Challenge to Manufacturing: A Proposal for a National Forum. (1988)

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.

National Manufacturing Policy: An Industry Perspective Laurence C. Seifert and Alfred D. Zeisler The United States needs a national policy for accelerating and sustaining productivity im- provements to regain global manufacturing competitiveness. This paper establishes a basis for formulating such policy initiatives and supporting government actions. The need for national initiatives has been widely discussed. Many programs aim at improving the productivity and performance of U.S. manufacturing firms—including De- partment of Defense projects, National Sci- ence Foundation research grants to universi- ties, and attempts to shape U.S. trade policy. Government policy makers appear to desire a proactive national manufacturing policy, but they lack a theory for selecting and integrat- ing components of a meaningful policy. Such a "theory," offered in this paper, would pro- vide a foundation for effective initiatives. OPPORTUNITIES IN A CHANGING ENVIRONMENT U.S. manufacturing is now widely perceived as having lost its world leadership position. There is evidence of rebound, albeit insuffi- cient to ensure recapturing that leadership. By working together, government, industry, la- bor, and academia can correct this situation. The basis for competitiveness rests with two principles: • End-to-end business performance, from com- ponent and product design to customer sup- port, must be of superior quality. • Competing in global markets mandates equal access to all markets for all competitors; e.g., "level playing field." The first principle is dealt with in this paper; the second is an issue for trade policy, and is currently being considered by the Congress. The broad perspective suggests that the fol- lowing aspects of manufacturing need atten- tion: • Development of process technology, the D of R&D. Current levels of process research, pub- licly or privately supported, may be adequate but we have not demonstrated the ability to use our technology advantage to achieve results comparable to those of foreign com- petitors. • Skills mobility among the existing work force. As a nation, we worry about effects of technology and foreign competition on our work force, yet there is a shortage of skilled workers. Constraints on work force mobility must be adjusted to accommodate competi- tive realities and put the right people in jobs that need doing. Our mindset must change from protecting the status quo of our work force to reshaping it. • Shortage of software skills. We must address both quantity and quality shortfalls. • Vulnerability in certain underlying technolo- gies. Materials and semiconductors, especially, are a vital underpinning for a wide range of industries. • Lagging deployment of process control tech- nology and systems, for both production and design processes. Proven benefits are not being realized. If we are to compete with other nations who maintain a lower standard of living, and therefore enjoy lower labor costs, we must use out technology to extend our process capabilities to lower overall costs. • Economic system disincentives to investment in manufacturing, as perceived by industry. More technical research, in and of itself, will not address these problems. Government support of the manufacturing sector has tra- ditionally focused on research programs and our fine university system. In these, we have maintained global leadership. These strengths 26 SE1FERT AND ZEISLER

are necessary, but not sufficient to achieve our broader goal. Moreover, despite our technical research, foreign competitors have largely caught up in technology innovation. Even where they have not, rapid diffusion of new technology has neutralized the traditional U.S. advantage. We must not delude our- selves with a technical Maginot line in our battle for manufacturing leadership. THE TARGET: PRODUCTIVITY GAINS The United States can no longer expect to compete globally based upon technology alone. Instead we must achieve continuous cost improvement in both the operations and infra- structure of manufacturing. Annual produc- tivity gains approaching 10 percent would put this country's manufacturing base on a course to intercept the productivity gains of the world's best in less than a decade (see Figure 1). Productivity gains at this level are possible—achieved in Japan between 1960 and 1973, and recently in several U.S. firms. Over the years 1979-1986, U.S. manufacturing productivity improved approximately 3.5 percent per year while that of Japan increased 5.6 percent, and Federal Republic of Germany 2.7 percent. Beginning in 1986, the United States may have initiated a sustainable rate of productivity increase that is greater than that of other nations. This country has always stretched to meet important, difficult goals. A 10 percent annual productivity increase is a worthy objective for U.S. manufacturing. However, each industry, and within it each firm, must be free to evolve its own model, set its own objectives, and select tools to meet the objectives. An Era of Change: Key Trends in Manufacturing We believe, based on discussion with others, that industry is in general agreement about changes affecting it during the latter decades of the twentieth century. These changes trans- late into challenges—or opportunities. The following are critical factors for success dur- ing the coming period. Creative approaches in these areas are needed to support a 10 per- cent per year national productivity initiative: • Traditional job categories often lack rele- vance to information-intensive contemporary manufacturing. Industry must have incen- tives and freedom to seek out talent for new jobs among workers in traditional classifica- tions and move those workers into training or new jobs—for example, blue-collar workers into the ranks of software workers. • The welfare of our natural environment has won the widespread support of manufactur- ing industry. But environmental policy re- quires stable ground rules, lest confusion hinder investment and deflect resources un- necessarily. • Process control has taken on paramount importance to all aspects of manufacturing. The tools of control (sensors, computers, soft- ware) must continue to improve and be inte- grated into systems—in the United States, by domestic firms, whenever possible. Process development costs must also be acknowl- edged, in many cases, as beyond the re- sources of individual firms. This accelerates the trend toward industrial alliances, includ- ing cross-border alliances. FIGURE 1 Manufacturing productivity growth. U.S. Japan Mfg Non-Farm Business Mfg. 1979-1986 1985 1986 3.5% 5.1% 3.7% 1.1% 5.6% 7.3% 2.8% 1977 1986 1998 SOURCE. Based on dat a extrapolated trorn Halsopoulous and Brooks "Capital Formation in It* U S and Japan' and MLR Dec 1987. Vol 110. No 12 (lor productivity ligures) NATIONAL MANUFACTURING POLICY 27

