Summary and Assessment
Transit systems in the United States serve large and varied markets. They provide critical transportation services to the elderly and people with disabilities; carry millions of commuters each day; and are the main, and often only, means of mobility for the urban poor and others with limited access to private automobiles. In many urban areas, transit is seen as having an important role in curbing traffic congestion and air pollution. Transit systems are expected to meet these diverse user needs while controlling fare levels and expenditures. In addition, they are subject to a mix of economic, institutional, and political influences that affect the kinds of services they can offer and the means by which they can provide these services.
Given their varied circumstances, public transit systems employ many different approaches to deliver their services. Many contract with other organizations—often private companies—to provide at least some services. Indeed, the public and private sectors have long collaborated in the provision of transit services in the United States. For much of the 20th century, private companies provided most transit services under the regulatory oversight of state and local governments. During the past four decades, the model for transit service has changed as state and local governments have assumed almost exclusive responsibility for transit planning and funding. Many state and local governments now use their own facilities, equipment, and workers to deliver needed transit services. Although the private sector’s role has diminished
in comparison with its role during the first half of the 20th century, private entities still provide many transit services, usually under contract with public transit systems.
The aim of this study was to gain a better understanding of the contracting approach to public transit provision in its many forms. During the past 20 years, numerous studies have examined the effects of contracting on service cost and quality, usually by scrutinizing the experience of individual systems. While the results from these studies indicate that cost savings are a main reason for contracting, they also point to a multitude of other reasons for the practice—from the desire for more flexibility in adding and withdrawing services to the special expertise needed for the provision of particular kinds of service, such as transportation for people with disabilities. Because so much contracting occurs under so many differing circumstances, it is impractical—if not impossible—to draw general conclusions about the practice on the basis of a relatively small number of system-specific studies.
To be sure, given time and resource constraints, a study-by-study evaluation was not a practical option for this project. In any event, the committee recognized from the outset that a comprehensive review of past studies on contracting would in all probability have generated more questions than answers. During the past decade, much of the debate over contracting has centered on the somewhat subjective matter of what proportion of transit agency overhead expenses should be allocated to contracted services. There is honest disagreement on which cost allocation model or accounting conventions are most suitable for particular circumstances; thus, the models and conventions used have often varied substantially from study to study. Likewise, previous studies of contracting’s effects on transit safety, on-time performance, customer satisfaction, and other aspects of service quality have varied widely in terms of study methods, assumptions, and data quality.
In the committee’s view, sorting out these differences among past studies and trying to use those studies to draw conclusions on the effects of contracting today would have proven futile given the time and resources available. Hence, the committee elected to conduct its own nationwide survey of transit service contracting practices and results. More than 500 public systems that receive federal transit aid were asked to report on their own experiences with contracting. Each public transit system has its own reasons for deciding whether to contract, and the systems’ general managers are in a good position to offer judgments on the results of those decisions. Therefore, the surveyed general managers were not given detailed instructions on how to define cost savings or measure service
quality. It was assumed that they have a sufficient understanding of their own circumstances to identify and offer reasonable assessments of results.
At the same time, the committee recognized that these respondents are likely to reach judgments based on their own vantage points as transit managers rather than as policy makers, and that they may be inclined to defend current practice and judge alternatives differently before and after a decision has been made. Thus, it is impossible to be certain whether the general managers participating in the survey accurately appraised the reasons their agencies contract (or do not) and the outcomes of contracting. Instead of simply acknowledging these uncertainties, it would have been desirable to control for them; however, time and resource limitations precluded a more elaborate survey and statistical analysis. Therefore, the survey results are presented here for what they are—the experiences and perceptions of contracting reported by general managers of hundreds of public transit systems across the United States.
Key Survey Results
The survey results are helpful in understanding the extent and methods of transit service contracting in the United States, as well as the reasons some transit systems contract and others do not. Although the survey findings do not paint a complete picture of contracting practice and experience, they reveal much about the amount of contracting that occurs, motivations for and deterrents to contracting, and levels of competition for contracted services. Collectively, the survey results show that contracting is monolithic in neither practice nor outcomes, and that contracting experiences and methods are both varied and complex.
