tates have become increasingly active in promoting industrial competitiveness and economic development in recent years. S Some of these efforts involve the reorientation of existing institu- tions and programs that provide training, small business assistance, and recruitment incentives. In addition, states have undertaken a variety of new initiatives with respect to transfer, venture capital, and the modernization of established firms. An extensive literature has emerged on state economic develop- ment efforts. The results, however, have not been of much help to states in terms of developing competitiveness strategies, for two major rea- sons. First, the materials are primarily descriptive, highlighting the actions of various communities, states, and regions. Little evidence is given on the success or failure of such experiences. Moreover, for many programs, not enough time has elapsed to evaluate effectiveness, at least over the long term. Second, state experiments and initiatives have not been viewed in a larger analytical framework that would permit generalization and an understanding of the dynamic processes underlying these changes. Lacking this larger context, information about the experiences of other states, no matter how detailed or successful, is of limited value to states operating under different industrial and technological conditions. Patricia M. Flynn This article adopts production life-cycle models as a framework in which to analyze systematically the interrelationships between indus- trial and technological change, human resource needs, and state eco- Dean, Graduate School of Business, nomic development policies. This framework--in which products, pro- and Professor of Econo~nics, Bentley duction processes, and are seen as dynamic phenomena College. This article is based on re- whose locational, skill, and training requirements change as they search conducted for the National Cen- evolve--provides a conceptual model useful for evaluating and design- ter on the Educational Quality of the ing state economic development policies. Work Force at the University of Penn- The life-cycle framework suggests that states that incorporate the sylvania. dynamics of industrial and technological change into their competitive- ness strategies will reap employment and productiv- ment packages, particularly those offered by south- ity benefits that technology can provide. In contrast, ern states. states that fail to address these issues increase their Historically, North Carolina has been noted for vulnerability to the negative impacts of technological its ability to attract manufacturing plants~a majority change, including widespread unemployment and of the Fortune 500 companies have at least one plant job loss. located in this southern, right-to-work state. More recently, Tennessee, Kentucky, South Carolina, and Alabama have been successful industrial recruiters. A L Trends in State Economic Nissan plant located in Tennessee in 1980, and in Development Efforts 1985 the state won its bid for the General Motors Saturn plant. Kentucky attracted a Toyota plant in State economic development efforts revolve 1985 and was first runner-up in the Saturn contest. around three major strategies: the recruitment of South Carolina was successful in recruiting a BMW firms to the state, the development of high-tech plant in 1992, and Alabama was the site selected in start-up firms, and the revitalization of established 1993 by Mercedes-Benz for its first North American businesses. All state economic development strate- plant. gies attempt to boost the local economy. States hope More generally, states throughout the country such steps will result in net increases in the private sought to recruit high-tech industries during the late 1970s and early 1980s. These efforts included various tax and financial concessions and promises of work forces trained to accommodate the needs of individ- ual employers. State economic development efforts Recruitment efforts continue to be an active com- revolve around the recruitment of ponent of many states’ economic development plans. The competition for the Saturn plant, for example, firms to the state, the included 38 states and 1,000 local communities. Fur- development of high-tech start-up ther, state recruitment packages have become more firms, and the revitalization of complex as well as more expensive. In its winning proposal for the Saturn plant, Tennessee provided a established businesses. significant property tax abatement and infrastructure improvements and promised to spend an extra $45 million on higher education, in order to offer a range of technical courses (such as robotics and automation) employment base (direct and indirect), in state and for upgrading General Motors employees. Michigan’s local tax revenues, and in long-term economic recruitment of a Mazda plant in 1986 included $19 growth. The number of jobs created or maintained is million to train new workers, and Illinois offered $64 not the only factor to consider. The quality and level million in 1988 in hiring and training assistance in its of income associated with the jobs and the potential successful bid for a Mitsubishi/Chrysler plant (U.S. for spin-offs and other positive externalities play key Congress, Office of Technology Assessment 1990b). roles in the long-term results. In the 1990s, the stakes escalated. South Carolina offered a $130 million incentive package in its success- ful bid in 1992 to lure a 2,000-job BMW assembly Recn~itment of Finns plant. South Carolina reportedly offered Mercedes In the 1960s and 1970s, state economic develop- Benz a similar package to that offered BMW but lost ment efforts focused on the recruitment of employers out to Alabama, which promised a record-setting and jobs, either luring existing plants to relocate or incentives package worth over $300 million. In addi- attracting new plants. Seeking to differentiate them- tion to the price, another unusual feature of the selves, states offered tax and financial incentives to package was Alabama’s agreement to pay the salaries encourage firms to relocate within their borders. A of the 1,500 workers (at an estimated $45 million) relatively low-wage work force and a good labor while they were being trained during the first year or climate--which generally meant accommodating la- so on the job (Applebome 1993; Browning and Coo- bor or no unions--were often highlighted in recruit- per 1993).

