functions were delegated to the FAA by "Federal Order." The text of that order follows: The FAA is responsible for Promulgating and enforcing regulations on all safety matters relating to the manufacture, operation, and maintenance of aircraft; Registering aircraft and recording rights in aircraft; Developing, modifying, testing, and evaluating systems, proce- dures, facilities, and devices needed for the safe and efficient navigation and traffic control of aircraft; Locating, constructing or installing, maintaining, and operating fed- eral aids to air navigation wherever necessary; Developing air traffic regulations, and administering air traffic con- trol of civil and military air operations within U.S. airspace; Providing grants-in-aid for developing public airports; and Promoting and encouraging civil aviation abroad through tech- nical aviation assistance to other governments [49 C.F.R. 1.4(c)]. Section 103 of the Federal Aviation Act contains among its policy declarations "promotion, encouragement and development of civil aero- nautics." That section was not involved in the delegations by the Secretary because DOT lawyers believe that those policy mandates passed directly to the FAA under the statute creating the department. Similarly, Section 305 of the act, which directed the Administrator "to encourage and foster the development of civil aeronautics and air commerce in the United States and abroad," was apparently not transferred to the Secretary by the DOT legislation. The agency is headed by an Administrator who is appointed by the President and confirmed by the Senate, as is the Deputy Administrator. All other personnel are appointed by the Administrator, or pursuant to authority delegated by the Administrator. Under a 1988 reorganization, four Executive Director positions were established, and the chief operating functions were assigned to those directors. The post of Associate Admin- istrator for Aviation Safety was also created. Early in 1990, the succeeding Administrator added a fifth Executive Director (for Acquisition) and trans-
318 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION ferred the airports responsibility to an Assistant Administrator reporting directly to him. ORGANIZATIONAL HISTORY The precursors to the FAA date from the Air Commerce Act of 1926. Air safety, promotion, and regulatory functions were vested in the Secretary of Commerce until 1938. Reflecting the growth of the industry, they were transferred to a five-member, independent Civil Aeronautics Authority, which was created by the Civil Aeronautics Act of 1938. Within the authority was a three-member Air Safety Board, which had accident in- vestigation and related responsibilities. As part of a reorganization program authorized by the Reorganization Act of 1939, a 1940 presidential reorganization plan renamed the authority the Civil Aeronautics Board (CAB) and the several aviation responsibilities were placed in the Commerce Department once more. However, the post of Administrator was removed from the board and assigned to an Ad- ministrator of Civil Aeronautics, who reported directly to the Secretary. He headed the Civil Aeronautics Administration (CAA), with responsi- bilities for airways, aircrews, and safety. The separate safety board was abolished, and its functions were assigned to the CAB. Spurred by several midair collisions and by concerns that the CAA's funding was inadequate and its personnel were demoralized, Congress accepted the recommendation of the Eisenhower administration that the CAA be removed from the Commerce Department. The Federal Aviation Act of 1958 renamed the CAA the Federal Aviation Agency, and made it an independent agency. The FAA was headed by an Executive Level II (the cabinet department under secretary level) when it was independent, as it is now. However, the salary was more competitive in those days, and the average tenure of Administrators was somewhat longer than in recent years. The FAA had no exemptions from the procurement, personnel, and budget systems. However, as discussed later, all of these systems have become more complex or restrictive since then. The FAA did receive authority to pay 10 (later 22) experts at salaries above the civil service salary ceiling. In 1965, then Administrator Halaby determined that it made good sense to have a department of transportation and that such a department would not be viable without the FAA. He also had become concerned that the agency's programs were not receiving the support required and hoped that the FAA would have more influence in getting the resources needed to
Appendix B: Organizational Options for the FAA 319 do its job if it had a cabinet secretary to represent its interests within the executive branch and before Congress. It was he who recommended to President Johnson the creation of the DOT, and the President then proposed legislation to do so. Thus, when Congress enacted that legislation in 1966, the FAA became part of the DOT. In summary, aviation functions have been vested twice in the Depart- ment of Commerce and have twice been removed to an independent agency. The DOT thus makes the fifth location since 1926. CONSTITUTIONAL, POLITICAL, AND ECONOMIC ENVIRONMENT Before introducing the problems that have been ascribed in whole or in part to the FAA's status, it may be useful to review some of the basic conditions that govern the operation of agencies in the federal executive branchâconditions that are taken into account in the evaluation of or- ganizational options presented later in the appendix. Further, certain events or developments that have had a significant impact on FAA's operations must be recognized before one can effectively assess the effects of its organizational status. Constitutional or Statutory Although the President is the Chief Executive, virtually all statutory pow- ers are vested directly in the heads of agencies, whether they be executive departments (sometimes called cabinet departments) or independent agen- cies. The latter term describes an agency that is not part of an executive department. "Independent regulatory" board or commission likewise means a reg- ulatory board or commission that is not in an executive department. (An exception is the Federal Energy Regulatory Commission, which is in the Department of Energy.) Although such agencies are headed by three or more members, they perform functions similar to the regulatory functions performed by many single-headed independent agencies, such as the En- vironmental Protection Agency, or by agencies within departments, such as the FAA. Congress exercises important oversight powers through its own inves- tigations or hearings and through studies by its staff agencies. Oversight can lead to enforcement of congressional wishes through informal pres-
320 WINDs OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION sures, or through legislation, including appropriations, the accompanying committee reports, or appropriations riders. Political When one or both houses of Congress are not under the control of the President's political party (as has been the case since the FAA's status began to cause concern), the opportunity for conflicting direction of ex- ecutive agencies is increased. Thus, the Reagan administration's policies of restraining federal spending and employmentâpolicies that were not always endorsed by Congressâcreated situations in which the policies and priorities of the DOT often differed from those of the FAA or of Congress. Events and Developments That Have Significantly Affected the FAA A number of changes in the FAA's environment have had a major impact on its operations and the way in which they are perceived. Chief among them were the following: The 1978 deregulation of airline service, with the resulting rapid growth in passengers, creation of new airlines, proliferation or expansion of hubs with greatly increased traffic, and emergence of substantial delays in arrivals and departures at these crowded hubs. The 1981 strike by the Professional Air Traffic Controllers Organi- zation (PATCO); the firing of 11,400 strikers. (a small number of whom were reinstated); and the long process of rebuilding the controller work force. Complaints about declining quality of airline service. A ballooning federal deficit resulting in White House, Office of Management and Budget (0MB), and congressional restrictions on po- sitions, pay, training, travel, retirement benefits, and expenditures from the Airport and Airway Trust Fund. It is important to keep these developments in mind as one considers the problems confronting the FAA that are discussed in the next section, particularly the discussion of the FAA's performance.
Appendix B: Organizational Options for the FAA 321 PROBLEMS CONFRONTING THE FAA The problems mentioned below have been described in a number of earlier studies or proposals. The principal ones are those by the Air Transport Association of America (AlA), a panel of the National Academy of Public Administration (NAPA), the Aviation Safety Commission (ASC), the Reason Foundation, the Heritage Foundation, the Office of Technology Assessment (OTA), and by a number of members of Congress. Generally, those commenting on the FAA's problems have placed major emphasis on the following: Lack of control over its financing, The short tenure of Administrators, Lack of flexibility regarding personnel and procurement, and Micromanagement by the Secretary of Transportation or the DOT staff. A second order of problems that have been widely noted but which have generally been judged to be less harmful to the FAA's ability to operate effectively concerns the following: Congressional micromanagement, Failure to perform well in the politics of Washington, The cumbersome regulatory process, A bureaucratic culture that is resistant to innovation or rapid change and more disposed to avoiding criticism, Lack of responsiveness to industry changes, and Conflicts of roles involving promotion versus regulation, and oper- ations versus regulations. All of these problems should be viewed in the context of the environment discussed above, as they might prove to be of lesser consequence under different circumstances. Structural or Statutory Impediments Status and Tenure of the Administrator The DOT Secretary is a presidentially appointed, Senate-confirmed (PAS) position, paid at Executive Level I, now $138,900 per year. The Deputy
322 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION Secretary is a Level II, and the Administrator of the FAA is also a Level II, which now pays $125,100. All of them are appointed for indefinite terms; that is, they serve at the pleasure of the President. Four Assistant Secretaries, the General Counsel, and the Inspector General (all PAS) are paid at Level IV. An Assistant Secretary for Administration is a non-PAS appointee who is paid at Level V. The Administrator of the FAA is one of some 500 PAS positions. (Although the Deputy Administrator is a presidential appointee, the po- sition has often been filled by a career civil servant or an officer of one of the military departments.) Members of regulatory boards and commis- sions are appointed by the President for fixed terms, but only a handful of other PAS appointees have fixed terms (see section on Other Federal Agencies or Corporations). Of the 11 Administrators since 1958, 3 served less than 2 years and only 2 served 4 years or more. The last 3 Administrators served an average term of 2.5 years, which is longer than the slightly more than 2 years' average for all PAS appointees. The average tenure from 1965 to 1981 (counting Administrator McKee's two appointments as one) was slightly longer, about 2.8 years. The average tenure for McKee's two predecessors was 3.3 years (including the longest term of any Administrator-53 months). Thus, tenure has grown shorter. One factor that affects the tenure of the Administrator is that of the Secretary because a new Secretary often wants to have new agency heads appointed. The DOT Secretary's tenure has averaged only about 2.5 years. In contrast to the short tenure of the Administrator (and the heads of other federal agencies) the average service as Chief Executive Officer (CEO) of a Fortune 500 corporation is 8 to 9 years. An Administrator may be willing to take such a high profile job despite the lower level of compensation and benefits, as compared with those for the kind of position such a person might hold in the private sector. Will- ingness to remain for an extended period, however, is another question. The short tenure of recent Administrators may reflect the frustration of trying to operate in the difficult federal environment, together with low pay and, as a result, enhanced prospects for moving into higher-salaried private employment. All who have recently studied the FAA have concluded that the short tenure of the Administrator is a serious problem. When an agency head serves only a short time, there are a number of adverse consequences, including a tendency to focus on short-term instead of more significant but delayed improvements (ASC 1988, 21); a lack of continuity or stability in priorities, especially those involving large capital expenditures; and a
Appendix B: Organizational Options for the FAA 323 resistance on the part of the bureaucracy to respond to new directions because it is expected that the Administrator won't be around long enough to follow through (ASC 1988, 22). Role of Secretary of Transportation and DOT Staff Offices One of the factors that engendered great criticism during the Reagan administration was what has been called micromanagement of the FAA by Secretaries Dole and Burnley. Examples are setting of ceilings on categories of FAA employees, or directives to increase employment in such categories, rather than assigning the FAA a personnel ceiling and allowing the Administrator to determine how best to use that ceiling. Micromanagement was a subject of criticism in most of the recent studies of the FAA. (ASC 1988, 20; NAPA 1986, 10; OTA 1988, 65; Reason Foundation 1986, 5) Micromanaging constituent agencies like the FAA is not unique to the DOT, nor does it stem entirely from the actions of the Secretary. Staff at the DOT has numbered about 1,000 persons during the period under consideration, most of whom are, of course, assigned to routine or ad- ministrative functions. But the five Assistant Secretaries, the General Counsel, and the Inspector General all have staffs to help them exercise some oversight of the FAA. The boundary between needed oversight and harmful intervention is not always clear. However, these staff groups have sometimes caused problems for the FAA Administrators, even when the personal style of the Secretary was not to be meddlesome. There might have been fewer cases of departmental intervention in the FAA's management during the 1980s but for the efforts of the Reagan administration to terminate, alter, or restrain programs. Because the FAA was in competition with the rest of the DOT for funding and personnel, the imposition of personnel ceilings and budget reductions by the White House and the 0MB put the DOT in a position in which it had to force the FAA to absorb some of the cutbacks that were ordered. (As discussed later, the OMB's restrictive actions are more visible when they affect an agency directly.) This sometimes took the form of hiring freezes, and it underlay the Secretary's refusal to allow the FAA to rebuild the strike- depleted controller work force as rapidly as the FAA proposed, or as Congress ultimately demanded. Training of the FAA's inspector work force was the subject of on-again, off-again priorities of the DOT, and the DOT also forced reductions in the numbers of maintenance technicians, beyond the requirement that much of their work be contracted out.
324 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION There is another aspect of the issue of departmental intervention. Former Chairman Burnett of the National Transportation Safety Board (NTSB) has testified that the DOT sometimes pressured the FAA to issue safety regulations based on NTSB findings and recommendations, but which the FAA opposed (U.S. Congress 1988, 51. Depending on one's views about the merits of such cases, this may be good or bad. (For example, some observers believe that the cases referred to by Mr. Barnett involve deci- sions made for public relations reasons, rather than for cost-justified safety reasons.) However, it may support the concern of some that the FAA identifies too closely with the industry. Political Gaming As noted earlier, the shared power provisions of the constitution create inherent opportunities for conflict between Congress and the executive branch, which are enlarged by a split in party control within or between the two branches. While the head of a department is more likely to side with the White House and the 0MB on such issues as hiring and spending, an administration such as the FAA is more likely to be sensitive to the priorities and pressures of Congress, before which it must support its legislative and funding requests. Indeed, the term micromanagement has often been used to describe congressional intervention as well as that by Secretaries or the 0MB (NAPA 1986, 11). Further, the tension between Congress and the executive branch has clearly contributed to the stop- and-start character of some vital staffing policies, such as when to hire controllers and inspectors, and how many to hire. Regulatory and Other Statutes As pointed out earlier, authorizations and appropriations are generally given directly to agencies. They typically prescribe fairly specifically what the agencies may or may not do, or what they are required to do. Thus, the Airport and Airway Trust Fund is established by statute, as are the sources and levels of revenues, and the requirement that the receipts can be spent pursuant only to appropriations. This provision has led to frequent changes by Congress regarding the percentage of the FAA's system op- erating expenditures that can be financed from the trust fund. Other provisions of law can substantially limit the discretion of an agency. For example, the Federal Aviation Act has been construed to
Appendix B: Organizational Options for the FAA 325 prohibit the imposition of a head tax on passengers, which could provide a significant source of revenues to finance badly needed airport construc- tion or improvement. (Legislation enacted in 1990 would expressly permit the imposition of passenger facility charges by airport operators.) Simi- larly, the act's nondiscrimination provisions have been construed to block certain types of user charges for aircraft landings and take offs, which could be used to smooth out traffic and thus reduce congestion and delays at airports. The complexity of the regulatory process stems from the elaborate due process requirements of the Administrative Procedure Act. However, as discussed below, the inordinate delays in rulemaking, such as those involved in prevention of misfueling accidents, requirements for more smoke- and fire-retardant materials and in strengthened seat mounting, are probably a reflection of political and economic considerations and of the agency's own rulemaking policies, practices and procedures than they are of statutory origin. Similarly, the interventions of the DOT and the 0MB in the FAA regulatory process are mostly in response to the same political and economic factors. The Aviation Safety Com- mission noted that there have been few cases in which the DOT or the 0MB reviews "changed even minor aspects of the technical content of the proposed rules" (ASC 1988, 19). However, this observation appears to have been based on the formal record of changes made after a proposed rule had been forwarded by the FAA and not upon the informal negotiating process which takes place among the FAA, the DOT and the 0MB before a rule is officially proposed. Some members of Congress and the project director of the OTA study have suggested that the so-called promotion responsibilities of the FAA may have inhibited more vigorous regulatory and enforcement actions (U.S. Congress 1988, 31, 34). However, a similar concern for their industries' welfare can be observed in all other federal regulatory agen- cies, none of which has any explicit promotional responsibilities. Financing The Federal Deficit Except self-financed corporations, no federal agency has escaped the ef- fects of the growing federal deficit. Sometimes, even corporate entities are affected. For example, then 0MB Director James Miller directed the
326 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION U.S. Postal Service to defer about $1 billion in capital expenditures, and to pay more than $400 million to the treasury for employee benefits, during fiscal years 1987 and 1988. This was done despite the fact that it was not even subject to appropriations (except for a half-billion dollar, revenue forgone appropriation). Particularly significant was the so-called Gramm-Rudman-Hollings law (GRH), which established deficit ceilings, year by year, and mandated compliance with them through automatic cuts. Those arrangements were supplanted by different provisions in the budget summit package enacted in 1990. In the main, these new provisions insulate agencies from the effects of increases in different segments of the budget, or from extraneous factors that increase the deficit. But they do require that increased funds voted for one program be offset by reductions in funds for others in the same segment. The FAA and the other agencies that perform vital public safety func- tions, such as the Secret Service, the FBI, and the Coast Guard, have been subject to the employment restrictions and spending cutbacks required by the deficit. No change is likely in that situation for the foreseeable future. Similarly, the FAA has not been permitted by Congress and the Pres- ident to spend as much as it sought from the trust fund for airport im- provements, or to spend as much of the trust fund as it wished for its operating costs. This has given rise to complaints that those who pay the fuel and ticket taxes that go into the trust fund have been denied the benefits promised them. On the other hand, as expenditures mount for the completion of the National Airspace System Plan (NAS Plan), it appears that the large surplus in the trust fund may be used up. It should be noted that a statute creating a trust fund can provide for expenditures from it without annual appropriations action, as is the case for the Social Security trust funds. Alternatively, the revenues now placed in the trust fund in the treasury could be placed instead in a public enterprise fund in the FAA, and the FAA could be empowered to spend those funds without appropriations action. While there are substanlial precedents for such arrangements (as discussed in Exhibit 1), they are not widely known, even by members of Congress (U.S. Congress 1988, 282). Those who have recently reviewed the FAA's problems have concluded that inability to plan and manage its finances has contributed significantly to its perfbrmance problems (OTA 1988, 6; ASC 1988, 20,23; NAPA 1986, 19; Reason Foundation 1986, 5). \