What Has Been Learned About Contracting Practices?
The survey results, augmented by National Transit Database (NTD) data collected by the Federal Transit Administration (FTA), provide important information on the extent of contracting among federal aid recipients, how contracts are structured, and the state of competition for contracted services.
Many transit systems contract for some services, but mainly for small amounts and typically for demand-responsive rather than fixed-route bus services.
About 20 percent of all transit systems receiving federal aid contract for all their transit services, 40 percent contract for some, and the remaining 40 percent
do not contract at all. Yet even with most systems (60 percent) contracting for at least some service, the total amount of contracting is relatively small, accounting for about 15 percent of all bus and demand-responsive services provided.
More than two-thirds of transit systems have contracts for demand-responsive services, while fewer than 40 percent contract for bus services. More than half of all demand-responsive services are contracted, compared with only 6 to 7 percent of bus services.
Large systems are more likely than small ones to contract for some transit services, although few contract for more than 25 percent of their services. In relation to their absolute numbers, small systems contract less than large systems; however, small systems are much more likely than larger ones to contract for all their services. Because of their small size, the former systems—often operated by municipal and county governments—contract for either all or none of their services. Large state and regional transit systems, with many more service offerings, have greater opportunity to contract for a portion of their services.
Transit systems that have experience with contracting have found ways to exercise control over their contracted services and have sought ways to promote competition.
Most contracts are for multiyear periods, usually for 3 years with two 1-year options. The duration is sufficiently long to avoid repeated costs associated with rebidding, but sufficiently short to discourage contractor complacency.
Systems usually provide the vehicles and other major assets for bus service contracts, and they often provide them for demand-responsive services as well. By owning these key assets, transit systems can readily take back and rebid a service if the contractor fails to meet responsibilities and expectations.
Most contracts are structured to pay contractors on the basis of a predetermined fee per unit of output produced—usually hours of revenue service. The contractor is therefore responsible for controlling costs; only one-quarter of systems pay contractors on a cost-plus-fixed-fee basis.
Monetary penalties to discourage poor performance are common in contracts; fewer contracts contain monetary rewards for good performance.
Most contracted services are competitively bid and attract multiple bidders.
Most transit service contracts are awarded through competitive processes.
Most contracts attract multiple bidders; the survey results indicate 2.8 on average. Small systems have the most difficult time attracting multiple bidders;
such systems reported 2.5 bidders per contract on average, compared with 2.9 for medium-sized systems and 3.6 for large systems.
The number of bidders on contracts has been stable in recent years; however, demand-responsive contracts are more likely than bus contracts to experience a decline in bidders during rebidding.
The contracts of larger systems change hands more often than those of smaller systems. Many contracts continue to change hands even after they have been rebid multiple times. Even those contracts that have had only one contractor typically continue to attract bidder interest when they are rebid.
What Has Been Learned About Contracting Experiences?
The results of the survey of general managers are highly informative about why some transit systems contract some or all of their services, while others do not contract at all. The results reveal many common experiences with contracting, including both positive and negative effects. They also suggest ways to make contracting work better.
General managers of most transit systems are generally satisfied with their current approach to service delivery, although most have used more than one method of delivery.
Nearly 80 percent of general managers of transit systems that currently contract reported that they would contract again given the choice. In contrast, nearly three-quarters of general managers of transit systems that do not contract reported that they are satisfied with their current method of service delivery.
About one in seven general managers from transit systems that now contract and about one in four from systems that do not contract would change their current method of service delivery if given the opportunity. About one-third of transit systems that do not currently contract have contracted in the past. These results suggest that the level and nature of contracting are dynamic and that the mix of advantages and disadvantages favoring contracting or in-house methods of service provision can change for a transit system over time.
The main reasons transit systems contract for service, according to transit managers, are to reduce costs and increase flexibility to introduce new services. A main reason some do not contract is the desire to maintain control over services and operations.