18 May/June 1994 New England Economic Review In recent years, recruitment efforts in many velop new firms have influenced private investment states have focused on attracting new plants of firms practices and filled gaps in capital markets. By the that are expanding, rather than trying to induce mid 1980s, most states had funded venture capital employers to relocate existing facilities. The trend has programs to finance new and emerging businesses. also been toward greater emphasis on international These programs, some of which require matching in~cestors, as states hope to lure plants of Japanese funds from the firms, are generally quite small. They and other foreign companies. often seek to expand or change existing lending practices in the private sector. They may support firms that might not have approached traditional High-Tech Job Creation sources of seed money, or encourage private invest- In the 1970s and early 1980s, many states began ments in potentially productive projects traditionally supplementing industrial recruiting strategies with bypassed because they were considered too risky. efforts to create jobs at home. The impetus behind These entrepreneurial venture capital programs this trend came partly from some states’ disappoint- have brought states into relatively unfamiliar territory ment with their lack of success in recruiting jobs from for public sector institutions. Traditionally, state in- outside. It was also in response to growing evidence dustrial development loan programs worked with nationally that the key to employment growth and existing firms that backed their loans with collateral. good jobs lay in "growing your own" (Grubb and In contrast, the new loan programs often focus on Stern 1988). start-up operations and new product development, The experiences of California’s Silicon Valley and for which collateral is often not reqttired (Eisinger 1988). Massachusetts’ Route 128 provided tempting exam- ples of the high-tech job creation approach. Seeking Revitalization of Established Businesses to replicate the success of these areas, many states adopted a range of high-tech development initiatives Recent years have also witnessed a shift in em- that focused on research, development, and technol- phasis in state economic development programs to- ogy transfer. ward assistance to established businesses (Ganzglass Efforts to stimulate technological and Heidkamp 1987; Osborne 1987; Rose and Kotlow- have taken a variety of forms, including research itz 1991). Efforts to help established firms in the centers, industry-university partnerships, matching United States historically have focused on the pre- grants, and research parks. Research centers, often vention of job loss or on the reemployment of work- operated in conjunction with universities, conduct ers displaced from their firms. Measures to retain jobs applied research and allow firms to pool their re- in mature or declining industries, for example, have sources for facilities and equipment. Research parks, often included import quotas, domestic content rules, which encompass concentrations of R&D firms, are restrictions on outsourcing, and protection against designed to generate the exchange of new ideas and unfair competition. hasten their transfer to the market. By the mid 1980s, At the state level, cost-reduction incentives (for approximately 150 research parks were in operation example, reductions in unemployment insurance, in the United States, almost double the number a workers’ compensation, or taxes and direct subsidies) decade earlier (Eisinger 1988). have been used in attempts to offset cost disadvan- Programs to support high-tech start-up firms tages in an area and to keep employers in the state. have also grown in recent years. All states now States have also taken an active role in seeking to operate programs to assist small businesses and most offset the adverse consequences of structural change. have programs designed to stimulate new firm for- Many states have developed worker assistance cen- mation. Traditionally, small business assistance pro- ters or emergency teams to assist with plant closings grams offered technical and managerial help; states and provide job search assistance, supplemental un- are expanding these efforts to include more entrepre- employment benefits, and assistance in moving. neurial and financial assistance. A few states have Some states have created programs to assist created small business "incubators," which provide existing firms before a shutdown becomes imminent. shared services such as legal assistance, conference Michigan’s Jobs Opportunity Bank, Delaware’s Blue rooms, accounting services, and research facilities at Collar Jobs Act, and the New Jersey Jobs Training relatively low rents to start-up firms. Program specifically target resources to retrain cur- Increasingly, state initiatives to create and de- rent workers and possibly forestall plant closings.