Appendix B: Organizational Options for the FAA 327 Delayed Funding Other problems inherent in the shared powers system have been exacerbated by the deficit and by the split in party control. For example, policy differences between the two branches have caused serious delays in the passage of appropriations bills. In some years, the whole government has been financed for a while (or for an entire year) by a continuing resolution, which generally authorizes agencies to spend during a defined period in accordance with a prescribed formula, pending the enactment of regular appropriations. Some- times, an agency may be funded by two or more short-term continuing resolutions before the final appropriations bill is enacted. Planning, especially long-range planning, is always difficult because of the year-by-year funding style of the federal government, but efficient management, as well as planning, suffers when the fiscal year is well under way before an agency knows what its funding and employment levels will be. Moreover, in some years, the 0MB has required changes in mid-year in order to deal with increased government-wide expenditures, as was the case in connection with several annual pay increases. Another consequence of inadequate and uncertain funding is that re- search and development (R&D) and long-range capital improvements sometimes get too little support. Thus, the FAA tends to lag behind the state of the art, and obsolete systems have been sorely taxed by the significant growth in volume. Furloughs Last year another example of the chaos created by the federal deficit was the issuance of notices of possible furloughs to federal employees. Because the expected deficit for fiscal year 1991 exceeded the ceiling in GRH, and because of statutory requirements for proper notice to employees, all FAA employees, including controllers and inspectors, were warned that they might be placed in a nonpay status for up to 22 days during fiscal year 1991. Federal law also prohibits employees from working voluntarily, that is, without pay. The budget summit legislation enacted in 1990 appears to obviate the need for furloughs, but some damage was done anyhow. The mere threat seriously disrupted morale and suggested to employees that the federal government is a hostile environment in which to work. The possible consequences if the furloughs had been implemented were even more troublesome in terms of an overworked staff striving to keep aircraft safe
328 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION and safely separated with some of the work force in furlough status. In order to mitigate such possible problems, the FAA announced plans to require drastic reductions in operations at about 40 airports if the furloughs were to become effective. Human Resources A number of human resource problems have already been mentioned, such as those of personnel ceilings, restrictions on travel and training funds, failure to keep federal pay competitive, restructuring and cutting back of retirement benefits, and threatened furloughs. Others that have been sig- nificant include the difficulty in reassigning experienced controllers to high-cost areas, and the lack of authority to pay relocation allowances (that is, beyond the actual costs of moving one's family and possessions). It was concluded in each of the three major studies of the FAA that serious human resource management problems exist in the FAA (OTA 1988, 65; ASC 1988, 20, 21; NAPA 1986, 19). Given the variety of constraints under the federal personnel and financial structures, it is not surprising that the FAA faces a number of difficulties in human resource planning and utilization. While some believe that the FAA could and should have done better than it has in anticipating or responding, the fact remains that serious problems have arisen or are imminent. These include an aging work force (especially technicians); a more technical work force (one that is growing harder to recruit or retain); the erosion of technical competence, as retired workers are not replaced or as their successors are not given needed training; and an ill-trained echelon of managers. The FAA's human resource management challenge is probably as great as that facing any federal agency. It has need for an extremely diverse set of personnel, including engineers, scientists, planners, controllers, technicians, inspectors, regulators and, not the least, managers. Although each category of employee poses different management issues, the FAA must deal with them under a rigid "one size fits all" federal personnel system. Its managerial weakness may be one of its most serious problems. The agency has developed a pattern of promoting controllers (who get promoted rapidly and burned-out early) to senior management positions, despite the fact that they are often ill-equipped by experience or training for such positions. Last year, the Bush administration persuaded Congress to change the retirement law once more, this time to cancel the right to take part of one's retirement benefits in a lump sum. Because Congress had not yet
Appendix B: Organizational Options for the FAA 329 acted by the time of the proposed cut-off date for retiring under the previous law, a number of senior managers (including three of the five Executive Directors) and other FAA employees elected to retire suddenly, before that date, so as not to risk the loss of the lump-sum option. Altogether, retirements roughly tripled in the month before the deadline. The deficit reduction package later agreed to by the President and Congress delayed the cut-off date, thus spurring another round of unexpected retirements. The results of all of these difficulties are not trivial. They include too much reliance on pre-announced inspections (ASC 1988, 37) and lack of capacity adequately to supervise or manage the increasing amounts of work contracted out, including R&D, operations and maintenance. They also include uneven staffing of facilities employing controllers, resulting in overtaxed controllers and traffic delays at busy, high-cost areas such as Chicago, New York, Los Angeles, Boston, and Washington, D.C. (The Office of Personnel Management has approved a demonstration pro- ject through which controllers in the first three cities are being paid higher salaries. Additionally, the pay reform legislation enacted last year can provide further relief from this problem.) Another result of personnel shortages is a temptation to rely on inappropriate "hardware" solutions when more staff might have addressed a problem both more effectively and more economically (ASC 1988, 23). Procurement The Federal Property and Administrative Services Act of 1949, the Office of Federal Procurement Polièy Act of 1974 and other procurement laws, such as the Competition in Contracting Act of 1984, place heavy reliance on competition. They give losing bidders multiple opportunities to protest, thereby delaying decisions for long periods. However meritorious the objective of strengthening competition, the result is a cumbersome, ex- pensive, complex, and exceedingly slow process. There is broad agree- ment that the FAA's difficulties in managing its acquisitions are a significant part of its problem (OTA 1988, 6; ASC 1988, 20, 22; NAPA 1986, 20; Reason Foundation 1986, 6). Former DOT Secretary Burnley has said, "it takes years even in fairly simple, straightforward purchases [and that, as a result] high-tech systems may be technological relics by the time they are actually installed" (ASC 1988, 22). In combination with congressional hearings to expose allegedly unfair or unwise procurements, General Accounting Office (GAO) investigations and statutorily authorized second guessing of agency decisions, and agency
330 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION Inspectors General who naturally search hard for errors of judgment as well as process, these laws have made agency procurement officers virtual procurement "czars." These officials zealously protect the system even against worthy objectives like efficient, effective, and economical dis- charge of an agency's programmatic responsibilities. Procurement offi- cers' emphasis on awarding a contract to the lowest bidder, despite significant quality advantages with other bidders, is one example of how procurement and program objectives often clash. Some observers have placed a large share of the blame for the slippage and cost escalation of the NAS Plan on the procurement system. On the other hand, the GAO responded to a congressional request to evaluate such charges by concluding that the problems stemmed from poor planning and management by the FAA, and not from defects in the procurement statutes (U.S. Congress 1988, 68). The previously noted growth in reliance on contracting out has, of course, made difficulties in the procurement process (whatever their origin) far more significant to the FAA's overall performance. One consequence of the cumbersome procurement system, together with the funding limitations and uncertainties mentioned previously, is that it is hard for anyone to be held accountable for bringing in programs on time and within budget. The FAA has recently created a new position of Executive Director for Acquisition. While this action could alleviate some of the difficulties that the FAA has caused for itself in the procurement process, it can not, of course, relax the statutory rigidities. Some changes in the statutory procurement regime for the FAA were enacted last year affecting, primarily, the ability to enter into multi-year contracts. The FAA's Performance Despite (a) termination of about two-thirds of its 16,000 controllers in 1981; (b) lack of effective authority to shift controllers to high-cost lo- cations; and (c) burgeoning traffic under the stimulus of deregulation and competition, the FAA kept the ATC system functioning effectively 24 hours a day. According to the reports of the ASC and OTA, no evidence of a decline in safety during this period is apparent. Indeed, the data in their reports show continued improvement in the industry's safety record (ASC 1988, 6-11; OTA 1988, 4). It is wise to keep in mind, therefore, that the FAA has demonstrated that it can perform effectively an exceedingly difficult and vital mission with its current structure and authority. One should be certain that any
Appendix B: Organizational Options for the FAA 331 reorganization designed to alleviate the FAA's management difficulties does not unintentionally impair its ability to continue to perform reliably its uninterruptable and highly critical operational and safety regulatory functions. On the other hand, the FAA has not demonstrated the capacity to anticipate or respond to rapid changes in technology or the industry that it serves. Therefore, "muddling through" may not be good enough to deal with the future demands upon the agency. FAA Culture The organizational culture of the FAA has been characterized as quasi- military, that is, one in which the hierarchy is rigid, upward communi- cation is weak, and personnel are expected to do what they are told without challenge. A contributing factor to the PATCO strike was said to be the agency's rather poor record in employee-management relations, which is characteristic of such a culture. It is, perhaps, significant that a substan- tially new controller workforce hired after the strike fairly soon created a National Air Traffic Controllers Association to represent them, with about 70 percent voting in favor. Controllers work in somewhat different circumstances from those of many federal workers. They interact with aircrews far more than with other federal employees. They spend much time in the company of other controllers, and it is easy for them to come to believe that their working conditions are bad and that they are underpaid in comparison with flight crews. While these factors may help to explain the existence of employee relations problems, they do not excuse the FAA's tardiness in recognizing and seeking to deal effectively with them. This is underscored by the fact that the FAA has experienced serious labor-management problems with other groups of employees who are not subject to the special circumstances of controllers. Representative Guy Molinari (R.-N.Y.) complained about an unpro- ductive mindset in the FAA. He referred to the lack of aggressiveness in pursuing solutions to compensation problems, its being a reactive agency, its failure to acknowledge problems until they become critical, and difficulties in its approach to the NAS Plan (U.S. Congress 1988, 190, 191). The agency, to its credit, has been working for several years on the problem of its culture or operating style. It has engaged consultants in organization development and in total quality management (TQM) to assist it. A key feature of TQM is employee empowerment.
332 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION Response to Industry Changes Virtually all of the problems discussed in the foregoing sections probably contributed in some way to what has been perceived as the FAA's slow response to dramatic changes in the industry. Certainly, funding limits, personnel ceilings, and difficulty in reassigning controllers were important factors. Perhaps the most significant shortfall in performance by the FAA was its failure to anticipate and respond to the changes in the industry resulting from deregulation. For example, the creation of new hubs and the rapid expansion of existing hubs outstripped the FAA's capacity. This resulted in enforced delays in order to cope with the increased traffic. These delays, in turn, engendered concerns about safety and service. The FAA also was ill-equipped to deal with the restructuring of the industry. The burden of certificating many new carriers in the first phase of deregulation, or of adapting to the acquisitions and mergers of carriers in the ensuing period, taxed the FAA's capacity greatly. It had an insuf- ficient work force of inspectors and its practice of assigning regional responsibility for carriers based within the region proved to be very cum- bersome in connection with the acquisitions and mergers (ASC 1988, 36). While FAA management must share some of the blame for these per- formance difficulties, the general perception of the FAA's inadequacies was probably unfairly influenced by extraneous factors. The authority to approve airline mergers and acquisitions was, for a defined period, vested in the DOT which, during the Reagan administration, approved all 27 of those proposed and rarely required any modifications. It bears repeating that, although the FAA did not prepare adequately for traffic growth (whatever the reasons) it did keep air travel safe by enforcing delays (flow control) where needed for aircraft separation. Rulemaking Another area of allegedly poor performance involves the promulgation of regulations. As noted earlier, federal rulemaking is complex because of statutory due process requirements. The inherent delays have been en- larged by two of President Reagan's executive orders (E.O. 12291 and E.O. 12498) establishing a regulatory review process in the 0MB. This review also encourages the DOT to play a larger role in FAA rulemaking than it might otherwise, as the FAA's proposed rules must go through the DOT on the way to the 0MB.
Appendix B: Organizational Options for the FAA 333 Nonetheless, the process within the FAA has been criticized as being too protracted. The ASC observed that "more than 200 separate steps must be taken when all parties agree on the rule, it typically takes more than 2 years to get a final rule, and some issues [are] still unresolved after more than 15 years" (ASC 1988, 19). It has been noted that agencies often add a great deal of their own procedural baggage to what the statutes require of them (NAPA 1983). So slowness in response to evident concerns about aging aircraft, mis- fueling, smoke-and fire-retardant cabin interiors, and weakly mounted passenger seats should be charged in part to the FAA' s own deficiencies, whether they reflect procedural impediments or indecisiveness. Conflict of Roles Two other concerns have often been expressed about possibly conflicting roles assigned to the FAA. The first is between its regulatory role (cer- tificating or licensing airlines, aircrews, etc.) and its operational role (ATC). The second is between its responsibilities to promote aviation and to regulate it. The evidence that the latter conflict causes serious difficulties appears to be scant. However, several members of Congress, including Senator Wendell Ford (D.-Ky.) and Representative James Oberstar (D.-Minn.), have proposed legislation to remove Section 305 of the Federal Aviation Act, the so-called promotion function, from the FAA (but see section on the FAA's Mission and Status, in which it is noted that Section 103 is equally relevant). The ASC noted the possible pressure to relax regulatory standards for economic reasons (ASC 1988, 24-25). OTA's project director also expressed concern about the conflict between regulating carriers and fostering their economic interests. She referred both to problems with commuter airlines and to those involved in the bunching of airline departures at desirable hours (U.S. Congress 1988, 31, 34). The ASC saw some problems resulting from mixed incentives, for example, in allocating resources or in trying to please the White House and the 0MB, in an agency which both performs safety functions itself and regulates the performance of others (ASC 1988, 24). These role- conflict issues are taken into account in the following evaluation of or- ganizational options.
334 WINOS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION ALLEVIATION OF PROBLEMS WITHOUT REORGANIZATION The preceding section reflects broad agreement that there are serious problems both in the FAA's performance (such as in regulatory delays and in responding to industry changes) and the statutory and political constraints that impair its response to the needs of a dynamic industry and the opportunities of new technology. While reorganization has been pro- posed to remedy these deficiencies, it is never accomplished without some difficulty and the investment of resources, as well as the risk of unintended consequences. Therefore, one should be certain that the problems to be addressed can not be adequately treated within the existing structure before turning to legislation to effect a reorganization as the remedy. Following is a discussion of the problems identified in the earlier section on Problems Confronting the FAA. In each case, account is taken of the conclusions reached by the other groups that have studied and made rec- ommendations about the FAA's status. Finally, attention should be given to whether there appears to be any risk that the FAA's operational capability might be weakened by a pro- posed reorganization. This last concern is addressed, wherever relevant, under the Advantages and Disadvantages captions in a later section on Recent Proposals to Reorganize the FAA. Tenure of Administrator Longer tenure could theoretically be achieved by a "contract" between the President and the Administrator. However, the tenure of the Secretary of Transportation and the fact that such a contract could not bind the President's successor must also be considered. Thus, longer tenure can realistically be achieved only through legislation providing a fixed term for the Administrator. Experience with fixed-term positions (see Exhibit 1) clearly demonstrates that average tenure is well over double that for other PAS appointees. In fact, it approaches the length of the statutory term. Reappointment to such positions is not uncommon. DOT Micromanagement DOT Secretary Samuel Skinner has greatly improved the relationship between the DOT and the FAA. However, department budget, manage- ment, policy, legal, personnel, audit, and other staffs will always have
Appendix B: Organizational Options for the FAA 335 their jobs to do and, under the most supportive Secretaries of Transpor- tation, the FAA has been subjected to a good deal of interference by such staffs that goes beyond responsible oversight. Micromanagement, thus, is likely to remain a problem so long as the FAA is a full-blown constituent agency of the department. Political Gaming Whether an independent agency or a unit of the DOT, the FAA will be subjected to the problems of political game-playing, some relating to the 0MB and the DOT, and some to Congress and the executive. Only as a self-financed entity, not subject to personnel ceilings and appropriations, can the FAA be freed from these problems. Perhaps more politically sophisticated leaders in the FAA and more understanding managers in the DOT could marginally alleviate the problem, but the prospects for the FAA's escaping from political pressures are not encouraging. Regulatory and Other Statutes The basic rules for the FAA are and will remain those prescribed by law. The FAA has no ability to avoid the strictures in authorization, appro- priations, and other laws. Adequacy of Funding The deficit problem will not likely be corrected in the foreseeable future. Under the budget summit agreement, agencies will generally be able to have their funds increased only to the extent that other agencies' programs in their part of the budget are reduced. Substantially increased funding can, thus, be achieved only through the conversion of the FAA to a self- financed entity. Of all the reforms proposed for improving the FAA's ability to cope with its current and anticipated problems, granting it fi- nancial independence is probably the most important. Delayed Funding As noted above, only freedom from the appropriations process will give the FAA the assurance of predictable and timely funding.