• Lack of educational preparedness and re- training opportunities retard our efforts to boost productivity. Many studies argue for new emphasis on manufacturing processes in engineering schools, anticipating shifts in training requirements, and tightening rela- tions between colleges and industry. • U.S. industry has not aggressively and promptly deployed available new processes that are flexible and support rapid changes in product output. Better deployment of new process capabilities is required where rapid changes in production rates and product models are called for, and to support the growing trend toward customization of prod- ucts and systems for individual users. • We need more process technology experts. The pool of expertise must be balanced be- tween citizens of the United States and non- citizens, who may be unable to serve in the defense sector of U.S. manufacturing. TABLE 1 Manufacturing Shipments Dec87($B) A86(%) Food (S.I.C. 20*) 28.3 + 5.4 Transportation Equiptment (37**) 27.0 -4.4 Electrical Machinery (36**) 20.7 +10.7 Nonelectrical Machinery (35") + 6.2 Chemicals (28**) 18.2 +10.5 Petroleum (29) 10.7 +10.8 Fabricated Metals (34) 10.4 + 1.2 Primary Metals (33) 102 +37.2 Paper (26) 10.2 +15.1 Printing (27) 9.6 + 5.5 Rubber and Plastics (30) 62 +11.3 Instuments (38) 56 + 3.5 Lumber (24) 4.9 +10.8 Apparel (23) 4.9 + 5.5 Stone, Clay, Glass (32) 4.6 +10.2 Textiles (22) 4.3 -1.0 Furniture (25) 2.9 +10.8 Miscellaneous (39) 2.5 +10.8 Tobacco (21) 2.2 + 3.5 Leather(31) 0.8 + 5.4 Total Manufacturing $204.5 + 7.1 ' Standard Industrial C&silicalion/Figures arc unadjusted lor seasonal variation Source US DepartmentolCommerce "Focus ol preliminary analysis SOURCE Industry Week, 3/7/88 DEFINITION OF MANUFACTURING Manufacturing industries in the United States are categorized by the U.S. Department of Commerce in the Standard Industrial Classifi- cation (SIC) Codes 20-39. This broad array of industries ships approximately $200 billion worth of goods per month and employs about 19 million people (see Table 1). In this discus- sion, manufacturing industries and compa- nies, not factories, are the focus. A traditional categorization views manufac- turing as a stand-alone function, supported by other organizations such as product de- sign, sales, distribution, and so on. Manufac- turing accounts for about 20 percent of na- tional employment, about 23 percent of total output, 60 percent of exports and 75 percent of imports. Manufacturing generates about 20 percent of the GNP. Manufacturing Redefined Manufacturing today must be regarded in a perspective sometimes called the "product- realization process" (the term used in AT&T). In this now widely understood view, the cus- tomer is at the heart of all the activities that result in a product, or service (see Figure 2). Manufacturing is no longer a stand-alone function but is instead seen as one element in an end-to-end process—from marketing or technological innovation, through product design and manufacture, to delivery and after-market service. Systems engineering disciplines are used in the manufacturing process to bring major bene- fits in efficiency and shorter concept-to-cus- tomer intervals without significant financial investments. Critical to product realization are revised internal measurement systems that are customer based and customer respon- sive. A successful process is determined in the marketplace, not simply by meeting internal criteria. Besides the product realization perspective, other factors also redefine manufacturing in the late twentieth century. Manufacturing industry now experiences growing depend- ence on real-time information transfer, soft- ware, unique materials, integrated circuits (including photonics), worker knowledge, 28 SE1FERT AND ZEISLER