May/June 1994 Nero England Economic Review 19 Skills corporations, in which business and academic focus their technology program funds on university institutions work together and share training and R&D and on assisting start-up firms, rather than on retraining costs, emerged in the 1980s to assist estab- the integration of new technologies into established lished firms that were growing rapidly and facing firms. For instance, only about 10 percent of the $550 skill shortages. million spent on various kinds of technology pro- Increasingly, states have begun to take broader grams in 1988 was spent on technology transfer and measures, which include programs for moderniza- on technical and managerial assistance. As of 1990, tion and the development of new, foreign markets, in only 10 states operated programs whose primary order to bolster the competitiveness of existing firms. function was to assist manufacturers in technological Michigan’s Modernization Services Program and adoptions. A mid 1980s survey by the Office of Massachusetts’ Center for Applied Technology, for Technology Assessment (1990a) showed that only 2 example, seek to revitalize the states’ traditional percent of small and medium-sized enterprises had manufacturing sectors, such as auto parts, apparel, received industrial extension services from the state. and cutting tools. These programs assist firms in the The recent Department of Defense "build-down" integration of new technologies by identifying both and growing defense conversion efforts will bring technological and training needs and by providing greater attention and funding to industrial modern- support and technical assistance. ization activities. The federal Advanced Research In a multistate effort, the Southern Technology Projects Agency will be providing hundreds of mil- Council Consortium for Manufacturing Competitive- lions of dollars nationwide for R&D and dual use ness was established in 1988 to utilize the states’ (defense and commercial) technologies. In October vocational schools and community colleges to assist 1993, for example, Massachusetts received $10.6 mil- small and medium-sized enterprises with new tech- lion in the Clinton Administration’s first round of nologies. Some states have begun experimenting defense conversion grants. These funds will be used with programs to stimulate exports by helping small to create a statewide Manufacturing Modernization and medium-sized enterprises market their products Partnership Program to help small and medium-sized overseas. firms diversify into commercial markets. Some state-financed training programs have shifted their efforts toward retraining the potentially unemployed and upgrading the skills of current H. Technology Life Cycles, workers. California’s Employment Training Panel, Competitiveness, and Economic the nation’s largest state-financed training program, Development 1 funded at approximately $55 million a year, was originally designed to assist firms moving into the state. Life-cycle models emphasize the evolutionary It now focuses on helping existing businesses retool character of production and employment needs. The and reorganize in order to enhance productivity. "industry life cycle" concept dates back to the 1930s, A few states have begun linking their training when industries were found to undergo a sequence funds for established firms to capital investments. of stages---experimentation, rapid growth, dimin- Indiana’s Basic Industrial Training Program, for ex- ished growth, and stability or decline--as they devel- ample, requires firms in mature industries (such as oped. Separate "life cycles" have subsequently been transportation, steel, and heavy machinery) that are delineated for products, for production processes, expanding or modernizing to invest in capital equip- and for technologies. ment in order to be eligible for retraining assistance. The state covers between 10 percent and 50 percent of training costs, depending on the level of investment. Technology and Skill-Training Life Cycles Illinois’ Industrial Training Program, which added a The technology life cycle, in particular, is a mature industry component to complement the tra- valuable tool in understanding the impact of indus- ditional support of new and expanding companies, trial change on jobs and employment (Ford and Ryan also makes training contingent on capital investment 1981; Shanklin and Ryans 1984). Technologies--such by the firms. as a numerical control technology, a microelectronics While the revitalization of established businesses has taken on increasing importance, the shift in this direction is still quite limited. Most states continue to This section draws heavily upon Flynn (1991, 1993).

20 May/June 1994 New England Economic Review Table 1 Skill Training Life Cycle

Introduction: Growth: Maturity: Decline: New and Increased Demand Slower Growth in Skill Emerging Skills for Skills Demand for Skills Obsolescence Nature of Tasks Complex Increasingly routinized Increasingly routinized Narrowly defined Type of Job Firm-specific Increasingly general General: transferable General: transferable Skills Effects on Job Job enlargement: new Emergence of new Relatively rigid job Elimination of Structure positions created when occupations hierarchy; occupations occupations significant change in associated with formal skill needs occurs education and related work experience requirements Skill Training Employer or equipment Market-sensitive Schools and colleges, Declining number of Provider manufacturer schools and colleges more generally schools and colleges; some skills provided by employer Source: Adapted from Patricia M. Flynn, Technology Life Cycles and Human Resources, Lanham, MD: University Press of America, 1993, p. 19.