336 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION Furloughs Stability in staffing levels will not be achieved as long as the FAA is subject to the budget and personnel ceiling controls administered by the 0MB. Only as a self-financed entity can the FAA expect to have control over its staffing level. Human Resources Even with the new pay reform legislation, there are numerous and serious shortcomings in the Civil Service laws. Under the new legislation, the administration has determined to pay geographical differentials in only a few cities, not including Washington, D.C. While the FAA could probably do a much better job in human resources planning, training, management development, performance evaluation, incentive systems, and the like (as could other agencies), some of the basic statutory restrictions, such as those with respect to hiring, discipline, incentives and firing, are beyond its discretion to change. With authority for a separate personnel system it could develop substantially different policies and practices for its diverse types of employees. Procurement There are possibilities for improvements in the way the FAA determines its requirements, converts them into an acquisition strategy and plan, and proceeds to contract for, test, evaluate, produce, and install the equipment. The appointment of a new Executive Director for Acquisition may help to streamline and better coordinate this complex process. But personnel and funding restrictions (discussed previously) contribute to the agency's poor planning and management in the area of procurement as well as in other areas. A bureaucratic culture that rewards those who do not stick out their necks, and not innovation, is a part of the problem. Moreover, no matter how much better the FAA plans its procurements, there are statutory impediments prescribing policies and practices, which give rise to many of the difficulties. Prominent among these are insistence on advertising and competition, even if only one or a few bidders are qualified; em- phasis on choosing the lowest bidder when others may offer superior quality and value; prohibitions on working with contractors during the
Appendix B: Organizational Options for the FAA 337 specification of follow-on procurements; and complex, time-consuming procedural requirements, including those affording allegedly aggrieved, unsuccessful bidders the right to multiple and protracted appeals. Culture Change Changes in organization culture are very slow. The FAA has a number of commendable efforts under way, but significant progress may easily take 5 years. A really major change in attitude and behavior depends on changing the reward and incentive system from a bureaucratic one that rewards those who do not make waves to one that encourages creative and innovative behavior. The challenge is difficult. One needs an organization whose values will continue to emphasize conservatism when it comes to the safety of aircraft operations, while seeking in- novation and dynamism in responding to industry and technology changes. That can best be achieved by a change to a corporate-type entity. Response to Industry Changes As with procurement and human resources, discussed above, there is plenty of opportunity for improvement in the FAA's performance. Seizing this opportunity, however, is also constrained by human resources and funding limitations. A self-financed entity, with the ability to respond quickly to the rapidly changing requirements of its rate-paying customers, is far more likely to be attuned to their needs because it will know that it can and must meet them. Rulemaking Much of the delay in the regulatory process is caused by the agency's own excessive caution and unnecessarily cumbersome procedures. Fun- damental change in the approach to the regulatory process is far more likely to occur in a new agency, whose statutory autonomy and flexibility would give it the confidence and the incentives to depart from existing methods of operation.
338 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION Conflict of Roles Removal of the alleged conflict between promotion and regulation requires legislation. The conflict between the FAA's operational and regulatory roles can best be addressed by administrative action rather than by leg- islation. That is, unless there is a realistic possibility of spinning off the ATC functions from the FAA's other safety functions, in which case legislation would be required. Summing Up With one exception (an organizational separation of operational from reg- ulatory functions) none of the problems listed above is likely to be sub- stantially alleviated except through the enactment of legislation. With two exceptions (a fixed-term for the Administrator and separation of the pro- motional and regulatory roles) legislative change is likely to be achieved only as a part of a significant reorganization. Based on the experience with other independent agencies, conversion of the FAA to independent agency status would probably not be accompanied by the kinds of statutory relief that are badly needed. Exhibit 1 contains a discussion of the ex- periences of the National Aeronautics and Space Administration (NASA), including its general lack of freedom from the restraints affecting the FAA, its performance difficulties, and the vagaries of the budget and appropriations process as they have affected it. NASA was selected be- cause it has a number of common problems with the FAA, although there are also substantial differences in the character of the two agencies. GREATER FLEXIBILITY VESTED IN OTHER FEDERAL AGENCIES A review was carried out through published sources, as well as inquiries of knowledgeable persons and the search of statutes, of various kinds of management flexibility accorded to members of several classes of federal agency. This was intended to illuminate the probability that such statutory freedom would be given to an independent agency, one within a depart- ment, or a government corporation. Because of its length, the results of that review are provided in Exhibit 1. The citation of examples in the exhibit is not intended to suggest that any of the agencies are analogous to the FAA. Indeed, in most significant respects, the FAA is unique. No
Appendix B: Organizational Options for the FAA 339 other federal agency has the responsibility to operate a large-scale, na- tionwide, complex system in a hazardous environment 24 hours a day throughout the year. The examples should be helpful in predicting whether the kinds of freedom from constraints are more likely, or likely at all, to be given to the FAA or to ATC (a) as it stands, (b) as an independent agency, or (c) as a public corporation. (It is assumed that an ATC corporation would have virtually complete flexibility if it were privatized.) For convenience, Table B-i shows which agencies and corporations have exemptions from several types of statute, or have special systems. This table does not reflect all of the provisions discussed in the exhibit, but the lessons that can be drawn from all of the material in the exhibit are discussed in the remaining sections of this appendix. Based on a review of the provisions cited in Exhibit 1, some of which are referenced in Table B-I, one can conclude which types of flexibility or freedom from constraints are more or less likely to be afforded to a given type of federal entity. Following are the conclusions that can rea- sonably be drawn from that review. Tenure The heads of independent agencies, agencies within departments, or cor- porations all appear equally likely to be given fixed terms if a case is made for the need for longer tenure. Provided that no limits are sought to be placed on the President's power to remove the appointee, such a proposal is not likely to be controversial. Financing Neither independent agencies nor agencies within departments are likely to be freed from the normal budget and appropriations processes. (An exception would be a requirement that the agency's initial budget request to the 0MB be furnished to Congress, but such a procedure is not likely to change outcomes much, if at all.) The prospects for allowing a self- supporting government corporation to be funded outside of those processes are very good, especially if one proposes to make the corporation subject to the' usually nonintrusive provisions for 0MB and congressional over- sight included in the Government Corporation Control Act.
340 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION TABLE B-I EXEMPTIONS FROM SELECTED MANAGEMENT STATUTES Statutes Agency Personnel'4 Procurement Appropriations0 Independent CIA x Federal Reserve Board x x x GAO x Nuclear Regulatory Com- mission (some) Smithsonian Institution (some) x NASA et al. None None None Within departments FBI x Foreign Service x Military intelligence x Power Administrations (DOE) x Public Health Service x U.S. Attorneys x Veterans Affairs (some) Corporations Commodity Credit Corp. x x Ex-Im Bank x FCIC x x x FDIC x x x FHA x Federal Prison Industries x GNMA x x OPIC x x Seaway Corporation x TVA x x x Postal Service x x x °Exemptions may be from pay or classification laws, or both. Some have separate systems prescribed by law. bSome receive appropriations for non-revenue-producing activities, or to cover losses. Human Resources Except for authorization to pay a limited number of experts salaries above the normal ceilings (which the FAA already has) there is little prospect for giving either an independent agency or an agency within a department special personnel authorities. Further, the likelihood of such action has been substantially reduced by the enactment of pay reform legislation last year. A government corporation is a more likely candidate to be given authority to establish an entirely separate personnel system.
Appendix B: Organizational Options for the FAA 341 Procurement Neither an independent agency nor an agency within a department is likely to be freed from the procurement statutes that cover almost all such agen- cies. While federal corporations are far more likely to be exempted, such action cannot be confidently predicted for such an FAA corporation. Micromanagement An independent agency would not, of course, be subject to microman- agement by a department, but might be more subject to detailed inter- vention from the 0MB, White House staff or cabinet councils. Agencies within departments are inherently vulnerable to intervention by depart- mental staff, even when the secretary is not so disposed personally. A government corporation, even if subject to prescribed policy supervision by a cabinet secretary, is most likely to be free from the worst features of secretarial or departmental micromanagement, particularly if (as sug- gested below) the secretary's powers with respect to the corporation are made nondelegable. Congressional micromanagement can be substantially mitigated only through freedom from the normal appropriations process, because that process affords Congress the most potent opportunities for intervention in agency management. Therefore, as noted under the paragraph on financing above, a government corporation has the best chance to escape that burden. RECENT PROPOSALS TO REORGANIZE THE FAA The section on the FAA's problems identified the sources of several proposals regarding the FAA's status (namely, the ASC, Heritage Foun- dation, NAPA, OTA, and Reason Foundation). The leading proposals are to (a) leave the FAA where it is, but strengthen its capacity; (b) make it an independent agency again; (c) reassign the ATC functions to a self- supporting government corporation; (d) privatize ATC functions; and (e) convert the entire the FAA to a self-supporting government corporation. The following sections briefly characterize the five proposals and list their principal perceived advantages and disadvantages. Generally, the various proposals focus on the FAA solely at the "corporate level." That is, they do not address the distribution of authority between headquarters and the field, or changes in the field structure or other operational duties
342 WINDS OF CHANGE: DoMEsTIC AIR TRANSPORT SINCE DEREGULATION TABLE B-2 PROBABLE EFFECTIVENESS IN CORRECTING OR PREVENTING PROBLEMS (ASSUMING ENACTMENT OF PROPOSAL) Status Independent FAA ATC Corporation Problems Quo Agency Corporation Public Private Financial independence Poor Poor Excellent Excellent Excellent (with right to borrow) Tenure Fair Good Excellent Excellent Excellent Compensation Fair Fair Excellent Excellent Excellent Personnel ceilings Poor Poor Excellent Excellent Excellent Procurement Poor Fair Good Good Excellent Micromanagement DOT Fair Excellent Good Good Excellent Congress Poor Poor Excellent Good Excellent Political gaming Poor Fair Excellent Good Excellent Regulatory Fair Good Good Good Excellent Culture change Fair Fair Excellent Excellent Excellent Role conflicts Fair Fair Good Excellent Excellent Maintaining system safety Good Good Excellent Fair Fair oExcept for the last entry, relating to maintaining system safety, probable effectiveness applies only to the spun-off ATC entity; performance of the remaining functions would not necessarily be improved at all, and might even be adversely affected, for example, by having inferior benefits or opportunities for the employees involved. or relationships. Exceptions are the provisions in Representative Oberstar's independent agency bill that would change the FAA regional structure (H.R. 1633) and the proposed transfer to airports of respon- sibility to operate some or all control towers under the privatization proposals. For the most part, this discussion focuses on the merits of each proposal. That is, it assumes the proposal can be put into effect. Table B-2 compares the five proposals in terms of their likely long-term effectiveness in the correction of selected problems, also assuming that the proposal in question is actually adopted by Congress and the President. Judgments about the likelihood of enactment are contained under the section on Observations On Political Feasibility. The assessments in the table take into account
Appendix B: Organizational Options for the FAA 343 the changes that may be reasonably anticipated in the character and com- plexity of the FAA's future operations. Leave In the DOT But Address Most Serious Difficulties A report by OTA, Safe Skies for Tomorrow, is perhaps the fullest ex- position of the contention that the FAA should remain in the DOT and retain its current status there, although the report does not contain an explicit recommendation to that effect (OTA 1988). Similarly, the study's project director essentially supported the retention of the FAA in the DOT in her 1988 testimony before the House Subcommittee on Aviation (U.S. Congress 1988). OTA urged that steps be taken to improve the FAA's capacity to function effectively. For example, OTA (1988, 7) recommended that Congress hold "the Administrator accountable solely for safety," allocate "stable and adequate funding sources from the General Fund and the Airport and Airway Trust Fund," and give the Administrator "a fixed term, perhaps up to 5 years." Advantages The national interest is best served by having a single transportation agency concerned with all modes of transportation, which does have to make trade-offs in allocating priorities and funds among federal trans- portation programs; Secretary Skinner's February 1990 report on national transportation policy, Moving America, illustrates the potential benefits (DOT 1990). A number of national and international policies that affect the FAA's operations (such as national security, macroeconomic policy, and inter- national relations) can best be interpreted and applied through a department secretary; without the strong backing of Secretaries Volpe and Lewis, it is doubtful that the FAA's tough position in two controller labor disputes could have gained White House backing. Efficient federal rianagement requires a reasonable limit on the number of agencies reporting directly to the President; the FAA is but one of many agencies for which independent status is sought, such as the much larger Social Security Administration. An agency within an important department, such as the DOT, has the Secretary to "go to bat" for it at the White House and the 0MB; as
344 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION an independent agency, it is likely to have less clout, and probably be subject to greater control by the 0MB or the Domestic Council. The FAA has recently undergone a number of internal reorganiza- tions (and may have undergone another, as the Administrator has elected, for now, not to fill three vacant Executive Director positions); some forms of reorganization could conceivably disrupt operations. Leaving the FAA where it is would assure that there is no disruption of ongoing ATC and safety functions. Disadvantages I. There is very little prospect that problems that stem from govern- ment-wide legislation, such as personnel and procurement, can be sub- stantially alleviated without a change in the FAA's status. (As noted earlier, an exception is the demonstration project allowing area differ- entials for controllers, which was finally approved more than 10 years after the FAA first requested the former Civil Service Commission for a similar arrangement under previous legislation. Legislation has also been enacted to introduce locality pay across-the-board for federal agencies.) In particular, OTA's recommendation that the FAA be allocated more stable and adequate funding (without a change in its status) seems wholly unrealistic in the light of the federal deficit situation, which is most unlikely to be improved during the next 5 years, at least. It is very unlikely that it would ever be allowed to leverage its funds by having authority to borrow. Although the duration of intensive DOT micromanagement to date seems to have been coincident with the tenure of the last two Reagan appointees as DOT Secretary, there is no assurance that these conditions would not reappear under another Secretary; further, DOT staff have often been vexing to the FAA under supportive secretaries. The possibility of substantially changing the FAA'S mindset or cul- ture in the short term is very slight without a significant change in its status, authority, and reward and incentive systems. Summing Up It seems unlikely that any significant statutory improvements will be achieved while the FAA remains an agency within the DOT, except for a fixed term for the Administrator and the separation of the dual roles of promotion
Appendix B: Organizational Options for the FAA 345 and safety. As mentioned earlier, legislation has been introduced to achieve the second change. Although it appears to have been drafted imperfectly, Congress or the executive branch could readily draft a suitable bill to transfer the promotion functions that are vested directly in the FAA to the Secretary of Transportation. If one concludes, therefore, that more stable funding, significantly increased statutory flexibility, especially regarding personnel and pro- curement, and a near-term change in the FAA's culture or mindset are essential to the achievement of an effective FAA, then a change in its status must be sought. The risk of losing the desirable coordination with, for example, other transportation policies or programs, can be addressed in designing an alternative status and governance structure, as discussed below. Independent Agency At the 1988 House hearings, former Administrator Halaby testified on behalf of an impressive group of aviation industry organizations in favor of making the FAA an independent agency, which would have a sta- tutory advisory committee, but no governing board (U.S. Congress 1988, 276). His testimony carries substantial weight since he was Ad- ministrator at the time the FAA was included in the DOT and strongly supported that change. Legislation to make the FAA independent has been pending for sev- eral years. Hearings have been held in both Houses. The Senate Com- mittee reported a bill in 1988, but no action was taken by the Senate. Although it was reintroduced in the 101st Congress, there were no hearings on it. The 1989 Senate bill, S. 1600, introduced by Aviation Subcommittee Chairman Ford, would have given the Administrator a 7-year term and raised the pay level to that for department heads. (5. 1600 in the 100th Congress provided two Deputy Administrators, one for Safety, but the 101st Congress bill provided a single deputy.) It would have provided some additional flexibility in personnel and procurement, but no broad exemptions. It would have converted the trust fund to a revolving fund, but still required annual appropriations. In the House, Representative Oberstar's 1989 bill, H.R. 1633, pro- vided a 5-year term, and authorized separate personnel and procurement systems. It also proposed to exempt the FAA from the 0MB' s regulatory review.
346 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEEGULATION Advantages It would free the FAA from micromanagement by the DOT. Its Administrator would have more visibility and freedom to argue the case for its programs with Congress and the public. Its legislative and funding requests would not have to compete within the DOT for priority; personnel ceilings imposed by the 0MB would not be traded off against other DOT programs. The President, the 0MB, and the Domestic Council would be more directly subject to criticism for shortchanging the FAA since the blame could no longer be shared with the DOT. Legislation to make the FAA independent would very likely include a fixed term for the Administrator. Converting the whole of the FAA to an independent agency would pose no threat to the continuity of ATC and safety functions. Disadvantages It would likely weaken or complicate transportation and other na- tional and international policy making; the agency would be subject to, perhaps, undesirable pressures or tendencies to advance the interests of those it regulates and serves, even if contrary to the national interest. It would sharply reduce the possibility that the DOT could success- fully press for safety improvements recommended by the NTSB, but resisted by the FAA. (Of course, this is a problem only if the NTSB's position is sound.) Except for a fixed term for the Administrator, and separation of the promotion and regulatory roles, such legislation (by the time it is subjected to executive branch reports and testimony) is unlikely to address, mean- ingfully, any of the other difficulties facing the FAA. In particular, fi- nancial stability and flexibility, including borrowing authority, is not likely to be gained. An independent FAA could, again, be a "weak sister" in the bu- reaucratic scheme, and might also be subject to greater micromanagement by the Congress. Changes of this magnitude can be made very infrequently; the new status would likely last for a decade or more, even if no improvements are evident, unless some crisis should arise to put the matter high up on the public agenda.