Half the general managers of transit systems that currently contract reported that reducing costs, increasing cost-efficiency, and introducing new services are the most important reasons for contracting. About one-third rated as important the desire to create a more competitive and flexible environment.
Nearly 60 percent of the general managers of transit systems that do not contract for services reported that maintaining control over their operations is a main reason for not contracting. About half also reported that they do not believe cost savings, if any, are sufficient to prompt them to contract.
Neither group of general managers—those that contract now and those that do not—cited the influence of policies or laws as especially important in the decision about contracting.
General managers of most transit systems that currently contract are satisfied with the cost savings achieved. However, many cite problems with the quality of the contractor workforce, employee turnover, and customer service as negative side effects of contracting.
More than half the general managers from systems that are currently contracting identified reduced operating costs as a positive effect of the practice. General managers from both small and larger systems reported this benefit. Those from small systems also reported benefits from reductions in staffing and administrative burdens assumed by the contractor.
General managers from agencies that currently contract, as well as those that have contracted in the past, cited the loss of operational control, poor service quality, and problems with customer service as negative aspects of contracting. Small systems are more likely than large ones to view contract monitoring responsibilities as problematic.
Most general managers of transit systems that currently contract believe their contracting programs are fully or partially meeting expectations.
More than 55 percent of the general managers from contracting systems reported that their expectations from contracting have been fully met, and 38 percent reported that their expectations have been partially met.
Nearly all general managers of systems that are now contracting reported cost savings. However, shortcomings in contractor service quality and associated time demands on agency staff to ensure quality are the main reasons some believe contracting has not fully met their expectations. More than half of the general managers with partially met expectations identified service quality as an important problem. Service quality was also identified as a problem by many general managers of systems that no longer contract.
Whether To Contract and How To Make It Work
By and large, the general managers of transit systems are satisfied with the cost savings from contracting, but less satisfied with service quality. Whether through trial and error or through preparation and foresight, the general managers of those systems that contract today have learned how to make their contracting programs work better to achieve cost savings and acceptable service quality. In particular, they offered the following advice to other general managers who may be considering contracting.
Anticipate the benefits and costs of contracting, and set realistic expectations.
Take an open-minded and realistic view of contracting’s advantages and disadvantages. Conduct a full analysis of the likely consequences—not only by examining budgetary effects, but also by weighing potential effects on service quality, workforce motivation and morale, and flexibility to respond to new and changing service demands.
Consider various approaches to structuring contracts, including the option of providing vehicles, facilities, and other costly assets. The contract should be structured to prompt sufficient competition and allow the agency to take the service back should the contractor fail to meet expectations.
Conduct a review of the costs and effects of providing services directly, and use this information to evaluate all contracting options.
Establish a competitive procurement process that invites high-quality proposals and screens out unrealistic proposals and unqualified contractors.
Scrutinize contractors beforehand, through both formal and informal means. In particular, seek out other transit systems to discuss their contracting experiences in general with individual contractors.
Review proposals carefully. Begin with a process that reveals the capabilities of prospective contractors—for instance, by asking contractors for technical and business information on startup plans; assumptions about wage rates; plans for hiring, training, and retaining workers; and the qualifications of their management team.
Make internal estimates of service costs, and use this information as a baseline in assessing the credibility of contractor proposals, including those that are priced too high and those that are priced too low.
Clearly spell out all contractor responsibilities, closely monitor performance, and communicate with the contractor.
Outline all of the duties and roles of all parties to the contract. Establish a clear mechanism for making changes in contract agreements, and define all expectations with respect to service quality.
Include penalty clauses and rewards in contracts to motivate good performance.
Routinely monitor contractor performance, and provide the contractor with candid and frequent feedback.
Maintain an open and collaborative relationship with the contractor.
Insights and Ideas for Further Evaluation
In preparing this report, the committee has been careful to present only findings that are directly substantiated by the information gathered during the course of the study, particularly from the survey findings. These findings have been presented in Chapters 4 and 5 and summarized above.