technology, or a data-processing technology---exhibit The firm-specific nature of skills required by the new patterns of development in which they are intro- technologies also means that employers must provide duced slowly at first, become more widely adopted as their own training or rely on equipment vendors to intensive efforts lead to do so. improved performance, and are then replaced by a As a technology becomes more widely adopted new, superior technology. and equipment standardized, skills that were once A clear understanding of the technology cycle firm-specific become general skills transferable can provide signals of impending changes in prod- among employers. Increased demand and standard- ucts and production processes. Rapid product inno- ization of skills permit their "production" on a larger vation accompanies the earliest phases of a technol- scale and at locations away from the R&D sites. As a ogy’s development, whereas process innovation result, skill development tends to shift from the peaks later in the technology’s cycle as product workplace to the formal education system as technol- design stabilizes. As a technology matures, uncer- ogies mature. Computer programming, keypunch- tainty about its capabilities and limitations declines, ing, and word processing are classic examples of this and products and processes can become more stan- transfer. dardized. in the later stages of develop- As technologies become obsolete, training fo- ment of a technology, if they occur at all, are primar- cuses on replacement needs and on the retraining of ily minor improvements in equipment rather than workers for other areas. A limited market for these major, fundamental changes in either product or skills and declining student enrollments result in the production processes. termination of school-based training programs in Just as the production processes change over the these fields. The responsibility for training to fill life cycle of a product, so do the skill and training relatively short-term, skilled replacement needs, needs of industry over the life cycle of a technology thus, shifts back to firms. (Table 1). The early stages of a technology, which are characterized by a high degree of product innovation, The Geographic Location of Jobs are relatively skill- and labor-intensive; professionals such as engineers and scientists perform most of the In addition to altering production processes and tasks later assumed by production and marketing skill needs, technology and production life cycles managers, technicians, and skilled craftsworkers. affect the geographic location of jobs. Patterns of

May/June 1994 New England Economic Review 21 regional specialization occur, as employers seek to products that incorporate technologies on the cutting locate different production activities in areas best edge and more mature consumer electronics goods, suited to their needs. Furthermore, changes in the such as radios and televisions. Firms manufacturing labor and skill requirements over a product’s life can the newer goods tend to concentrate their production trigger geographic shifts in employment over time. operations near R&D. More mature products are The "regional life cycle model" suggests that the produced in lower-cost areas. Similarly, while an attractiveness of regional and local economies varies increasing share of the world supply of semiconduc- with the skill needs of products at different stages of tors is produced outside the United States in coun- development (Rees and Stafford 1984). Early stages of tries with relatively abundant supplies of low-cost product innovation and development occur in areas labor, the design and development work is still in which highly skilled professional and technical highly concentrated in Silicon Valley. workers are available to conduct R&D. Standardiza- The computer industry shows similar patterns of tion and increasing output of the product trigger regional specialization and employment trends (Hek- reduced skill requirements, inducing production man 1980). R&D, design, and production of state-of- shifts to geographic areas characterized by lower the-art equipment continue to be geographically con- labor costs. centrated in Massachusetts and California, along with company headquarters. In contrast, the large- scale production of relatively standardized computer components and routinized assembly activities have The "’regional life cycle model" dispersed away from R&D centers, taking place in large branch plants in states with relatively low labor suggests that the attractiveness of costs (such as Tennessee, South Dakota, and North regional and local economies and South Carolina) or in low-wage countries (such varies with the skill needs as Mexico, Hong Kong, and Taiwan). of products at different stages of development. III. State Strategies and Life Cycles When viewed in the life-cycle framework, the evidence on recruitment, high-tech job creation, and Similarly, on a global level, the "international business revitalization strategies sheds new light on product cycle model" posits that firms initially locate the role of states in fostering economic development. close to the source of demand for their newly devel- oped products so they can rapidly communicate Recn~it~nent Strategies market information into product changes (Wells 1972; Vernon 1979). As foreign markets emerge for the Relocation incentives will have different effects product, they generate exports for the producing on different types of production activities. In the early country. At some point, depending on the nature of stages of product development, firms compete the products and the characteristics of foreign de- mainly via innovation and through product differen- mand, the expanded foreign market attracts its own tiation. In contrast, for firms that produce relatively production base. When production costs abroad are standardized products, competition is mainly a func- low enough to compensate for transportation and tion of cost. Incentives such as low wages and tax other costs, such as tariffs, the country that originally abatements will, therefore, be a greater inducement produced the product becomes a net importer of the to plants operating at the later stages of production good. At the final stages of product development, cycles than to firms involved primarily with R&D and production activities may shift from the sites of product entrepreneurial activities. Similarly, short-term cus- demand to lower-cost areas in other countries. tomized training programs are likely to appeal to Industries usually rely on a range of technolo- employers engaged in large-scale, mass production gies, have products in several phases of develop- processes, but be of little value to firms characterized ment, and are characterized by diverse skill needs by complex, nonstandardized activities, which re- and employment patterns. The electronics industry, quire relatively high-skilled and broadly trained for example, produces both highly sophisticated workers.