Appendix B: Organizational Options for the FAA 347 It would not likely accelerate the kinds of cultural change that seem to be desirable. It would greatly weaken the DOT as an important major-purpose department; it could encourage efforts to remove the Coast Guard as well. Summing Up An independent NASA has suffered many of the same ups and downs in presidential and congressional support as has the FAA. Making the FAA an independent agency, without much more, would probably prove to be unhelpful, and could even be harmful both to aviation and to broader national interests. It would virtually preclude attention to more effective remedies for the foreseeable future. Government Corporation To Operate ATC In 1985 the ATA proposed that ATC be splitoff from the FAA and assigned to a user-funded government corporation. Significantly, in order to discourage opposition from general aviation, the ATA proposed that the legislation "grandfather" the existing proportionate shares of user taxes. The proposed National Aviation Authority would be headed by an Administrator appointed to a 10-year, nonrenewable term. There would be a statutory advisory board, but no governing board. Two consultants prepared a report for the ATA in 1985, A Study of the Safety/Regulatory Functions of the Air Traffic Control System. They con- cluded (Belanger and Newpol 1985, vii): that the proposed establishment of a National Aviation Authority would not have an adverse impact on the operation of the Air Traffic Control System provided the regulatory authority for the airspace that makes up the system is retained by NAA and safeguards regarding coordination such as in FAR 11 [notice and hearings requirements] are incorporated in the legislation. They proposed a number of different dispositions of authorities now assigned to air traffic, aviation standards, or airports. Some would remain where they now are; some would be reassigned; and some would be split. As to others, the consultants said the functions could go to either of two units. Throughout their report, there were situations where they said that coordination and concurrence would be essential. The report does not
348 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION detail how that coordination would be achieved, except as to reliance on the notice and hearings requirements of the Administrative Procedure Act. The ATA requested that a NAPA panel study the desirability and fea- sibility of its ATC corporation proposal. The NAPA panel concluded that such a split of FAA's responsibilities was not wise and recommended that the whole of the FAA be converted to a self-financing government cor- poration (NAPA 1986, 23-26). [Although the ATA continued to support the ATC corporation proposal (Journal of the Air Traffic Control Asso- ciation 1987, 8, 9), it later joined with the industry associations that supported former Administrator Halaby's 1988 testimony in favor of mak- ing the FAA an independent agency.] Legislation to carry out the ATA proposal was introduced in 1987 by Senator Daniel Inouye (D.-Hawaii), by request, S. 1159. It was not re- introduced in the 101st Congress. Then DOT Secretary Burnley recom- mended such a reorganization in his House testimony, but he furnished no details as to the proposed powers, structure, or governance of such a corporatipn (U.S. Congress 1988, 240). Nor did he discuss the matter of coordination between such a corporation and the government agency in which the other the FAA functions would be vested. Advantages According to the 1981 NAPA panel, the principal rationale for cre- ating a government corporation is that it be revenue producing and po- tentially self-sustaining through charges to those it serves (NAPA 1981, 27, 28); clearly, this applies to ATC. Such a corporation would have control over its financing, including the authority to leverage its funds through borrowing. Changes in the culture, mindset, incentive systems, and operating style of ATC personnel are far more likely to occur through a dramatic change in status, as would be involved in converting ATC to a self-financed corporation. The bulk of the FAA's activities, including both personnel and money, is involved in the development and operation of ATC; corporate status would offer the prospect of alleviating the problems discussed earlier for that large portion of the FAA's programs. Any conflict between the FAA's operational and regulatory roles would be resolved since the regulatory function would remain in the DOT; in fact, the remaining FAA could perform a more meaningful safety oversight role with respect to ATC than is now performed by anyone in the FAA.
Appendix B: Organizational Options for the FAA 349 Leaving regulatory and other functions in the FAA would permit them to be effectively coordinated with other transportation policies and programs. Disadvantages Safety functions were once split between the CAA and the CAB and were combined in the FAA, in part to remedy coordination problems. OTA (1988, 66) concluded, "The ATC function is inextricably linked with aviation safety and is a central component of an integrated safety system." A NAPA panel reached the same conclusion (NAPA 1986, 16). Since ATC claims the bulk of FAA personnel and money, the re- mainder might be too small or weak to be effective; improving ATC's status might weaken prospects for similarly needed improvements for the residual FAA. Of all the options being considered, this one raises the most serious possibilities for substantially disrupting a complex program which, despite major obstacles, has proven to be safe and reliable; even some of the supporters of the proposal have pointed to many cases in which close coordination and concurrence would be required across agency lines. Summing Up Clearly, there are significant coordination requirements among ATC, air- ports, aviation standards, regulation and certification, systems develop- ment, and R&D, which supports all of the foregoing. These functions now report to three different Executive Directors and one Assistant Ad- ministrator in the FAA. If adequate coordination can be achieved across these lines internally, in theory, it could also be achieved across agency lines. In practice, coordination across agency lines is likely to be far more difficult even if (as would be appropriate) some formal oversight roles were assigned to the residual FAA in regard to the new ATC corporation. On the other hand, the case could be made that an organizationally separate regulatory overseer of ATC would enhance safety because the overseer would not be faced with the same pressures to make compromises. However, the greater difficulty of achieving interagency coordination is, perhaps, un-
350 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION derscored by the testimony of OTA's project director that existing coor- dination within the FAA is not satisfactory (U.S. Congress 1988, 44). Before recommending that the FAA be split along these lines, one would want to have a detailed study of the way in which existing coordination problems are handled, as well as an analysis of new coordination problems that would arise. This would be essential in dividing responsibilities be- tween the two resulting organizations, and in fashioning coordination mechanisms, so as to assure that safety would not be jeopardized. The kind of analysis required for confidence that the split would be practicable has not yet been offered by the supporters of the proposal. Such a study was beyond the scope of this review of current proposals. Privatize ATC There have been proposals by the Reason Foundation (1986) and the Heritage Foundation (1982) to assign ATC to a private corporation, pref- erably structured as a user-owned cooperative. The Reason Foundation proposal also suggests that the co-op might contract out, through com- petition, the operation of all or parts of the national system and that airports be allowed to operate their own control towers. The 1987 President's Commission on Privatization recommended privatizing or contracting out major portions of the ATC system. Contracting out should be distinguished from privatization. Under the former, the contracting agency retains full authority and responsibility for the program. Its contractor is merely an agent, to which specified authority is delegated and to which funds are remitted for the cost of performing the delegated functions. A few ATC facilities are already operated under contract with the FAA or the Air Force. Since the FAA, under several of the options discussed in this paper, could use contracting out as a method of executing more or all of its ATC functions, contracting out is not considered further as a separate organizational option. Under privatization, the authority and responsibility for carrying out a program is actually divested by the government and it is up to the private corporation carrying out the program to finance the operations. The government can still play a role. For example, the Communications Satellite Corporation (ComSat) is a private corporation that is chartered by Congress. The enabling statute vests the President with the authority to appoint three members of its board of directors. ComSat's rates for certain services are subject to regulation by the Federal Communications Commission.
Appendix B: Organizational Options for the FAA 351 There are a number of precedents for private operation of ATC, often by nonprofit, user-owned companies. Aeronautical Radio, Inc. (ARINC) originated ATC services in the United States and performed them during 1935 and 1936. There are now privately-run ATC services at a number of smaller airports. Some foreign countries have, or had in the past, private companies operating ATC. Examples are International Aeradio (a British Airways subsidiary) in Commonwealth countries, the Middle East, and the Caribbean; Lockheed and Bendix under successive five-year contracts to Saudi Arabia; and two non-profit corporations in Mexico and Cuba, which have since been nationalized. Radio Suisse (nominally a private company, but 96% of whose shares were owned by the government) operated ATC services in Switzerland from 1931 until 1987 when it was replaced by SWISSCONTROL, a private company. In the Netherlands, a proposal to privatize ATC and commu- nications is to be taken up by the parliament this year, with approval anticipated. In the United Kingdom, the Civil Aeronautics Authority must compete with private companies for operation of airport control towers. Reassigning ATC functions to a private corporation, while leaving the rest of the FAA's functions in it or a successor federal agency, would raise issues of coordination with airports, aviation standards, regulation and certification, systems development, and R&D, similar to those that were detailed in the preceding section, relating to a public ATC corpo- ration. As noted there, except for reliance on a notice and hearings re- quirement, no one has spelled out how these vital coordination issues would be resolved. It was, therefore, concluded that such a split would give rise to serious concerns about the possibly adverse effects on safety and that it should not be considered further until a detailed study. and analysis was conducted to assure that such adverse effects could be avoided. Other issues arise, but they can be addressed by familiar methods. For example, a principal criterion for privatization is whether the function is one in which competition could be expected to develop and survive. However, ATC has the characteristics of a natural monopoly. The Reason Foundation proposed that the user-owned nature of the nonprofit company would protect the users and the public from excessive charges and from "gold plating." However, an issue would surely be raised as to whether the public could rely on this corporate form to assure that user fees were fair and reasonable and, accordingly, that the rates charged the traveling publië would be fair. Hence, like other public utilities, its rates might have to be subject to federal regulation. Similarly, its operations should be subject to federal oversight or regulation from a safety perspective. There is ample precedent for federal safety regulation of private entities,
352 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION such as that by the Federal Railroad Administration and the U.S. Coast Guard. A New Option In light of the difficulties posed by splitting ATC from the remainder of the FAA's functions, another possibility has been proposed. It is to create a private ATC corporation but to assign the remaining functions of the FAA to a public entity embedded within it. Specifically, the private cor- poration would be user-owned and would be financed by user charges. The public agency within it would be vested with virtually all other FAA functions (although a few functions might be left in the DOT), preserving precisely the existing allocations of functions within the FAA. That entity's funds would be provided by the corporation, but in such amounts as would be annually prescribed by the 0MB. This body would have the authority to oversee the corporation and regulate its operations from a safety perspective, and might also be given the authority to regulate the rates charged by the corporation. These functions might be analogous to the role of an inspector general or to that of an independent auditor. The closest precedents for such an arrangement are probably the Federal Energy Regulatory Commission (FERC) and the Postal Rate Commission (PRC). In both cases, however, there are significant differences. FERC is within the Department of Energy, which is a purely federal agency, and FERC's regulatory functions apply to private utility companies, not to those of the Energy Department. PRC performs a regulatory review function regarding the rates of the U.S. Postal Service (USPS). Although it is an independent regulatory agency outside of USPS, its budget is included within that of the Postal Service. However, it is subject to direct appropriations from Congress and is not financed by USPS. Advantages I. ATC is like many other public utility services that are typically offered by the private sector; it can readily be self-sufficient through charges levied on users of the system; thus it would be free from all of the budgetary problems of the federal establishment, those that are pro- cedural as well as those that stem from insufficient resources. It could plan effectively for changes in the industry and in technology. Through
Appendix B: Organizational Options for the FAA 353 borrowing, it could leverage the funds which it would raise from user fees. A private corporation would be free from both executive and leg- islative micromanagement. It could use the best practices of the business sector for its pro- curement and personnel systems, and thus acquire both high quality per- sonnel and equipment. Any conflicts between the FAA's promotional and regulatory roles or between its regulatory and operational roles would be removed as the ATC corporation would not have promotional functions; its regulatory functions would be limited to those inherent in operating the ATC system. The tenure of its head would likely approximate the substantially longer tenure of other private companies that are similar in size; together with its independent financing, this would result in a corporation that could do effective long range planning. Costs would be controlled effectively; any tendency by the corpo- ration to build a "gold plated" system would be countered by its user- owned character and governance structure (or by a federal rate-review process). Existing coordination of the relevant functions within the FAA is achieved largely through informal means. Through continuation of the current division of responsibilities and with the organizational and physical proximity of the two entities, close coordination between ATC functions and related functions of the governmental body could be achieved. Perhaps most important, a private corporation and its personnel would be freed from the mixed, and often perverse, incentive systems inherent in a federal agency and would be able to build a culture based on rewards and penalties that are genuinely related to innovation and performance. Disadvantages The non-ATC functions (such as safety regulation, and rate regu- lation if that should be a feature of the reorganization) are, however they would be characterized, inherently public; creating what amounts to a public agency within a private corporation is unprecedented and, therefore, raises novel challenges to which there are not clearly satisfactory answers. If the entity performing the public functions is fully independent from the ATC corporation (as it must be in order to avoid constitutional problems) then there would seem to be basically a landlord-tenant rela-
354 WINOS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION tionship. In that case, the so-far unresolved problems of how to assure coordination between the safety and operational functions of the public entity and those of the private corporation would have to be confronted. The proposal to maintain the existing division of functions does not address how any disagreements would be resolved, as there would be no one in charge of both parts of the new structure. Thus, there could be some jeopardy to the continuity and reliability of the allegedly intertwined safety functions of ATC and regulation. 3. Additional problems would be presented in developing and imple- menting a user fee schedule adequate to finance the corporation, partic- ularly with respect to continuing or ending the apparent subsidy to general aviation in its use of the ATC system. Summing Up The possibility of creating a private corporation to perform ATC functions is attractive because, insofar as those functions are concerned, it would clearly provide an effective remedy for the vexing problems that have confronted the FAA. (As shown in Table B-2, it appears to be clearly superior in many respects to the other options.) It is the only option that could assure the replacement of a federal agency culture and modus op- erandi with the incentive systems of a private company. However, the difficulties presented by the splitting of the FAA would seem to make such a proposal premature. Perhaps a careful study of the possibility and means of separating ATC from other FAA functions would demonstrate that this could be done efficiently and without jeopardizing safety. However, that study has not yet been conducted and it was beyond the scope of this paper. Convert To Government Corporation Two groups that have addressed the issue of how to strengthen the FAA's status have recommended that the entire agency be converted to a self- financed government corporation. The first was the 1986 NAPA panel that was commissioned to evaluate the ATA proposal for a separate ATC corporation (NAPA 1986). The second was the congressionally chartered, presidentially appointed ASC, which reported in April 1988 (ASC 1988). The NAPA panel proposed that there be a single administrator appointed by the President and that there be an advisory board, but no board of
Appendix B: Organizational Opt ions for the FAA 355 directors. Further, the corporation president would be subject to the su- pervision and direction of the Secretary of Transportation. Such a role for the Secretary would, of course, include the review and approval of the all-important schedule of charges to be levied by the corporation on its customers. The NAPA panel's proposal for a link to the Secretary of Transportation seeks to confer the advantages of corporate status while preserving some of the benefits of leaving the FAA where it is. The ASC proposed that the FAA be converted to an "authority" gov- erned by a nine-member Board of Governors. The board would be chaired by the Administrator and would include a Director of Aviation Safety, both of whom would be appointed by the President for 7-year terms. The Secretaries of Transportation and Defense would be on the board, ex officio, and there would be five other presidentially appointed members. One would be from the public at large and four would be selected so as to assure consideration of the needs of "various interests such as airline passengers, air carriers, regional airlines, general aviation, aircraft man- ufacturers, airport operators, and airline labor" (ASC 1988, 29). Under the ASC's proposal, the Director of Aviation Safety could par- ticipate in the Administrator's rulemaking process and disapprove pro- posed rules, as well as initiate and promulgate rules. Either official could appeal rulings of the other to a five-member Safety Committee of the Board of Governors, to consist of those two officials, the two department secretaries, and the public member of the board. It should be stressed that the simple creation of a corporation by statute endows it with only those powers or exemptions that are expressly in- cluded. While there is ample precedent for having the Congress pass and the President sign the kind of legislation proposed by these two groups, there is no assurance of success either as to passage or as to the provisions which may be included. To date, no legislation has been introduced along the lines of either of the two proposals for a corporation or authority. In the United Kingdom, ATC en route functions are performed by the Civil Aeronautics Authority, a quasi-public corporation that competes with private companies for operation of airport control towers. The authority is subject to policy direction from the Department of Transportation. It must recover its costs from users, but its charges are subject to disapproval by the government if they are found to be too high. A number of other countries are converting, or have recently converted, their civil aviation agencies to public corporations which are subject to policy direction or other forms of influence or control by the government ministries. These include Australia, New Zealand, Finland, Sweden and Portugal.