and sophisticated process technology. Aggres- sively addressing these areas now will enable United States manufacturing to step deci- sively ahead. The application of manufactur- ing technology by itself will provide less com- petitive differentiation than at present. (We must nevertheless take better advantage of available new technologies than we now do.) A new relationship between producers and suppliers is evolving to support higher levels of quality and customer responsiveness. Plants are being opened close to suppliers and customers in order to reap the economies of just-in-time procedures. In international markets in-country value-added regulations will continue to impact supplier decisions. The traditional supplier/customer model is changing to a partnership arrangement. Manufacturing of all types will depend in- creasingly on the intellectual capabilities and skills of its workers. The "information trades"—specialists who can integrate infor- mation processing in a company or industry, and replicate those processes in global operations—are of growing importance. The information tools—for resource planning, accounting, just-in-time—will be less indus- try-specific and more generic, as the customer base becomes more global. Critically impor- tant will be the ability to facilitate technology transfer between individuals, companies, and geographic locations. A depository of infor- mation and an easily accessible system to permit acquiring the information will be of immeasurable value. As assembly and fabrication costs improve, white-collar infrastructure costs and produc- tivity must improve. All processes within a manufacturing company must receive rigor- ous scrutiny and undergo the discipline of quality and productivity initiatives. Keeping Our Eyes on the Ball It is possible to rationalize away the chal- lenges faced by U.S. manufacturing. Recent positive trade and productivity figures may give an appearance that we have turned a corner—that perhaps there is no need to forge national initiatives. Missing in this conclu- sion, however, are a series of facts, including the following: • Aggregate statistics are deceiving. A major shift out of declining industries has taken place. This has seemed to boost productivity FIGURE 2 AT&T product realization system. Market and Productflanmng Cusiomjir Needs Product Design and Process Developmenl Account Management •:": Orftrs Customer Service 4 Materials Component and I Device Manufacture Product Distribution. Staging. Installation, and Repair Products External Suppliers rates, but does not take account of the pos- sible detrimental impact on long-term na- tional interests. • U.S. influence on the evolution of the world's industrial style (i.e., the equipment used, design configurations, standards and management) has diminished steadily. As a result, our ability to sell plant and industrial tools to other nations is lessened. • Technological leadership is now generally distributed among companies and among nations. What leadership there is is more transient than ever. Thus, all industry is at higher risk now than in the past. For example, leadership in transportation equipment may depend on composite material leadership; leadership in communications equipment depends on photonics leadership; future power supply leadership may depend on leadership in superconductivity. • Apparent gains in manufacturing produc- tivity must be viewed in light of shifting cur- rency levels, which lead to increases in vol- ume and capacity utilization that may be transitory. Maintaining high utilization rates is not assured without progress in production fundamentals. Only this will allow the United States to compete internationally with a broad range of products and over fluctuating ex- change rates. NATIONAL MANUFACTURING POLICY 29

• Employment in the growing service sector is rising, but this sector is intimately tied to manufacturing industries, not independent of them. The service sector exhibits less produc- tivity growth than manufacturing and is ripe for further competitive inroads. Any reduc- tion in our manufacturing base would have a broader impact on employment than in the past. • Manufacturing capital investment is often delayed by fluctuations in, and uncertainty about, the tax treatment of such investments. Comparative R&D Investment Today R&D expenditures in the United States have risen in recent years, but from a manufactur- ing perspective the expenditure on process development, as opposed to research, is insufficient. A reapportionment of resources is called for. The share of industry's investment in R&D expended for "development" fluctuates be- tween 65 percent and 75 percent of a $65 bil- lion total, that is, about $40 billion per year. Yet there appears to be insufficient spin-off to generate world class processes. This may in part account for the U.S. lag in innovation and time-to-market. Japan today leads this country in R&D invest- ment in several industrial segments (food, textiles, metals, rubber), and is closing the gap in others. The R&D intensity of manufac- turing firms has risen faster in Japan than in the United States, although there is some evi- dence that domestic increases are occurring. Company-financed R&D in manufacturing was 1.3 percent of sales in Japan in 1970 and 2.7 percent in 1980. In the United States the figures were 2.2 percent and 2.8 percent re- spectively. Japan's rate of increase in R&D exceeds our own. It would appear that what we do spend on R&D is not buying what we need. For ex- ample, large sums have been spent on robot- ics, but the results are questionable. Major investments in computers have also returned less than promised productivity. U.S. innova- tion is being openly questioned. Evidence suggests that the Japanese are often able to develop and introduce commercial products more quickly and less expensively than American firms. Why does a nation using less automation than our own appear so productive? It is noteworthy that R&D in Japan is undertaken at the suggestion of customers and produc- tion workers twice as often as it is in the United States. This might indicate that R&D expenditures directly responsive to customer desires and supportive of the needs of pro- duction have immediate economic benefit. In the United States, the Department of De- fense is one entity with a demonstrated abil- ity to move from research to development, with a commitment to improving and imple- menting new manufacturing processes, qual- ity controls, and management techniques. Expansion of programs here could well serve the longer-term commercial interests of the United States. CRITICAL TECHNOLOGIES FOR NATIONAL COMPETITIVENESS There are four critical technologies that are, or will be, a basis for future U.S. competitive- ness—materials, semiconductors, software, and process control equipment. Materials The world of advanced materials is changing drastically. In some cases, the United States has lost its international competitive edge. We increasingly depend on overseas suppliers of advanced or technologically critical engi- neered materials. The new materials acquire their special value from integration into complex systems. For example, as an ingot, aluminum is worth about $1 per pound; shaped, it is worth about $5 per pound; but applied as a microconduc- tor, its value rises to about $5 million per pound on silicon chips. The materials field suffers from a chicken-egg syndrome. It is too expensive for a supplier to develop a material for an application without a commitment by a user. But users do not commit without know- ing what the material can do. Many U.S. firms are having difficulty justifying development costs and risks, and the United States is losing its domestic supplier base. From an industry viewpoint, what is needed is: • A refocusing of education to balance the use of resources in research and applications. • Development of related processes for high- quality, effective bonding, forming, and drilling. 30 SE1FERT AND ZEISLER