22 May/June 1994 N~o England Economic Review The life-cycle framework accentuates the need to stance, has been attributed to trained and malleable look beyond industry aggregates in fashioning re- labor, low wages, and cheap land. With respect to cruitment strategies for economic development. Most workers, in particular, the Germans are said to have industries and many, especially larger, firms encom- found "a work force willing to tolerate management pass products, processes, and technologies at various practices that Americans often find idiosyncratic, if stages of maturity. Industrywide data, therefore, not obnoxious .... Such adaptability has more than combine production activities requiring different cap- made up for the skill levels of many of the workers" ital and labor requirements, and with diverse location (McCarthy 1993). In 1993 Mercedes-Benz sought a needs. U.S. location in order to move closer to the vast In the life-cycle perspective, the concept of a high American market and avoid a 25 percent tariff on technology industry is a misnomer. "High tech" is a imported trucks. The Alabama site selected offered dynamic and relative concept that describes the ear- relatively low labor costs and a tax and incentive liest phase of development. "High-tech employ- package that will result in Mercedes paying the ment" should refer only to those jobs involved with equivalent of $100 for the site (Applebome 1993). R&D, innovation, or nonstandardized production activities--jobs that exist across a wide range of industries, including those that are relatively mature. "Low-tech" or routinized production activities (at the Few businesses move their other end of the development cycle) also are found across a variety of industries, including computers operations between states, and and electronics. very little employment growth In attempts to recruit "high-tech" employers during the late 1970s and early 1980s, many states is attributable to the migration used incentives including tax abatements and short- of jobs into a state. term customized training programs to pursue a list of "high-tech" industries. While the industries were selected on the basis of their relatively high propor- tions of R&D expenditures and of professional and While many states continue to actively recruit technical workers, the bulk of the employment in employers, a relatively small number of states can be these industries was in blue-collar and clerical jobs. expected to launch effective recruitment strategies Many states succeeded in recruiting only the rela- that contribute significantly to the number of "good" tively low-skilled, standardized manufacturing jobs jobs and to long-term economic development. Few (for example, the assembly of printed circuit boards) businesses move their operations between states, and in these industries. very little employment growth is attributable to the Earlier recruitment activities yielded similar re- migration of jobs into a state. sults, with jobs relocating from other states primarily Moreover, recruitment strategies, even those ini- in manufacturing branch plants (Malecki 1983). These tially appearing quite successful in terms of numbers jobs are more apt to involve relatively standardized of new jobs, can actually undermine long-term eco- production activities and be more vulnerable to fur- nomic growth. For instance, if tax and other financial ther dispersion to lower-cost locations than are jobs incentives have a negative impact on the quality of in firms indigenous to a geographic area. Many of the life by restricting education and services in the area, northern firms that relocated to southern states to relocation incentives could deter the entry of employ- take advantage of a low-wage work force and com- ers whose work force contains relatively high propor- pany-specific training, for instance, subsequently re- tions of professional and technical workers. In addi- located to still lower wage areas (Southern Growth tion, the recruitment of new industries and firms can Policies Board 1988; Rosenfeld 1992). backfire if, in the process, incentive packages to new The bulk of recruitment incentives used by states firms impair the competitiveness of established em- are still those (for example, tax and financial abate- ployers or prompt their "premature" departure from ments, customized training) that appeal primarily to the area. Expensive recruitment packages, for in- plants with relatively low-skilled and low-wage posi- stance, can drain resources from more traditional tions. The attractiveness of the Carolinas to German sources of employment, which comprise the bulk of firms locating plants there in recent years, for in- all jobs in local economies. Existing companies may