356 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION Advantages The self-supporting criterion for corporate status advanced by the 1981 NAPA panel was deemed to be suitable for the whole of the FAA by the 1986 NAPA panel. A user-financed government corporation would offer the maximum prospect for changing the reward and incentive systems and, therefore, the organizational culture of the whole of the FAA. Po- tential micromanagement by the DOT or by Congress could be eliminated or sharply curtailed. Based on the review presented in Exhibit 1 of the incidence of exemptions from burdensome, government-wide legislation (such as per- sonnel, procurement, and funding), and summarized in Table B-i, it appears that the prospects are far higher for granting such exemptions to a user-funded FAA corporation than they are for either an independent agency or for the agency if it remains within the DOT. It would have control over its financing, including the right to leverage its funds through borrowing. Any conflict between the FAA's promotional and regulatory roles could be readily resolved by leaving the promotional role in the DOT; any conflict between the regulatory and operational roles could be miti- gated by the creation of a post with safety oversight responsibilities. Longer tenure for the Administrator would likely be achieved under most corporate models; long-range planning and stability of funding would be greatly enhanced. Converting the entire FAA to a corporation would probably be less disruptive than the typical and recurring internal reorganizations, and certainly less disruptive than splitting off ATC; no significant threat to the continuity of performance of vital ATC or safety functions would be raised. Disadvantages The somewhat complex governance structure proposed by the ASC illustrates the difficulty of seeking to make such a corporation fully in- dependent while protecting the public and various segments of the industry from being overcharged or otherwise ill-served or disadvantaged. (The Postal Service has a different, but equally complex governance and rate- setting structure.) If the corporation is governed by a board, experience suggests that one should expect to see at least some of the board appointments used for
Appendix B: Organizational Options for the FAA 357 patronage purposes and also expect to see significant gaps in filling vacancies. A substantially autonomous corporation, with a single head and no board of directors, could be a problem in the event of a poor choice of Administrator. (Note, however, that it is not proposed to constrain the President's power to remove the Administrator.) The single-headed structure proposed by the 1986 NAPA panel, providing for policy oversight by the Secretary of Transportation, leaves open the possibility that an unwise DOT Secretary could still intervene unduly in the affairs of the corporation. The separate status of a corporation would (depending on whether it has a link to the DOT Secretary) raise some of the same issues of possible fragmentation of national and international policy that were dis- cussed previously with regard to independent agency status for the FAA. Summing Up Corporate status for the whole of the FAA (except for the promotional role) would seem to offer the greatest prospect for mitigation of the prob- lems confronting the agency, while introducing few, if any, new problems (except, of course, those of a transition). Allowing it to operate on a businesslike basis would be in the public interest, provided that legitimate concerns about policy fragmentation or about vesting too much rate-setting authority in it can be dealt with adequately. These concerns would be substantially alleviated by the supervisory role to be assigned to the DOT Secretary under the NAPA proposal. However, as suggested below, care should be taken to assure that the corporation does not become de facto a part of the DOT. OBSERVATIONS ON POLITICAL FEASIBILITY The purpose of this appendix was to explore both the desirability and feasibility of adopting one or more of the proposed organizational re- sponses to the FAA's problems. The preceding sections focused primarily on desirability. This one emphasizes issues affecting political feasibility. The next section offers suggestions fOr organizational change taking into account both desirability and feasibility. Although micromanagement of the FAA by the DOT has often been a problem, it was a serious issue for some 6 years only, of the 24 years
358 WINDs OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION since the DOT was established. It does not seem to be a vexing problem at this time. As a result, congressional interest in changing the FAA's status appears to have flagged. Legislation effecting major reorganizations is seldom passed without executive branch support or acquiescence, and is rarely (if ever) passed over a presidential veto. Since there is now no concerted effort within Congress to reorganize the FAA, an executive branch initiative may be required to precipitate action. It is clear that the DOT is opposed to converting the FAA to an independent agency. While the privatization of ATC has been under discussion since 1986, it has attracted little support from industry players, other than controllers, or from members of Congress. Both OTA and a NAPA panel concluded that ATC could not safely be split from certain other FAA functions. Without a strong champion in the executive or legislative branch, it is hard to see how such a proposal can get off the ground. The proposal could undoubtedly gamer more support if a convincing analysis were done to show how the various elements of the safety regulatory and operational functions could be separated, but coordinated well enough to avoid jeop- ardy to aviation safety. Because legislation would be required to solve a number of the FAA's problems, and because such legislation is unlikely to be enacted over executive branch objections, one needs to address the question of what sort of proposals might, at least, pass muster with the executive branch, and preferably command its strong support. The following points address that objective. Providing the FAA more stable and adequate funding can be accom- plished only through legislation; a self-financed corporation offers the best vehicle for accomplishing this goal. Additional procurement and personnel flexibility would also require legislation. The recent enactment of government-wide pay-reform legislation greatly reduces the prospect that the FAA, as a regular federal agency, would be given any new personnel authority or flexibility. A user-fee financed corporation would be consistent with current ad- ministration policies and practices, including those recommended in Sec- retary Skinner's February 1990 report, Moving America (DOT 1990); such a corporation might well be provided the same personnel and procurement exemptions that have been given to other government corporations. A corporation subject to the business-type budget provisions and the audit provisions of the Government Corporation Control Act of 1945 (GCCA) would raise fewer executive branch or GAO objections than
Appendix B: Organizational Options for the FAA 359 would one exempt from the GCCA. It would also have better prospects of acceptance by the Congress. Such a corporation would not need to receive appropriations for its operations (except possibly for intentional subsidy of certain users). Seek- ing to place it off budget would cause additional executive branch con- cerns, and the benefits of being off budget (provided it is not subject to appropriations) appear to be marginal. The prospects of executive branch support for converting the FAA to a government corporation without some link to the DOT Secretary are weak. Such a link (if carefully defined) could preserve some of the benefits of departmental connection while substantially reducing the disadvantages. Because the option of converting the FAA to a government corporation seems feasible, as well as desirable, and because the others discussed appear to be less so, further consideration is given to a possibly salable model in the following section. POSSIBLE CONFIGURATIONS FOR FAA CORPORATION Scope It would include all functions of the the FAA, except for promoting aviation, which would be vested in the DOT. Governance Options are (a) a single head appointed by the President for a fixed term, with an advisory board and a supervisory role in the DOT Secretary, or (b) some kind of representative governing board, of which the Secretary would be a member. Comment The former would be easier to sell to the executive branch and probably to Congress. From the standpoint of good management and accountability, it is superior to the governing board. Fears about undesirable and unwise intervention by the DOT Secretary must be weighed along with the pos-
360 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION sibility that an appointee as Administrator might also turn out to be a mistake. There is no way to ensure good appointments in either position. With that in mind, a fixed term for the Administrator should probably be limited to 5 years, but with no prohibition on reappointment. It is clear from the history of fixed- term appointments that they do result in longer service (see Exhibit 1). The second option might be easier to sell to the industry, but endless arguments would probably ensue over how many seats should be assigned to reflect the perspectives of which interest groups. More important, such a board, if created, would be a forum for possibly bitter debates over who should bear what portion of the system costs. Those disappointed with the decisions of the board would no doubt then try to get Congress to intervene. An advisory board would create similar, though substantially lesser problems, but would seem to be essential if there were to be no governing board. On balance, the first option seems preferable. However, to mitigate the possible resurgence of DOT micromanagement, the Secretary's supervi- sory role should be made nondelegable. That would not preclude the Secretary's use of staff to study issues and make recommendations, but it would require the Secretary personally to communicate approvals or disapprovals to the FAA Administrator. Although such communications could, of course, be in writing, the Administrator and the Secretary would both know that an Administrator's appeal of a secretarial decision would have to be heard by the Secretary. Selection of the option providing for the DOT Secretary's supervisory role is assumed in proposing the other features that follow. Financing The corporation should be entirely self-financed and its revenues should be placed in a public enterprise fund in the corporation, not in the treasury. No appropriations should be made to the corporation for military use of the ATC system. The Defense Department should be charged for using the system (with appropriate credit for the services it contributes) and should acquire the needed funds to pay the charges through its own ap- propriations. The corporation should also have borrowing authority, which would allow it to leverage its funds and, thus, to make capital iiivestnients with a long term payoff. A more difficult issue concerns the possible impact of a user-fee fi- nanced system on general aviation (GA). Fears that GA would oppose the
Appendix B: Organizational Options for the FAA 361 ATA's proposed ATC corporation led the ATA to suggest that, existing shares of the tax burden be "grandfathered." That would mean contin- uation of the cross subsidy from commercial airlines and their passengers to GA, which apparently now exists. An alternative would be for Congress to appropriate funds to the FAA corporation to pay for such subsidies without charging the other users to do so. This would be analogous to the revenue forgone appropriation to the Postal Service for mailings for the blind, etc. If this were to be done, one should seek to include the same arrangement as in the postal legis- lation, whereby the subsidized rates automatically revert to the standard rates if Congress should fail to appropriate the funds. Although an un- desirable departure from the entirely self-financing model, the Tennessee Valley Authority (TVA) and USPS have both had relatively good expe- rience with the combination of appropriated and nonappropriated funds. Rate Setting The Administrator should have authority to set rates, subject to the ap- proval of the DOT Secretary. He would consult his advisory council and should be required to have an independent rate counsel reporting directly to him. That officer would advise him on the rates proposed by the operations officials. Rate setting would be in accord with the provisions of the Administrative Procedure Act, and the legislation should provide some broad policies regarding the rate structure. There are a number of models for the setting and collection of fees by federal corporations [e.g., Federal Deposit Insurance Corporation (FDIC), Pension Benefit Guarantee Corporation (PBGC), Farm Credit, TVA and Securities and Exchange Commission (SEC)]. The enabling statute should address this issue. Because of the operational nature of the ATC system, one would not want to have disputes about unpaid bills. However, because all aircraft must be registered, it would be easy to impose sanctions for non-payment of bills. For example, failure to pay the fees when due would result in the suspension of the right to fly. Public corporations can also be given the power to seize the assets of those delinquent in their payments. While the corporation would, of course, need personnel not now employed by the FAA to administer such a system, the costs of collection should not be prohibitive and would be covered by the fee structure. (The experience of Eurocontrol suggests that fee collection is not likely to be a problem. However, if there should be any concern about the ability to enforce fee
362 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION collections, provision could be made to require that fees be paid in ad- vance, based on estimated usage. Charges for the ensuing period could adjust for actual usage.) Budget Status Good public management principles would call for the corporation to be subject to GCCA and on budget. The executive branch and the GAO are likely to call for such provisions if they are to support the creation of a corporation. Annual appropriations can be avoided entirely in a number of ways, as is the case for TVA's revenue-producing activities. In addition to placing the fund in the corporation, rather than in the treasury, there is the per- manent, indefinite appropriations model as used by Social Security. So long as annual appropriations are not required, there would be little risk to program autonomy if the corporation were subject to GCCA and if its funds were reflected in the budget, as is the case for most government corporations, including those that receive no appropriations.' Safety Regulations It should be explicitly recognized that there ought to be a regulatory review of the ATC operations. This should be provided by a safety counsel to the Administrator. The functions would be similar to those proposed for the Director of Aviation Safety by the ASC, but there would not need to be such an intricate appeals process 'as was proposed by the ASC if a single Administrator is in the decision making role. The creation of such a position would substantially reduce any conflict between the regulatory and operational roles of the FAA. A corporation's regulations would not automatically be exempt from the 0MB regulatory review process but its enabling act could so provide, as was proposed in Representative Oberstar's H.R. 1633. Making the DOT Secretary's supervisory role nondelegable would significantly reduce the DOT's involvement in FAA rulemaking, as well as in other matters. Effects of General Management Statutes A provision modeled on that in the Postal Reorganization Act should be proposed that would exempt the corporation from the pay, classification,
Appendix B: Organizational Options for the FAA 363 and procurement laws, and so forth. General policy provisions might be endorsed, such as those favoring pay comparability with the private sector and competition in procurement to the degree practicable. Labor-Management Relations As a government agency, its employees would be prohibited from striking. Collective bargaining should probably be authorized, but with provisions for resolving impasses. Alternatives would be compulsory arbitration (as in the USPS case), compulsory arbitration using the final offer selection model (which is used for public safety functions in a number of local governments), or placing final authority to set rates in the Administrator, with the DOT Secretary's approval. Relations with the Defense Department Except for the financing arrangements, discussed above, the corporation's relations with the Defense Department would be unchanged, including the provision for military control of ATC in war time. SUMMARY There appears to be little disagreement about the nature and serious sig- nificance of FAA's management problems. The most important of these relate to the following: Lack of control over its financing; The rapid turnover of Administrators; Excessive constraints under legislation relating to personnel and pro- curement systems; and Micromanagement by the DOT, the 0MB, and Congress. Only a few of the needed improvements in FAA management can be accomplished without legislation which, in most cases, would seem un- likely to be enacted except in the context of institutional reform. For example:
364 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION Longer tenure for the Administrator requires legislation and would be relatively easy to accomplish. Micromanagement by the DOT, the 0MB, and Congress can not be significantly reduced unless FAA is given a far more autonomous structure and method of financing. The needed reforms in personnel, funding, and procurement processes are likely to be achieved only through a change in FAA's institutional status in the federal system. Desirable changes in FAA's culture can be accelerated through an entirely new structure and funding method that would permit it to create different reward and incentive systems. The conclusion is that the FAA must have a different organizational form and status. Other than leaving it in the DOT, the leading proposals are: Granting the FAA independent agency status; Establishing a public corporation to operate the ATC system; Privatizing the ATC system; Converting the FAA to a public corporation responsible solely to the President and Congress; and Converting the FAA to a public corporation with some specific policy review authorities in the Secretary of Transportation. Making the FAA an independent agency once more, without the other legislative reforms that are needed (and which are not likely to be accom- plished in such legislation) would probably prove to be worse than taking no action at all. Theoretically, splitting off ATC from the remainder of the FAA should be feasible, but the strong warnings against such a move by OTA and a NAPA panel dictate caution. A careful study could provide the basis for such a reorganization design, but such a study has not been conducted. The same caution argues against the otherwise attractive option of privatizing ATC. In all these circumstances, making the FAA a public corporation appears to be both the most desirable and feasible option. Severing it entirely from the DOT would have some disadvantages which could, however, be over- come by creating a policy link to the Secretary of Transportation, but not to the DOT. This could be accomplished by making the Secretary's au- thorities nondelegable. None of this can be done without legislation. That, of course, requires the cooperation of the executive branch and Congress. Therefore, it is
Appendix B: Organizational Options for the FAA 365 hoped that the Secretary of Transportation will see the advantages of the recommended structure and, with the approval of the White House, will work with Congress to develop legislation that will equip the FAA to cope with the ever-growing problems of managing one of the most complex and vital programs in the nation.
366 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION Exhibit 1: How Other Federal Agencies or Corporations are Equipped to Cope With Problems Like Those Confronting FAA The organizations discussed in this exhibit are grouped under three cat- egories: (a) independent agencies, including independent regulatory boards or commissions, (b) agencies within cabinet departments, and (c) gov- ernment corporations. The cases cited are, essentially, all those that have any relevance to the problems faced by the FAA. The following criteria were used in selecting the provisions or circumstances discussed below: If a problem for the FAA is not relevant to the agencies discussed here, it is simply not mentioned. If an agency has the same problems as the FAA does and has no special authority or flexibility to deal with them, the problems are generally not discussed here. Agencies that have exemptions from the laws or regulations consid- ered to be problems for the FAA, or have separate systems, are listed under the relevant subjects. Provisions of special interest available to selected agencies are also described. This approach is designed to highlight special provisions which tend to be associated, or not associated, with one type of agency or corporation. This should assist in evaluating the possibility that a given type of or- ganizational status might be accompanied by increased flexibility. Independent Agencies Structural and Statutory Provisions Independent agencies, except regulatory boards and commissions, are typically headed by single, PAS officials. While they nominally report directly to the President, their access is necessarily limited. Except in special (or ceremonial) circumstances, they are unlikely to meet with the
Appendix B: Organizational Options for the FAA 367 President. Their supervision comes, in effect, from White House staff, the 0MB, and cabinet councils. The absence of a department head in a supervisory role has both pluses and minuses, as discussed below. In brief, as far as the agency's self- interest is concerned, such agencies do not compete for resources with other agencies reporting to one official below the President, but they do not have a department head to "go to bat" for them with the White House and the 0MB. Tenure Virtually all agency heads are appointed for indefinite terms, and their average length of service is like that of all PAS appointeesâslightly more than two years. Sometimes, in highly technical or professional agencies, the tenure tends to be longer. There are a few agency heads who have fixed terms: Director of the National Science Foundation (6-year term since the agency was established in 1950; the average service has been almost 5 years, with one, director serving two terms). Director of the Office of Personnel Management (OPM) (4-year term since the former Civil Service Commission (CSC) was converted to OPM in 1978; the average serviceâincluding the first director's service also as CSC chairâhas been almost 4 years). Chairman of the Federal Reserve Board (4-year term, as Chairman, since 1934; the average service has been almost 8 years, with six chairmen serving two or more terms). Comptroller General (15-year, nonrenewable term, since the estab- lishment of the GAO in 1921; the average service has been more than 11 years). The GAO is regarded as a congressional staff agency although it performs a number of executive functions. The Postmaster General was appointed for a 4-year term before the department was converted to the Postal Service. Since that post was often filled by a former head of the political party of the President, it was clearly not a position where the incumbent was expected to survive a turnover in the White House. This illustrates that a fixed term, absent any statutory restrictions, does not limit the President's power to remove the officeholder.
368 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION Reporting Relationships Most agencies are subject to oversight from one or more Executive Office of the President agencies, such as the 0MB and the Office of Science and Technology Policy, and one or more cabinet councils. Special statutory or administrative arrangements exist regarding: The Director of Central Intelligence, who operates under the direction of the President and of the statutory National Security Council, which is chaired by the President. The Administrator of NASA, whose programs are overseen by the National Space Council (chaired by the Vice President and established by presidential instead of statutory action). Regulatory Review The promulgation of regulations by agencies, boards, and commissions is subject to the Administrative Procedure Act. However, the independent regulatory boards and commissions are not required to participate in the OMB's regulatory review. Financing With few exceptions, independent agencies are fully subject to the same budget, appropriations and apportionment processes that govern the FAA. That means that independent agencies, such as NASA, are just as vul- nerable to the vagaries of the legislative, budgetary, and political processes as is the FAA. In recent years, Congress has directed that a number of agencies submit their annual appropriations requests to Congress in the same form as furnished to the 0MB. This is designed to discourage reductions by the 0MB by making them more visible and to ensure that Congress will not have to probe for the amount of the agency's initial budget request. NASA's programs have experienced cyclical increases and decreases, and it has been subjected to some of the short-term stops and starts from which the FAA has suffered. NASA has no extraordinary freedoms or flexibility under the budget and procurement statutes. The Washington Post on September 14, 1990, reporting on NASA Administrator Richard
Appendix B: Organizational Options for the FAA 369 H. Truly's comments to the Advisory Committee on the Future of the U.S. Space Program, stated: T]he real problems, from his perspective, include his need for more authority and flexibility to run the program and "a better match" between NASA's programs and its resources. He cited "government-wide management prac- tices" that limit managers' ability to move money and people and to procure services needed to carry out programs. Human Resources Almost all independent agencies operate under government-wide appoint- ments, removal, pay, and classification statutes. Following are agencies that are exempt from some of the general laws, but they are usually subject to the same salary ceilings as are applicable elsewhere: Central Intelligence Agency (CIA), GAO (since 1980), Nuclear Regulatory Commission (for some positions transferred from the former Atomic Energy Commission, which had an independent per- sonnel system), and The Panama Canal Commission (scheduled to be terminated in 1999). NASA has authority to pay a limited number of personnel at rates above the civil service ceiling. Procurement Except for the Federal Reserve Board and the Smithsonian Institution, no examples have been found of independent agencies that are exempt from the competitive procurement statutes. However, certain statutory provi- sions authorize exceptions (e.g., for national security ,.reasons). Independent Agency Performance Independent agencies have generally functioned satisfactorily for long periods without raising serious questions about their status. However, a large agency's constituents will often seek to have it elevated to cabinet status. Recent examples include Energy and Veterans Affairs, while the
370 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION proposed elevation of the Environmental Protection Agency has not yet occurred. The FAA (from 1958 to 1966) Among the leading reorganization bills introduced in the last Congress were those by Senator Ford (S. 1600) and by Representative Oberstar (H.R. 1633) to make the FAA an independent agency. There seems to be a common belief that the FAA prospered during the years of its in- dependence, and the motivation for such legislative proposals is, in part, to restore that situation. Former Administrator Halaby's concerns about the FAA's lack of support when it was independent have not received much recent attentionâperhaps because he has now proposed to remove the FAA from the DOT. His earlier concerns included the assignment of lead responsibility for supersonic transport development to Defense instead of the FAA. He also believed that the agency could better withstand industry pressures as a part of a cabinet department. Also overlooked is that the FAA seemed to do reasonably well during the first 17 years of its existence as an agency within the DOT. That is why the sections on Events and Developments and on Structural and Statutory Impediments in this appendix focused on changes that have occurred more recently in the FAA's environment. The CIA The CIA has had its share of failures or apparent failures. While these have sometimes attracted substantial congressional and public attention, they have not generally led to proposals to change its status. Rather, proposals have focused on changing the nature and extent of congressional oversight. NASA NASA was once widely praised as an example of outstanding government performance at the frontiers of technology. Its recent record of difficulties and failures raises some interesting questions: Did NASA simply burn out as an agency and lose the esprit and quest for excellence that was evident during the launches to the moon?