However, the richness of the committee’s insights extends beyond these findings. Each of its 12 members is knowledgeable about contracting experiences across the country; thus each joined the committee with well-founded views on transit contracting. During a period of 10 months, the committee met five times and communicated frequently through telephone conference calls and e-mail. Emerging from this extensive interaction was a consensus that reached beyond the empirical results obtained during the study. These insights are offered in the remainder of this chapter. First is a discussion of some of the additional information the committee would like to have had during the course of the study—information that was not recognized as being important until after deliberation of the survey results. This is followed by the committee’s own ideas—or hypotheses for further evaluation—on how transit contracting is occurring and evolving in the United States.
Additional Information Needed
As is the case with all research efforts, the committee’s work raised questions that could not be addressed adequately during the course of its deliberations. Several issues arose repeatedly and warrant further attention.
First, the transit general managers who responded to the survey were not asked to offer assessments of the cost and performance of their internal, directly provided services. In designing the survey, it was necessary to balance the desire for more information with the need for a survey that would not be overly
burdensome. Nevertheless, information on each agency’s satisfaction with internal operations would undoubtedly have been helpful in providing an appropriate context for examining the general managers’ reports about benefits and problems associated with contracting.
Second, it would have been desirable to have more information on the extent to which the political and policy environments affected the decision to contract. Again, survey recipients were intentionally asked for a limited amount of information. Although the survey did address the effect of policies and laws on contracting, the level of detail in the responses was insufficient to support a general conclusion about the influence of the political and policy environments on contracting decisions. The survey was distributed to general managers, who have a particular vantage point and are but one group of individuals involved in the contracting decision. The experiences of individual committee members suggest that the political and policy environments, especially at the local level, can be important influences on the decision whether to contract. More information derived from a wider range of participants in transit policy making is needed to appraise political influences on the decision to contract.
Third, the committee was unable to determine with any specificity the magnitude of the cost savings from contracting and whether they change over time, nor were respondents asked to verify their perceptions of cost savings. As discussed above, the committee relied on the general managers to use their best judgment, and did not advise them on what constitutes cost savings or how to calculate those savings. The committee believes that follow-on studies aimed at quantifying the effects of contracting on transit operating and administrative costs, service reliability, safety performance, and other factors affecting the level and quality of service (such as labor productivity and relations) are warranted and will benefit from the information obtained from the survey conducted for the present study.
Finally, the committee did not examine the structure of the contracting industry to assess how trends toward industry consolidation may affect competition in the years ahead. The survey results provide a snapshot of conditions today and in the recent past. The committee does not know, nor can it predict, how the overall state of competition in the industry will change over time.
Ideas for Follow-On Study
Contracting is a dynamic undertaking: some transit systems are contracting, others are not, and some are about to undertake or discontinue the practice.
Individual circumstances can change over time, affecting the comparative advantages of contracting and direct service provision, or creating an opportunity for some other method of service delivery. The data gathered by the committee suggest that contracting, at least as perceived by the general managers surveyed, often entails a trade-off between cost containment and service quality. That some services are taken back in house may be the result of cost differentials diminishing over time. In other cases, the original desire to achieve cost savings or increase the amount of service through contracting may be replaced by a greater emphasis on improving service quality. As transit systems exert more control over service quality by imposing stringent quality requirements in contracts, it is reasonable to assume that contractor costs will increase over time. At the same time, labor unions may agree to changes in collective bargaining agreements that make direct service provision more cost-competitive, blunting differences between in-house and contractor costs.
A final and related insight concerns the nature of transit contracting. As discussed in Chapter 3, there is evidence that certain aspects of transit service contracting are difficult to articulate in a written agreement. As an example, transit systems can come to value particular qualities of individual contractors that cannot readily be specified in a request for proposals or a contract document; a transit manager may, for instance, be reluctant to switch from a contractor that provides demand-responsive services with few complaints from customers. The continuation of such relationships may be advantageous to riders as well as to the agency and contractor involved. The extent to which such service contracting relationships exist today in the transit industry and can be fostered and maintained to the benefit of riders deserves further consideration in follow-on studies.