• Stable administrative procedures to deal with questions of environmental hazards associated with material processing. • Development of a methodology to support costly material process scale-up. To ease the industry efforts under way, there are several roles government can play. These include: • Focus on the issue of technology transfer in materials and related process technologies. • Support advanced materials engineering in academia. • Provide expanded support for materials consortia composed of industry and universi- ties. • Ensure a level playing field for both import and export of materials. • Expedite release of classified information on materials, consistent with security needs. Semiconductors For the semiconductor industry a most critical issue is the inability of individual firms to fund adequately the broad range of support- ing technologies. Among the efforts consid- ered most important are: • The development of alternative litho- graphic technologies that will allow the manufacturer to choose the most cost effec- tive integrated circuit (1C) patterning tech- nique. • The development of fully automated flex- ible processing lines that provide rapid device and product turn-around times. • The availability of packages that will bear multiple chips, and the supporting package technology, within five years. • Accelerated improvement in test capabili- ties, including computer-generated test programs, networking of test systems world- wide, and broad application of built-in self- testing capabilities. The semiconductor industry will continue to be a global industry. Influencing the locations of design and fabrication of semiconductors will be the cost of labor, the distance to tech- nological leadership and educational institu- tions, and ease of access to controlled markets for select products. Nationally based support of 1C wafer fabrica- tion, assembly, and test will accelerate the dispersion of 1C manufacturing throughout Europe and Asia. Asian governments, such as in China and India, are willing to subsidize much of the capital required for a strong posi- tion in 1C manufacture. Another influence on the location of tomorrow's design centers is the large proportion of trained university graduates relative to available jobs in the Asian countries, coupled with the relatively low investment for establishing 1C design capability there. To make way for the United States to regain a preeminent position in design and fabrication of semiconductors, several policy changes and legislative actions, in addition to corpo- rate initiatives, are suggested: • Continue to encourage joint partnership through favorable antitrust and tax legislation and R&D incentives. • Provide appropriate incentives to facilitate the availability of capital necessary for invest- ment in plant and people. • Through combined govemment-indusrry- academic efforts, increase emphasis on educa- tion in 1C design and manufacturing, and related software. • Stimulate high school students to major in the sciences and engineering in college. • Develop a government program to support industry education programs to augment university training in critical technical fields. Software As the software content of products and proc- ess expands, the need for software grows faster than industry's current ability to create it. Yet rapid creation and support of software are essential to meet such objectives as re- sponsive product customization. The development time for finished software is often unpredictable, and its cost often higher NATIONAL MANUFACTURING POLICY 31