Appendix B: Organizational Options for the FAA 371 Did the lack of stability in Presidential or congressional commitment to the space program and its funding erode the agency's ability to maintain monientum and direction? Did NASA prosper at a time when the competitive procurement system was not so constraining as it is now? Has NASA's capacity to give adequate supervision and direction to the contractors on whom it relies so heavily (which was probably never strong enough) declined markedly? If so, is it because NASA lacks sufficient numbers of personnel to do the job? Or, is it because it has lost, and not been able to replace, the technical expertise required to manage a largely contracted out program as a result of salary scales that have become noncompetitive? Does NASA delegate too much authority to its research centers and to its engineering personnel, while lacking a meaningful strategic or man- agement planning capacity? The issues raised by these questions are relevant to the evaluation of options for FAA reorganization, especially the option for independent agency status. Agencies Within Cabinet Departments Structural and Statutory Provisions Generally, agencies within departments exercise their functions under delegations from the Secretary, in whom the statutory powers are vested. There are a number of exceptions, however. In the DOT enabling act, for example, there was a provision intended to vest the aviation safety func- tions directly in the Administrator, instead of in the Secretary. Under the first Secretary, however, this provision was so interpreted by the depart- ment as to vitiate the legislative intent. Manyof the agencies within departments are headed by PAS appointees, who almost always serve indefinite terms (with service averaging little more than two years). A few have fixed terms: After J. Edgar Hoover's 47-year tenure as Federal Bureau of Inves- tigation (FBI) Director ended in 1972, the post was converted to a PAS position. In 1976, the law was changed again to prescribe a 10-year term. The first appointee under the new provision served in the post for more
372 WiNos OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION than 9 years before accepting appointment as the Director of Central Intelligence. The Administrator of the St. Lawrence Seaway Development Cor- poration has a 7-year term (see the following section on corporations for more details). Since the 1989 establishment of the Department of Veterans Affairs, the Chief Medical Director and Chief Benefits Director are PAS appointees for 4-year terms. The Commandant of the Coast Guard serves a 4-year term. Financing Many special provisions impose restrictions or conditions on the expen- diture of funds by various agencies within departments, but there are few significant exemptions from the general budget and appropriations pro- cesses. Principal exceptions involve industrial funds, public enterprise funds, and revolving funds. Industrial funds are chiefly used for transactions that are internal to the government. For example, sums received for purchases from a supply agency can be spent to replenish supplies without going through the ap- propriations process again. Public enterprise and revolving funds, on the other hand, involve business-type functions and transactions with the public. If the authorizing legislation so provides, such funds can be established in an agency, rather than in the treasury, and the receipts can then be spent without further appropriations action. An example is the Bonneville Power Administration in the Department of Energy. Human Resources As with independent agencies, the civilian agencies within cabinet de- partments operate under the government-wide appointments, removal, pay, and classification statutes. Some separate systems are prescribed by statute, such as the Foreign Service Personnel System. Following are agencies that are exempt from some of the general laws, but they are usually subject to the same salary ceilings as are applicable elsewhere: Military Intelligence Agencies: The Defense Intelligence Agency, the National Security Agency, and the civilian intelligence components of the
Appendix B: Organizational Options for the FAA 373 three military departments are all exempt from the classification and pay statutes. Department of Justice: Employees of the FBI are not covered by the appointments and appeals provisions of the Civil Service laws, but they are subject to the pay and classification provisions. U.S. Attorneys are exempt from the Civil Service laws. Department of Veterans Affairs: The Veterans Health Service and Research Administration continues to have a number of personnel features carried over from the former Veterans Administration that are different from those in the Civil Service system. Department of Health and Human Services: The commissioned officer corps of the Public Health Service is subject to a personnel system similar to those used by the other uniformed services. Department of Energy: The five power agencies of the Interior De- partment (such as Bonneville and Southeastern) are exempt from the Civil Service laws. Other technical agencies within cabinet departments: A number of highly technical agencies within cabinet departments experience personnel problems similar to those of the FAA. These include Defense Advanced Research Projects Agency and other components of the Defense Depart- ment, the National Oceanic and Atmospheric Agency, the National In- stitute of Standards and Technology (NIST), the National Institutes of Health, the Food and Drug Administration, and components of the En- vironmental Protection Agency and the Department of Energy. In the case of several agencies within departments, there are provisions that authorize salaries above the civil service ceiling, or noncompetitive appointments, for limited numbers of personnel. Examples are the National Institutes of Health (115 positions) and NIST. NIST is also authorized by statute to carry on a demonstration project entailing greater flexibility in classification and pay, along the lines of the Navy's China Lake project. (The Office of Personnel Management has interpreted its authority to approve such demonstrations under the Civil Service Reform Act to pre- clude a second demonstration of the same techniques.) Otherwise, these agencies generally have no special flexibility in pay and personnel systems. Procurement No examples have been found in which agencies within departments are exempt from the competitive procurement statutes. As noted earlier, these
374 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION statutes do provide a number of grounds for exceptions, including for national security reasons. Corporations or Authorities In contrast to the two categories of agency discussed above, where only a small number in each class have any special flexibility, all of the members of this group of corporations or authorities have one or more dimensions of flexibility in personnel, budget or procurement authority. Some of these corporations are subject to the Government Corporation Control Act of 1945. Others are not. The following sections identify those characteristics of government cor- porations that are related to the list of FAA problems discussed previously. Most of the information is derived from a report (covering 39 corporations) prepared by a NAPA panel for the 0MB (NAPA 1981); a Congressional Research Service report (covering 31 corporations) (Moe 1983); and a study (covering 45 corporations) prepared by the GAO for the House Committee on Government Operations (GAO 1988). Data relating to corporations that have been terminated are sometimes included because their characteristics may still be instructive. The three reports mentioned above include as many as 26 government-created, pri- vate corporations, or government-sponsored enterprises. In the latter group, some are clearly government agencies, although they are sometimes mis- classified as private. The Communications Satellite Corporation is an example of the government-chartered, private category. The Legal Ser- vices Corporation (which is on budget and supported by appropriations) is an example of a non-revenue-producing corporation that is incorrectly called private. These two groups are generally not treated below. Structural and Statutory Provisions Most government corporations are governed by boards of directors to which the CEO reports. Typically, the boards are composed in whole or in part of government officials, serving ex officio, and of other members appointed by the President. The CEO is usually appointed or designated by the President, but in some cases is appointed by the board, its chairman, or another supervisory officer.
Appendix B: Organizational Options for the FAA 375 Other Supervisory Arrangements In a few cases the corporation is single-headed and there is no board: St. Lawrence Seaway Development Corporation (Seaway Corpora- tion) (headed by an Administrator), Government National Mortgage Association (GNMA) (headed by a President who is a PAS appointee), and Federal Housing Administration (FHA) [headed by a Federal Com- missioner who is a PAS Assistant Secretary of Housing and Urban De- velopment (HUD)]. The Administrator of the Seaway Corporation is a PAS, Executive Level IV appointee who, since 1975, has been appointed to a 7-year term; two Administrators have served full terms since then. The Seaway Corporation has an advisory board, but the other two single- headed corporations do not. The 1981 NAPA panel noted that the practice of having boards of directors was borrowed from the private sector, but that it had more disadvantages than advantages (NAPA 1981, 31, 32). The NAPA panel preferred the combination of an advisory board and vesting a supervisory role in a department head. This matter is discussed further in the evaluation of governance arrangements for the corporate organizational options earlier in the appendix. At this time, all three corporations that do not have a board of directors are within departments, the first in the DOT and the others in HUD. However, as discussed below, the Seaway Corporation originally was independent, had a single head and no board of directors. The following are also in departments: Commodity Credit Corporation, Federal Crop Insurance Corporation (FCIC), Federal Prison Industries, Inc. (in the Justice Department de facto), and PBGC. The following are or were subject to the supervision or direction of other officials or agencies: Federal Financing Bank,
376 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION Federal Savings and Loan Insurance Corporation (FSLIC) (termi- nated), and Rural Telephone Bank (mixed ownership). Seaway Corporation The corporation is now in the DOT. When it was established as an in- dependent agency in 1954, it was headed by a single administrator, ap- pointed for an indefinite term, and subject to the supervision of the President or of the head of such agency as the President might designate. Before the DOT was established in 1966, the Commerce Department was the principal transportation agency and the President had designated the Sec- retary of Commerce as the supervisor of the Seaway Administrator; earlier, the President had designated the Secretary of the Army. The bill submitted by the executive branch to create the DOT proposed to continue that statutory arrangement, with the understanding that, after enactment, the President would then transfer the supervisory role to the Secretary of Transportation. Congress, however, made the Seaway Cor- poration one of the components of the DOT, with no organizational dis- tinction from the other modal administrations in the DOT. As a result, it has lost most of the operational flexibility it once had and is subject to fairly detailed oversight by DOT staff. Corporation Sole There are some programs that are vested in a single officer in such a way that the officer has all the powers of a corporation. This is called a corporation sole. For example, the following language appears in the 1991 budget: The Secretary of Transportation is hereby authorized to make such expen- ditures and investments, within the limits of funds available pursuant to section 1306 of the Act of August 23, 1958, as amended (49 U.S.C. 1536), and in accordance with section 104 of the Government Corporation Control Act, as amended (31 U.S.C. 9104), as may be necessary in carrying out the program set forth in the budget for the current fiscal year for aviation insurance activities under said Act. Similar corporate powers have been vested in the HUD Secretary.
Appendix B: Organizational Options for the FAA 377 Independent Corporations The following corporations have the same status as independent agencies, except for their corporate form: Export-Import Bank of the United States (Ex-Im), Federal Deposit Insurance Corporation (FDIC), Inter-American Foundation, Overseas Private Investment Corporation (OPIC), Pennsylvania Avenue Development Corporation (PADC), TVA, and USPS (not incorporated or called a corporation in its statute). Except TVA and USPS (whose governing boards are described below), the corporations with independent status may receive policy guidance from other officials who serve as regularly appointed, or ex officio, members of the boards of directorsâto the extent that such officials choose to offer it in their capacity as board chairs or members. TVA is headed by a PAS, three-member full-time board, each of whom serves a 9-year term. The chair is designated by the President for an indefinite term. There is no special provision for supervision, except that provided through the budget process for corporations (see below), or other central management functions such as legislative clearance. The USPS is headed by a Board of Governors, which consists of 9 part-time members appointed by the President from the public at large for 9-year, overlapping terms, and the Postmaster General (PMG) and Deputy Postmaster General. The 9 presidential appointees choose their own chair and appoint the PMG to an indefinite term, and those 10 appoint the deputy. The board has set the PMG's salary at the cabinet secretary (Level I) rate, which is the maximum authorized under the Postal Reorganization Act. Regulations With respect to regulations and the regulatory review functions assigned to the 0MB by executive order, corporations are subject to the Admin- istrative Procedure Act. However, corporations that perform regulatory functions and are defined as independent regulatory commissions, such as FDIC, are not subject to 0MB reviews since the executive orders do not apply to independent regulatory boards and commissions. The "gov-
378 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION emment in the sunshine" laws apply to meetings of corporate boards but, of course, not to single administrators. Financing Budget Status The following corporations were on budget in 1988, even though some are self-financing and received no appropriations: Commodity Credit Corporation, Ex-Im Bank, FCIC, FDIC, FHA, Federal Prison Industries, Inc., FSLIC, GNMA, Inter-American Foundation, OPIC, PADC, PBGC, Rural Telephone Bank, Seaway Corporation, TVA, and USPS (now off budget once more). The Government Corporation Control Act of 1945 (GCCA) The GCCA listed but did not define 42 "wholly owned" or "mixed- ownership" corporations that existed at that time. Consistent with Amer- ican legislative practice, government corporations established since then are subject to the GCCA, or not subject to it, based on the provisions of their own enabling statutes. Similarly, they have only those powers or exemptions which are explicitly stated in the legislation. Typically, an enabling act authorizes a corporation to "determine the character and
Appendix B: Organizational Options for the FAA 379 necessity for its expenditures, and the manner in which they shall be incurred, allowed and paid." What that is intended to mean, and has in fact meant, is that the corporation is not subject to detailed, periodic, line-item financial reviews, either by the 0MB or Congress. The stated purpose of GCCA was "to bring Government corporations and their transactions and operations under annual scrutiny by the Congress and provide current financial control thereof." Toward that end, the act provided for audits by the GAO and, with respect to wholly owned corporations, it provided for the annual preparation and submission of a "business-type budget." Although the business-type budget was subject to changes by the Pres- ident, and therefore by the then Budget Bureau, the intent was that there be a program review, rather than a line item review, and no appropriations. Congress was expected to approve the program in toto. Following is an example of the language used by Congress in appropriations acts to ap- prove a corporation's program, without specifically appropriating funds to it (0MB 1990): The Saint Lawrence Seaway Development Corporation is hereby authorized to make such expenditures . . and to make such contracts and commitments as may be necessary in carrying out the programs set forth in the Corporation's budget for the current fiscal year. The GCCA also allows for appropriations to be made to wholly owned corporations. For example, appropriations are made annually to TVA for its nonrevenue-producing programs. According to the 1981 NAPA panel, Congress' expressed intent that such enterprises operate with "reasonable autonomy and flexibility" was achieved. The panel report (NAPA 1981) also stated, "In many respects the GCCA could be called the government corporation de-control act." Of course, Congress can always enact language to place limits or prov- isos on its approval of a corporation's program pursuant to the GCCA, but it must do so by a deliberate act, that is, it can not reduce or increase an item in the corporation's program. However, the appropriations act will often contain such a limit on how much of its nonappropriated funds a corporation can spend for administrative expenses. The NAPA panel surveyed 39 corporations and asked them whether they had found GCCA to be a problem. No corporation recommended any changes in it, but the General Counsel of TVA complained about the "tendency of some to overlook the Act and to treat government
380 WINDs OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION corporations in the same manner as Federal departments for budgetary purposes." All of the wholly owned corporations, except USPS and FDIC (which is incorrectly classified as mixed ownership), are subject to the GCCA. One indicator of the relative freedom accorded to corporations covered by the business-type budget provisions of the GCCA is whether the cor- poration is subject to the provisions of the Anti-Deficiency Act. That act is the basis for the OMB's apportionment process, through which agencies are authorized each quarter to spend sums that the 0MB concludes will not cause the agency to exceed amounts available in an appropriation or fund. In 1988, all of the corporations listed above were subject to the act, except the Commodity Credit Corporation, FDIC, the Federal Financing Bank, and USPS. For those subject to the act, however, apportionment generally applies only to amounts subject to the administrative expense limitations men- tioned above. In some cases where there are appropriations, the 0MB apportions the funds in an annual amount, rather than quarterly, thus emphasizing that corporations are generally subject to annual program reviews only. Self-Financing Corporations The Seaway Corporation was intended to be self-financing through toll revenues, although that has not proved feasible. It was authorized to set toll rates for the U.S. portion of the Seaway, subject to the approval of the President. TVA sets its own.rates for the electrical power it produces and sells. Receipts from these revenue-producing activities are available to TVA for expenditure under its authorization statute without apportionment by the 0MB or congressional appropriations, although they are submitted to the 0MB and included in the President's budget. The USPS was converted by the Postal Reorganization Act of 1970 (PRA) from the former Post Office Department to what is, effectively (though not so-termed in the statute or actually incorporated) a government corporation. Its Board of Governors has final authority to set postal rates, subject to certain limitations. An independent, five-member, Postal Rate Commission (PRC) must review the proposed rates and make its own recommendations to the Board of Governors. The Board can put the PRC's recommended rates into effect. The Board can submit new proposals if it does not approve those proposed by the PRC. However, it can put its own
Appendix B: Organizational Options for the FAA 381 originally proposed rates into effect, notwithstanding the PRC's disap- proval, but only by unanimous vote. The USPS is exempt from a number of government-wide management statutes, including the budget and procurement statutes. Title 39 of United States Code 410 provides the following: Except as provided by subsection (b) of this section, and except as otherwise provided in this title or insofar as such laws remain in force as rules or regulation of the Postal Service, no Federal law dealing with public or Federal contracts, property, works, officers, employees, budgets, or funds, including the provisions of chapters 5 and 7 of title 5, shall apply to the exercise of the powers of the Postal Service. Pursuant to this authority, the Postal Service will spend about $45 billion in fiscal 1991 without any Congressional action. An additional one-half billion dollars is still appropriated to the Postal Service for revenue forgone as a result of certain statutorily mandated, below-cost services, such as mailing for the blind. However, if Congress fails to appropriate the funds for these subsidized classes, their postage rates automatically revert to those paid by others for the same services. When the PRA was enacted, the nonappropriated funds were off budget. Then 0MB Director David Stockman placed them on budget for fiscal year 1986. After the 1987-88 expenditure reductions were ordered (as described previously), Congress began to consider action to take USPS off budget again, which it finally took for fiscal year 1990. However, it seems that Congress was inclined to do so simply because 1990 was a year with a Postal Service deficit, pending the 1991 rate increase. The FDIC has been self-financed through the premiums charged banks for federal insurance of their deposits. Congress has set limits on these premiums by statute. Because of Congressional fears that the banks might follow the unhappy example of the savings and loan industry, legislation is expected to be enacted that would remove these ceilings and allow the FDIC to establish the prejnium rates. In the meantime, the rates have been increased by statute. Human Resources All wholly owned government corporations are subject to the regular civil service and other management legislation unless specifically exempted by their statutes. Thus all the corporations discussed above are subject to the civil service classification and pay systems, except as follows:
382 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION FCIC is exempt from the pay statutes, but applies them anyway. FDIC is exempt from the classification system, but applies it anyway. TVA is exempt from both. USPS is exempt from both, but has its own personnel system, which is established in considerable detail by statute. For example, collective bargaining is authorized, and compulsory arbitration is provided for in the event of an impasse. Generally, corporations that are exempt from the pay statutes are still subject to the same, or similar, salary ceilings as those applying elsewhere. Exceptions include the former U.S. Railway Association, the former SynFuels Corporation, and a number of others, such as the Resolution Trust Corporation and the Office of Thrift Supervision, recently created to deal with the numerous savings and loan failures. Procurement About half of the corporations discussed above are subject to the com- petitive procurement statutes. Corporations that apply the statutes at their discretion (GNMA and USPS) can readily waive them. The following are or were exempt: Commodity Credit Corporation, FCIC, FDIC, GNMA, OPIC, TVA, USPS, and U.S. Railway Association (terminated). REFERENCES ABBREVIATIONS ASC Aviation Safety Commission DOT U.S. Department of Transportation GAO General Accounting Office NAPA National Academy of Public Administration 0MB Office of Management and Budget OTA Office of Technology Assessment
Appendix B: Organizational Options for the FAA 383 Journal of the Air Traffic Control Association. 1987. June. ASC. 1988. Final Report and Recommendations. July. Belanger, R., and C. Newpol. 1985. A Study of the Safety/Regulatory Functions of the Air Traffic Control System. DOT. 1990. Moving America. Feb. GAO. 1988. Profiles of Existing Government Corporations. Washington, D.C., Dec. The Heritage Foundation. 1982. Air Traffic Control: The Private Sector Option. Washington, D.C., Oct. Moe, R. 1983. Administering Pub/ic Functions at the Margin of Government: The Case of Federal Corporations. Congressional Research Service, Dec. NAPA Panel. 1981. Report on Government Corporations. Washington, D.C., Sept. NAPA Panel. 1983. Revitalizing Federal Management: managers and their overbur- dened systems. Washington, D.C., Nov. NAPA Panel. 1986. The Air Traffic Control System: Management By A Government Corporation. Washington, D.C., March. 0MB. 1990. Budget of the United States Government (1991 Budget). Washington, D.C., Jan. OTA. 1988. Safe Skies for Tomorrow. Congress of the United States, July. Reason Foundation. 1986. Privatizing the Air Traffic Control System. Santa Monica, Calif., Nov. U.S. Congress. House. 1988. Committee on Public Works and Transportation. Hear- ings Before the Subcommittee on Aviation.
Appendix C Airport Identification Codes U.S. commercial service airports are identified by a three-letter code assigned by the International Air Transport Association. Codes for the airports mentioned in this report are listed below. ABQ Albuquerque International Airport, Albuquerque, New Mexico ALB Albany County Airport, Albany, New York AMA Amarillo International Airport, Amarillo, Texas ANC Anchorage International Airport, Anchorage, Alaska ATL William B. Hartsfield Atlanta International Airport, Atlanta, Georgia AUS Austin Robert Mueller Municipal Airport, Austin, Texas BDL Hartford/Springfield Bradley International Airport, Windsor Locks, Connecticut BHM Birmingham Airport, Birmingham, Alabama BIL Billings Logan International Airport, Billings, Montana BNA Nashville Metropolitan Airport, Nashville, Tennessee BOl Boise Air Terminal (Gowen Field)., Boise, Idaho BOS Boston General Edward Lawrence Logan International Airport, East Boston, Massachusetts BUF Buffalo International Airport, Buffalo, New York BUR Burbank-Glendale-Pasadena Airport, Burbank, California BWI Baltimore-Washington International Airport, BWI Airport, Maryland CAE Columbia Metropolitan Airpoil, West Columbia, South Carolina CHS Charleston International Airport, Charleston, South Carolina CLE Cleveland Hopkins International Airport, Cleveland, Ohio 384
Appendix C: Airport Identification Codes 385 CLT Charlotte/Douglas International Airport, Charlotte, North Carolina CMH Columbus International Airport, Columbus, Ohio COS Colorado Springs Municipal Airport, Colorado Springs, Colorado' CRP Corpus Christi International Airport, Corpus Christi, Texas CVG Cincinnati International Airport, Covington, Kentucky DAL Dallas Love Field, Dallas, Texas DAY Dayton (James M. Cox) International Airport, Dayton, Ohio DCA Washington National Airport, Washington, District of Columbia DEN Denver Stapleton International Airport, Denver, Colorado DFW Dallas/Fort Worth International Airport, Dallas/Ft. Worth, Texas DSM Des Moines International Airport, Des Moines, Iowa DTW Detroit Metropolitan (Wayne County) Airport, Detroit, Michigan ELP El Paso International Airport, El Paso, Texas EWR Newark International Airport, Newark, New Jersey FLL Ft. Lauderdale-Hollywood International Airport, Ft. Lauderdale, Florida FMY Southwest Florida Regional Airport, Fort Myers, Florida GEG Spokane International Airport, Spokane, Washington GFK Grand Forks Mark Andrews International Airport, Grand Forks, North Dakota. GRR Grand Rapids (Kent County) International Airport, Grand Rapids, Michigan GSO Piedmont Triad International Airport, Greensboro, North Carolina GSP Greenville-Spartanburg Airport, Greer, South Carolina HNL Honolulu International Airport, Honolulu, Hawaii HOU Houston William P. Hobby Airport, Houston, Texas HRL Harlingen Valley International Airport, Harlingen, Texas lAD Washington Dulles International Airport, Washington, District of Columbia lAG Niagara Falls International Airport, Niagara Falls, New York IAH Houston Intercontinental Airport, Houston, Texas ICT Wichita Mid-Continent Airport, Wichita, Kansas IND Indianapolis International Airport, Indianapolis, Indiana ISP Ronkonkoma (Long Island) MacArthur Airport, Islip, New York ITO Hilo General Lyman Field, Hilo, Hawaii JAX Jacksonville International Airport, Jacksonville, Florida JFK New York John F. Kennedy International Airport, Jamaica, New York KOA Kailua-Kona Keahole Airport, Kailua-Kona, Hawaii LAS Las Vegas McCarran International Airport, Las Vegas, Nevada
386 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION LAX Los Angeles International Airport, Los Angeles, California LBB Lubbock International Airport, Lubbock, Texas LGA New York LaGuardia Airport, Flushing, New York LGB Long Beach Airport (Daugherty Field), Long Beach, California LIH Lihue Airport, Lihue, Hawaii LIT Little Rock Regional Airport, Little Rock, Arkansas MAF Midland International Airport, Midland, Texas MCI Kansas City International Airport, Kansas City, Missouri MCO Orlando International Airport, Orlando, Florida MDW Chicago Midway Airport, Chicago, Illinois MEM Memphis International Airport, Memphis, Tennessee MIA Miami International Airport, Miami, Florida MKE Milwaukee General Mitchell International Airport, Milwaukee, Wisconsin MSN Madison (Dane County) Regional Airport (Truax Field), Madison, Wisconsin MSP Minneapolis-St. Paul International Airport (Wold-Chamberlain Field), St. Paul, Minnesota MSY New Orleans International Airport, New Orleans, Louisiana OAK Oakland International Airport, Oakland, California OGG Kahului Airport, Kahului, Maui, Hawaii OKC Oklahoma City Will Rogers World Airport, Oklahoma City, Oklahoma OMA Omaha Eppley Airfield, Omaha, Nebraska ONT Ontario International Airport, Ontario, California ORD Chicago-O'Hare International Airport, Chicago, Illinois ORF Norfolk International Airport, Norfolk, Virginia PBI Palm Beach International Airport, West Palm Beach, Florida PDX Portland International Airport, Portland, Oregon PHL Philadelphia International Airport, Philadelphia, Pennsylvania PHX Phoenix Sky Harbor International Airport, Phoenix, Arizona PIT Pittsburgh International Airport, Pittsburgh, Pennsylvania PVD Providence Theodore Francis Green State Airport, Warwick, Rhode Island PWM Portland International Jetport, Portland, Maine RDU Raleigh-Durham Airport, Morrisville, North Carolina RIC Richmond International Airport (Byrd Field), Richmond, Virginia RNO Reno Cannon International Airport, Reno, Nevada ROC Rochester International Airport, Rochester, New York SAN San Diego International Airport (Lindbergh Field), San Diego, California
Appendix C: Airport identification Codes 387 SAT San Antonio International Airport, San Antonio, Texas SAV Savannah International Airport, Savannah, Georgia SDF Louisville Standiford Field, Louisville, Kentucky SEA Seattle-Tacoma International Airport, Seattle, Washington SF0 San Francisco International Airport, San Francisco, California SJC San Jose International Airport, San Jose, California SJU San Juan Luis Munoz Mann International Airport, San Juan, Puerto Rico SLC Salt Lake City International Airport, Salt Lake City, Utah SMF Sacramento Metropolitan Airport, Sacramento, California SNA Santa Ana (Orange County) John Wayne Airport, Santa Ana, California SRQ Sarasota-Bradenton Airport, Sarasota, Florida STL St. Louis-Lambert International Airport, St. Louis, Missouri SYR Syracuse Hancock International Airport, Syracuse, New York TPA Tampa International Airport, Tampa, Florida TUL Tulsa International Airport, Tulsa, Oklahoma TUS Tucson International Airport, Tucson, Arizona TYS Knoxville McGhee Tyson Airport, Alcoa, Tennessee
Appendix D Dissenting Statement Melvin A. Brenner Ibelieve that some parts of the committee's report help to clarify im-portant issues regarding the present status of the airline industry. How- ever, on other issues, I believe the report leaves misleading impressions, or reaches unwarranted conclusions, by failing to apply tests of reason- ableness drawn from the airlines' empirical experience. By wayof illus- tration, the comments that follow cite three specific examples which have the most current policy significance: (a) the issue of barriers to entry; (b) the proposal to impose congestion fees at airports; and (c) the discussion of fare levels at hub airports. BARRIERS TO ENTRY With the industry's increased concentration in recent years, a policy ques- tion has been raised in Congress and elsewhere whether such concentration has developed because of "barriers to entry" erected by the larger carriers to block entry (or expansion) of other carriers. The so-called barriers have been listed as carrier ownership of computer reservation systems (CRSs), long-term airport gate leases, frequent flyer programs, commission over- rides to travel agents, and code sharing. The committee endorses the idea that these factors have indeed limited entry. It admits that their effect may be individually small, but then dilutes that admission by referring to some unidentified interaction among them, which it says gives them "more powerful influence." With specific ref- erence to the issue of CRSs, it claims that such systems "offer considerable competitive advantages to the carriers that own them," and it recommends regulatory intervention to change aspects of CRS operation. WKV
Appendix D: Dissenting Statement 389 I submit that the industry's evolution to its present degree of increased concentration has been the result of a series of carrier mergers and failures, which are traceable to a variety of marketplace or managerial causes totally unrelated to the so-called barriers to entry. The CRS issue in particular has become a "strawman," with no evidence of any connection between ownership of such systems and airline success or failure. The three best profit margins within the industry have been achieved by carriers that did not own CRSs: Southwest has had the most consistent and most favorable profit margin, without ever participating in a CRS. For the 12 years of dereg- ulation, its net profit margin has averaged 7.1 percent, which is more than double that of the largest CRS owner, and more than five times higher than that of the second largest owner. Piedmont, another nonowner, averaged a 5.5 percent profit margin during the deregulation years of its independent existence, which again was several times better than the experience of the two largest CRS owners. USAir did not buy into the Apollo CRS until 1988. Before that, while it was still a nonowner, it averaged a 6 percent margin, and during that period was second only to Southwest in average profit margin. In contrast, the record of the five original CRS owners is as follows: Eastern went into bankruptcy in 1989 and ceased all operations in 1991. Trans World has been close to bankruptcy, has missed interest pay- ments on its debt, been threatened with creditor recapture of its aircraft, and has been forced to sell off major parts of its route system to raise cash. Delta has been generally profitable, but clearly cannot attribute such profit to its CRS status. Its CRS was so weak that it had to give up its independent operation and merge into another carrier's system. The remaining two original CRS owners (American and United) have averaged net profit margins of under 3 percent and 2 percent, respectively, during the period of deregulationâscarcely outstanding records. Incidentally, in recent years, several carriers acquired equity positions in CRS systems (i.e., USAir, Northwest, and Continental). In no case was such change in CRS status accompanied by improved financial results. More generally, the claim of CRS advantage fails to take adequate account of the substantial changes that have occurred in these systems
390 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION since 1984. The biased display of flight listings on CRS screens was eliminated in November of that yearâmore than 6 years ago. Further, technological changes now enable travel agents to access directly the internal reservation systems of carriers other than their CRS supplier, thus overcoming the initial concern that they would lack confidence in the accuracy of flight or fare information for such other carriers. Virtually no extra time is required to employ this direct access and there is no transaction cost to the travel agent for doing so. If---as has been suggestedâtravel agents had any present concern about the accuracy of schedule or fare information of non-host carriers, the direct access technology gives them a painless way to overcome such concern. Nor is there any sound basis for attributing to CRS ownership some greater ability to design pricing and other marketing programsâas claimed in the committee's report. All carriers can obtain marketing information from any CRS system, and therefore (if they choose) can equally apply such information to the design of their pricing or other marketing programs. If the concept of barrier to entry is to have more than academic semantic meaning, it surely should signify an incumbent's advantage that is so strong that it truly deters entry by another firm. The above financial record of CRS owners versus nonowners, coupled with the changes in display rules and technology since 1984, make it highly implausible that CRS ownership is in fact a meaningful barrier to entry. Nor does this record justify the regulatory intervention into CRS operations, which the com- mittee's report recommends. PROPOSED CONGESTION-RELATED AIRPORT FEES The committee recommends that the U.S. Department of Transportation (DOT) encourage communities to impose congestion fees as a means of reducing the airlines' scheduled activity at times and locations which involve congestion. Discussed in the report are a variety of practical problems associated with this conceptâbut the possible ramifications of those difficulties are inadequately explored. A few comments will illustrate why the committee's consideration of the issue has not advanced to the point of justifying its affirmative recommendation. First, the report notes that the cost of the increased landing fees could be substantialâbut it does not give even a ballpark indication of just how large a new cost this may impose on an industry already beset with financial difficulties. As one clue to possible dimension, it is worth noting the cost
Appendix D: Dissenting Statement 391 projected in one of the studies cited by the committee's report, in support of this concept (i.e., the Morrison and Winston study of 1989). That study estimated an annual landing fee increment of about $11 billion in 1988 dollars. That would be close to a 10-fold increase over existing landing fees. For further perspective, such annual increase in landing fees would be roughly three times the highest operating profit the industry experienced in any year since the start of deregulation. Before endorsing the concept, the committee should either have satisfied itself that such projected cost impact is unrealistically high or that the potential benefits would be com- mensurate with so extremely high a cost. Second, the committee's report is internally inconsistent as to how the congestion fee would be administered. At one point, it recognizes that communities might divert these substantial funds to purposes other than airport expansion, and (to avoid That risk) the report suggests a federal congestion fee, which would go into a central national fund. But that idea is not carried over into the report's recommendation. The report recom- mends that "the DOT should permit and encourage airports to experiment with congestion pricing..." (emphasis added). In other words, the rec- ommendation calls for local administration rather than a federal fee, and therefore would create the very opportunity for local diversion of the substantial revenues, which the report concedes to be a potential problem. Furthermore, if the concept were to take the form of federally admin- istered fees, that would open up other issues, not discussed in the report. The report suggests that the revenue from such federal fees would be distributed to those airports which present "the most cost effective pro- posals for expanding [capacity]." Because of the varying ability of dif- ferént airports to undertake capacity expansion, this concept would appear to involve considerable cross-subsidization between different locations. This. can be illustrated with reference to the situation at New York's LaGuardia Airport. The report indicates that in 1989 LaGuardia had the highest delay ratio of any airportâand presumably would therefore have particularly high congestion fees to force some of the activity away from this airport. The ability to plow these funds back into expansion of LaGuardia, however, would be limited by the locational aspect of that airport and its surroundings. Therefore, it would seem that the airlines (and pas- sengers) using LaGuardia would be expected to pay substantial added fees, which (after passing through a central fund) would be used to subsidize airport expansion elsewhere. Thus, what started out as a proposal for easing congestion would end up as a nw form of geo- graphic redistribution of funds.