TABLE 2 Potential Impact Matrix PROGRAM GRANTS Universities National Laboratories Consortia Private Industry Primary/Secondary Education INDIVIDUAL GRANTS Faculty Fellowships Student Loans Capitation DEVELOPMENT PROJECTS Military Special Consortia Industry INCENTIVES R&D Tax Credit Capital Investment Tax Credit Cooperative Venture Relief H=High M=Medium L=Low L M H H n/a M L L H L M M H L L M H H H M H M H H H H M H L M H H n/a L M L L The training of software developers today falls largely to industry itself. Better partner- ships between industrial and academic com- munities would help increase the exposure of young programmers to industry require- ments. Process Control The key to quality products is the ability to monitor and control all aspects of processes being used—in other words, process predicta- bility. State-of-the-art process control is built around a combination of special-purpose hardware, high-precision sensing and effect- ing mechanisms, statistical quality control algorithms, software, mini and micro comput- ers and multichannel information transfer networks, all assembled into an integrated system. The future of process control technol- ogy will be based on real-time computing as technology in the physical sciences brings this field along. Enhanced process control grows more impor- tant as processing moves into suprahuman modes (tasks not executable by people). An uninterrupted supply of world class proc- ess control equipment may require national initiatives in the form of incentives, use of the resources in our national laboratories, and support of consortia. than projected. For companies with a long history of using software, the costs can be highest of all, because they not only face the cost of software development for new activi- ties, but must also support the maintenance of large volumes of software currently resid- ing in working systems. Software itself is complex and requires highly trained people to create it. There is a shortage of such skilled people. Opportunities to work around the shortage include: "open-systems" computing architectures; more modular con- struction of programs; improved diagnostic systems; greater functionality on ICs; and greater use of intelligent systems, thereby decreasing error carry-through. These are now being called CASE (computer-aided software engineering) capabilities. RECOMMENDATIONS A U.S. public policy that supports essential technology building blocks would be useful. Specific recommendations are as follows: 1. Stimulate process development and deployment by fostering an atmosphere in which a signifi- cant proportion of available government and industry R&D expenditures is utilized for this effort. Parallel efforts can be taken to bring this to pass; first rebalance R&D funding to focus on this goal; second, support a program to facilitate related process technology trans- fer on a national level (consortia formation and capital investment incentives would help meet this need); third, provide financial sup- port to help government and academic re- search personnel become familiar with industry's manufacturing facilities and needs. 2. Enhance work force mobility through innova- tive education and reeducation programs. 32 SE1FERT AND ZEISLER

Government could support vocational re- training to prepare workers for change; pro- vide financial incentives to industry and indi- viduals for continued education; reimburse industry for broad national training pro- grams; and improve basic primary and secon- dary education, emphasizing science and math. 3. Support software R&D by promoting soft- ware skills in schools and among the existing work force; providing resources to enable expansion of CASE projects; and increasing funding for software R&D in universities. 4. Support materials and semiconductors R&D by supporting industrial consortia; facilitating rapid release of related Department of De- fense information consistent with national security; encouraging Ph.D. study by U.S. citizens; and providing incentives for indus- try research. Who Should Act? Public policy in the United States is decentral- ized and its tools are many. These include a range of program grants, individual grants, supported projects, and incentive programs. The tools are applied through institutions such as universities, schools, national labora- tories, private industry, the military, industry groups, and the tax system. Table 2 displays this information as a matrix with a range of policy tools and institutional tools on one axis and four areas of manufac- turing in need of policy attention on the other. Each intersection marked with an H is a locus for programs that could have high impact in support of manufacturing competitiveness. In work force mobility, for example, effective programs might be mounted by private in- dustry with grants and development projects, and through government student loans and grants to schools. The range of tools is broader than shown here, and the areas for application of the tools undoubtedly more numerous. Further, the impact assessments indicated are subjective. But this matrix is a start on identifying na- tional initiatives that show promise for U.S. manufacturing competitiveness. Final determination of the elements of a na- tional policy should be based on more quanti- tative evaluations of potential impacts of specific programs. Further deliberation by a national resource such as the National Re- search Council Manufacturing Studies Board could refine and augment plans for initiatives and develop a better assessment of the advan- tages of each. CONCLUSIONS Initial deliberations suggest that a national manufacturing initiative aimed at sustaining 10 percent productivity improvement per year, along with an international trade initia- tive aimed at a level playing field, would be useful. A manufacturing initiative must ad- dress the real performance needs of industry. Public programs that stimulate manufactur- ing process development and deployment, enhance work force mobility, strengthen our software R&D skills base, and support essen- tial underlying materials and semiconductor technologies would be the elements of a manufacturing initiative. As this work is based on discussion with a limited set of U.S. industries, additional study is recommended. Laurence C. Seifert is vice president of engineer- ing, manufacturing, and production planning at AT&T. Alfred D. Zeisler is AT&T's manager of industrial automation. NATIONAL MANUFACTURING POLICY 33