392 WINDs OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION The problem of airport congestion is serious, and certainly calls for urgent attention. But more harm than benefit can result from recom- mending a "solution" that oversimplifies the problem and leaves many serious questions unresolved. FARE LEVELS AT HUB AIRPORTS There have been a number of allegations in recent years that the airlines have taken advantage of their position at concentrated hubs to charge excessive fares at those locations. The committee's report is somewhat equivocal on this issue, stating that it is unclear "whether these higher yields [at concentrated hubs] stem partly from monopoly profits earned by carriers exercising market power." However, in the discussion before and after that sentence, the general impression is left that market power is probably involved. In leaving that impression, the committee has failed to bring to bear on this issue the overriding reality of the industry's extremely poor fi- nancial record. For the 5-year period 1985-1989, the airlines' net profit margin averaged a mere 1 percent. Then in 1990 a disastrous $4 billion net loss wiped out even that meager margin. At this writing, in addition to the collapse of Eastern, four other carriers are in bankruptcy, and Trans World is on the edge of bankruptcy. There may be some tendency to dismiss recent losses as attributable to the abnormal conditions associated with the Gulf War. But the record was poor, and worsening, even before the August invasion of Kuwait. Calendar 1989 was considerably worse than 1988, and the first half of 1990 was worse still. For the 12 months ending June 30, 1990, the industry had a net loss of more than $300 million, which represented a negative swing of $2 billion from the results 1 year earlier. These overall results do not represent even a reasonable subsistence level of earnings, let alone any semblance of monopoly profits. I believe the committee should have made specific reference to this general financial record in the discussion of hub fare levels. There is obviously a serious shortfall in overall system revenue. In that circumstance, the mere fact that some locations have somewhat higher fares than other locations pro- vides no basis for regarding the former as excessive. Given the overall revenue inadequacy, a far more plausible hypothesis is that the low end of the pricing spectrum is too low (partly because of fare wars), and therefore cannot provide a valid yardstick with which to evaluate or crit- icize pricing at hubs.
Study Committee Biographical Information Joel L. Fleishman, Chairman, is Senior Vice President of Duke Uni- versity and Professor of Law and Policy Sciences. Mr. Fleishman received his A.B. and J.D. from the University of North Carolina and his LL.M. from Yale University. He has served as legal assistant to the Governor of North Carolina, Associate Provost of Urban Studies and Programs at Yale University, Associate Director of the Institute for Social Research, Vice- Chancellor for Education and Research in Public Policy at Duke Univer- sity, and Vice-Chancellor of the University. Mr. Fleishman is Chairman of Duke University's Capital Campaign for the Arts and Sciences. He authored, along with Bruce L. Payne, Ethical Dilemmas and the Education of Policymakers (1980); coedited Public Duties: The Moral Obligations of Public Officials (1981); and edited The Future of American Political Parties (1982) and The Future of the Postal Service (1983). Mr. Fleishman is also on the Boards of Directors of the North Carolina School of Arts Foundation, the National Institute for Dispute Resolution, and the North Carolina Center for Public Policy Research. He is a trustee of the Urban Institute, the John and Mary Markle Foundation, and the Center for Phi- losophy and Public Policy. He is a member of the Association for Public Policy Analysis and Management, the National Academy of Public Administration, Order of Golden Fleece, Century Association, and Phi Beta Kappa. George J. Bean is Executive Director of the Hillsborough County Aviation Authority, which owns Tampa International Airport. His civil aviation career began after World War II when he joined Northeast Airlines in Worcester, Massachusetts. He became the Manager of the Worcester air- port in 1953, and subsequently became Manager of the Greater Wilmington Airport and then the Tampa airport. Mr. Bean is the Past President of the American Association of Airport Executives (AAAE) and the Airport Operators Council International. He received the AAAE President's Award for his outstanding contributions to the airport management profession. 393
394 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION Mr. Bean is a past member of the Executive Committee of the Trans- portation Research Board (TRB). Langhorne M. Bond, Bond & Associates, received his B.A. and LL.B. from the University of Virginia and performed postgraduate work at the Institute for Air and Space Law at McGill University and the London School of Economics. He served as a Special Assistant to the Under Secretary for Transportation of the U.S. Department of Commerce, Special Assistant to the Secretary of the U.S. Department of Transportation (DOT), Assistant Administrator of the Urban Mass Transportation Administration, Executive Director of the National Transportation Center in Pittsburgh, and Secretary of the Illinois Department of Transportation. He was Pres- ident of the American Association of State Highway and Transportation Officials for 1975-1976. From 1977 to 1980 he served as Administrator of the Federal Aviation Administration (FAA): Melvin A. Brenner is President of Melvin A. Brenner and Associates. During his career in the commercial aviation industry, Mr. Brenner served as Vice President of Market Planning for American Airlines and Vice President of Market Planning for Trans World Airlines. He was responsible for all planning functions concerned with marketing for passenger and cargo services, including schedule planning, pricing, market research, route planning, fleet planning, and development of corporate marketing plans. Since 1977 Mr. Brenner has been engaged as an airline consultant. As an economist, he formerly served in a number of transportation-related policy positions in the U.S. government, including Assistant to the Vice Chairman of the Civil Aeronautics Board and Aviation Advisor to the Under Secretary of Commerce. He has been active in industry association affairs with both the Air Transport Association of America and the International Air Transport Association. He has published several articles on aviation and deregulation and is a coauthor of Airline Dereg- ulation (1985). John J. Fearnsides is Senior Vice President and General Manager at The MITRE Corporation and Director of its Center for Advanced Aviation System Development. Dr. Fearnsides directs MITRE's aviation and air traffic control work in support of the FAA and foreign governments. He is also responsible for all of MITRE's weather-related research in support of the FAA, the National Weather Service, and the National Oceano- graphic and Atmospheric Administration. Before joining MITRE, Dr. Fearnsides served for 7 years in various capacities at the DOT, including
Study Committee Biographical Information 395 the positions of Deputy Under Secretary, Chief Scientist, Acting Assistant Secretary for Policy, and Acting Administrator of the Research and Special Programs Administration. From 1968 to 1972 he served on the technical staffs of Bell Telephone Laboratories and Bellcomm, Inc., where he spe- cialized in the application of mathematical systems theory to investment decision, traffic management, and spacecraft altitude control systems prob- lems. He received his B.S., M.S., and Ph.D. in electrical engineering. From 1962 to 1968 he served on the faculties of the electrical engineering departments at the University of Maryland and Drexel University. He has served on several National Research Council panels on transportation and as Adjunct Professor of Engineering and Public Policy at Carnegie-Mellon University. Cornish F. Hitchcock is Legal Director of the Aviation Consumer Action Project. He received his B.A. from the University of Chicago and his J.D. from Georgetown University Law Center. He served as Director of the Aviation Consumer Action Project (ACAP) for several years before also joining the Public Citizen Litigation Group in 1981, where he spe- cializes in aviation issues. He has testified before Congress on more than 40 occasions concerning a wide range of aviation, transportation, and consumer issues. Mr. Hitchcock is or has been a member of the FAA's Regulatory Negotiation Advisory Committee, the Special Aviation Fire and Explosion Reduction Advisory Committee, and the International Board of the International Foundation for Airline Passengers. He has published articles on airline deregulation, air safety, and other aviation issues in law journals and other periodicals. Adib Kanafani is Professor of Transportation Engineering and Director of the Institute of Transportation Studies at the University of California, Berkeley. He received his B.S. in engineering from the American Uni- versity in Beirut and his M.S. and Ph.D. in civil engineering and trans- portation from the University of California, Berkeley. During the past 20 years Dr. Kanafani has been involved in many aspects of transportation research and has special expertise in aviation issues. He has studied all aspects of aviation: demand, airport feasibility studies, capacity analysis, and economics. Dr. Kanafani has also consulted directly with airports all over the world. He is Associate Editor of Transportation Research and two other transportation journals and has authored or coauthored articles on aviation safety since deregulation, as well as numerous other articles on aviation. Dr. Kanafani received the Walter Huber Research Prize from the American Society of Civil Engineers in 1984.
396 WINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION Todd R. La Porte is Professor of Public Administration in the Department of Political Science at the University of California, Berkeley. He received his B.A. from the University of Dubuque and his M.A. and Ph.D. from Stanford University. He has taught public administration at the University of Southern California and Stanford University. He joined the faculty of the University of California, Berkeley, in 1971, where he served as As- sociate Director of the Institute for Government Studies. He has served as a consultant to a variety of organizations, including the Jet Propulsion Laboratory, the National Science Foundation, Stanford Research Institute, the National Aeronautics and Space Administration Ames Research Cen- ter, and the Naval War College. Dr. La Porte has served as a member of the National Research Council Committee on Radioactive Waste Man- agement and the Panel on Implementation of Environmental Standards for the Geological Disposal of Nuclear Waste. He has published papers on aviation, air navigation, and traffic control through professional societies and organizations such as the American Institute of Aeronautics and As- tronautics and the Institute of Navigation, in which he holds memberships. He is also a member of the American Association for the Advancement of Science, the American Political Science Association, and the American Society of Public Administration. Michael E. Levine is Dean of the Yale University School of Organization and Management and William S. Beinecke Professor of Management Studies. He received his B.A. from Reed College and his J.D. from Yale University. From 1968 to 1987, Dean Levine served as a member of the faculties of the California Institute of Technology, where he was Henry R. Luce Professor of Law and Social Change in the Technological Society, and the University of Southern California (USC). In 1978 he became the first Director of the Bureau of Pricing and Domestic Aviation of the Civil Aeronautics Board (CAB) and subsequently became General Director of International and Domestic Aviation. For his work at the CAB he received the board's award for Excellence and Distinguished Service. After leaving the CAB, Dean Levine joined Continental Airlines as Executive Vice President for Marketing. He then served as President and Chief Executive Officer of New York Air. Dean Levine returned to academia in 1984 as the William T. Dalessi Professor of Law at USC. He became the General George Rogers Clark Professor of Management Studies at the Yale Uni- versity School of Organization and Management in 1987 and Dean of the school in 1988. He has published numerous articles in scholarly journals, including several on airline deregulation, including his 1965 article, "Is Regulation Necessary? California Air Transportation and National Reg-
Study Committee Biographical Information 397 ulatory Policy" in The Yale Law Journal and his 1987 article, "Airline Competition in Deregulated Markets: Theory, Firm Strategy, and Public Policy" in The Yale Law Journal on Regulation. He served as a member of the President's Aviation Safety Commission in 1987-1988. Dean Lev- ine sits on the boards of Reed College, the UNR Trust, and the Institut du Transport Aerien in Paris. James W. Mar retired in 1990 from his position as Hunsaker Professor of Aerospace Education at the Massachusetts Institute of Technology (MIT). Dr. Mar received his SB., SM., and Sc.D from MIT. He served as head of the structural test section of Curtiss-Wright Corporation before joining the structural research group of MIT. He became a member of the faculty in 1950 and was head of the Department of Aeronautics and Astronautics from 1981 to 1983. Dr. Mar has served on several panels and committees of the National Research Council, including the Panel on Thermal Pro- tection Systems of the Materials Advisory Board, the Committee on Ther- mal Protection for Aerospace Vehicles, and the Committee on Structural Design with Fibrous Composites. He served as Chief Scientist of the Air Force from 1971 to 1972 and as a member of the Vehicle Panel of its Scientific Advisory Board. Dr. Mar is a member of the National Academy of Engineering. John R. Meyer is the James W. Harpel Professor for Capital Formation and Economic Growth of the Harvard University Kennedy School of Government. He received his B.A. from the University of Washington and his Ph.D. (with the David A. Wells prize) from Harvard. Dr. Meyer began his career at Harvard as a Junior Fellow in 1953 and subsequently became a Professor of Economics. He has been James W. Harpel Professor for Capital Formation and Economic Growth since 1983. He is or has served on the Boards of Directors for several corporations; he was Chair- man of Union Pacific Corporation and is currently a Director. Dr. Meyer has coauthored several books; his works in transportation economics in- clude Competition in the Transportation Industries (1959), The Urban Transportation Problem (1965), Autos, Transit and Cities (1981), Airline Deregulation: The Early Experience (1981), Deregulation and the New Airline Entrepreneurs (1984), and Deregulation and the Future of Intercity Passenger Travel (1987). Dr. Meyer has also contributed to several articles in professional journals. He has served as a Trustee of Mutual Life In- surance, President of the National Bureau of Economic Research, and a Member of the Presidential Task Force on Transportation and the Presi- dential Commission on Population Growth and American Future. He is a
398 IVINDS OF CHANGE: DOMESTIC AIR TRANSPORT SINCE DEREGULATION Fellow of the American Academy of Arts and Sciences, the Econometric Society, the American Economic Association, the American Statistical Association, the Council on Foreign Relations, and the Economic History Association. Robert P. Nebschel is Professor of Corporate Governance of the J.L. Kellogg Graduate School of Management and Director of the Transpor- tation Center at Northwestern University. Mr. Neuschel received a B.A. from Denison University and an M.B.A. from Harvard Business School. He retired as a senior partner in 1979 after 30 years with McKinsey & Company, Inc., an international management consulting firm, where he served clients in North America, Europe, and South America in strategic planning, organization, management development, and logistics. He serves as a member of the Board of Directors of several industrial corporations and is a trustee of Loyola University in Chicago. Clinton V. Oster, Jr., is Professor and Associate Dean of the Indiana University School of Public and Environmental Affairs. He received his B.S.E. from Princeton University, M.S. from Carnegie-Mellon Univer- sity, and Ph.D. from Harvard University. Before joining the faculty at Indiana, Dr. Oster was a Research Fellow at the Harvard Business School and Harvard University. Dr. Oster served as the Research Director for the President's Aviation Safety Commission in 1987-1988. With John Meyer he has written major evaluations of aviation deregulation: Airline Dereg- ulation: The Early Experience (1981), Deregulation and the New Airline Entrepreneurs (1984), and Deregulation and the Future of intercity Pas- senger Travel (1987). He has coauthored several articles on aviation safety and airline competition and is coauthoring a book on aviation safety. Judith A. Rogala is President and Chief Executive Officer of Flagship Express, a $60-million cargo air carrier with worldwide operations. She has more than 28 years experience in the aviation industry. Before joining Flagship, Ms. Rogala was the Senior Vice President for Service Integration of Federal Express, where she was in charge of integrating the Flying Tiger work force with that of Federal Express. She also served as the Senior Vice President for Central Support Services, which included re- sponsibilities for alihub operations, properties, logistics, and sort systems, and management responsibility for 19,000 employees. Before joining Federal Express in 1980, Ms. Rogala spent 18 years with Trans World Air- lines in numerous capacities, including General Manager of the Pacific Northwest.
The Transportation Research Board is a unit of the National Research Council, which serves the National Academy of Sciences and the National Academy of Engineering. The Board's purpose is to stimulate research concerning the nature and performance of transportation systems, to disseminate the information pro- duced by the research, and to encourage the application of appropriate research findings. The Board's program is carried out by more than 300 committees, task forces, and panels composed of more than 3,700 administrators, engineâ¬rs, social- scientists, attorneys, educators, and others concerned with transportation; they serve without compensation. The program is supported by state transportation and highway departments, the modal administrations of the U.S. Department of Trans- portation, and other organizations and individuals interested in the development of transportation. The National Academy of Sciences is a private, nonprofit, self-perpetuating society of distinguished scholars engaged in scientific and engineering research, dedicated to the furtherance of science and technology and to their use for the general welfare. Upon the authority of the charter granted to it by the Congress in 1863, the Academy has a mandate that requires it to advise the federal gov- ernment on scientific and technical matters. Dr. Frank Press is president of the National Academy of Sciences. The National Academy of Engineering was established in 1964, under the charter of the National Academy of. Sciences, as a parallel organization of out- standing engineers. It is autonomous in its administration and in the selection of its members, sharing with the National Academy of Sciences the responsibility for advising the federal government. The National Academy of Engineering also sponsors engineering programs aimed at meeting national needs, encourages ed- ucation and research, and recognizes the superior achievements of engineers. Dr. Robert M. White is president of the National Academy of Engineering. The Institute of Medicine was established in 1970 by the National Academy of Sciences to secure the services of eminent members of appropriate professions in the examination of policy matters pertaining to the health of the public. The Institute acts under the responsibility given to the National Academy of Sciences by its congressional charter to be an adviser to the federal government and, upon its own initiative, to identify issues of medical care, research, and education. Dr. Samuel 0. Thier is president of the Institute of Medicine. The National Research Council was organized by the National Academy of Sciences in 1916 to associate the broad community of science and technology with the Academy's purpose of furthering knowledge and advising the federal government. Functioning in accordance with general policies determined by the Academy, the Council has become the principal operating agency of both the National Academy of Sciences and the National Academy of Engineering in providing services to the government, the public, and the scientific and engineering communities. The Council is administered jointly by both the Academies and the Institute of Medicine. Dr. Frank Press and Dr. Robert M. White are chairman and vice chairman, respectively, of the National Research Council.
Transportation Research Board National Research Council 2101 Constitution Avenue, N.W. Washington, D.C. 20418 ADDRESS CORRECTION REQUESTED NON-PROFIT ORG. U.S. POSTAGE PAID WASHINGTON, D.0 PERMIT NO. 8970