May/June 1994 New England Economic Review 23 also suffer if the state subsidizes the entry of firms and develop these new and emerging firms. Research that are their direct competitors. on the location of technology-based entrepreneurial The external control inherent in branch plant firms confirms these life-cycle hypotheses with re- economies, whereby major corporate decisions are gard to the importance of R&D, venture capital, and made elsewhere, suggests that local employment and skilled labor in high-tech development strategies other community concerns may not be a top priority (Malecki 1990, 1991). in discussions of firm location and restructuring. "High-tech" job creation strategies are not likely Further, given their mix of production activities and to be very effective for many states (Browne 1983; occupations, branch plants are less likely than indig- Gittell and Flynn 1994). Historically, small technolo- enous new firms to act as a "seed bed" or "growth gy-based firms, and high-tech employment more pole" in stimulating spin-offs and new employment generally, have accounted for a relatively small pro- opportunities in an area. portion of all employment. High-tech employment Recent anecdotal evidence does indicate, how- in the United States is geographically concentrated, ever, that several foreign auto assembly plants (for with most found in New England, California, and example, Toyota in Kentucky, Honda in Ohio, Nis- Texas. R&D activities, in particular, remain geographi- san in Tennessee) have attracted supplier branch cally concentrated in a few areas of the country. plants to the area. Moreover, if state recruitment strategies provide longer and more complex educa- tion and training programs than in the past, states may be able to attract better-quality jobs. More highly States with significant university skilled and more broadly trained work forces are R&D, venture capital, and highly incentives that appeal to firms in innovative, non- standardized activities in earlier stages of develop- skilled labor have the most ment. Michigan, for example, one of the top three potential for implementing a contenders for a Saturn plant in 1985, offered a recruitment package that encouraged development of successful competitive strategy "world-class" manufacturing and engineering talent. based on entrepreneurial While it lost its bid for the manufacturing plant, it won the company headquarters and R&D facilities, new firms. and the relatively high-skilled jobs that accompany these functions (Fosler 1988). States with significant university R&D, venture capital, and highly skilled labor have the most poten- High-Tech Job Creation Strategies tial for implementing a successful competitive strat- The life-cycle framework helps to clarify the role egy based on entrepreneurial new firms. In addition, of new and emerging businesses in economic devel- an established base of high-tech employment pro- opment. The creation and development of new en- vides an area with a competitive edge in the creation trepreneurial firms require strategies that focus on of new entrepreneurial firms. An existing agglomer- the characteristics and needs of products and tech- ation of firms in similar or related sectors is a princi- nologies during their early stages. pal determinant of both birth rates and the distribu- In the high-tech success stories of the Silicon tion of small technology-based firms. Concentration Valley and Route 128, growth was driven by local of these resources in one area enhances the firms’ start-ups and spin-offs from companies already in the productivity by creating external economies of scale area. The technical infrastructure of both areas en- in production and marketing. A self-sustaining "crit- compasses applied research and product develop- ical mass" of employers can develop, as the concen- ment at universities, informal local tration of entrepreneurial firms attracts additional networks, a scientific and technical labor force, and firms and venture capital, strengthens the technolog- proximity to complementary and competitive firms ical infrastructure, attracts and retains skilled profes- and to distributors and markets. These examples sionals, further promotes informal communication accentuate the importance of innovation, research, networks, and encourages innovative activities (U.S. product design, and non-routine production activi- Congress, Office of Technology Assessment 1984; ties. Venture capital can provide the means to create Malecki 1990, 1991).

24 May/June 1994 Nezo England Economic Review The flow of venture capital further highlights the Business Revitalization Strategies advantages of an established high-tech base and the presence of research universities in the formation of The life-cycle framework also sheds new light on new firms. The availability of venture capital varies strategies to revitalize traditional and established widely by state and region, with funds flowing from firms, whose activities are primarily beyond the ini- U:S. financial centers like New York and Chicago to tial stages of development. Some established firms centers of innovation and technology. California, involve "mature" production activities. Representing Massachusetts, and Texas regularly attract venture the extreme opposite of high-tech activities, mature capital, with California alone often accounting for activities are those in which technologies and prod- one-third to one-half of all U.S. venture capital. In ucts are relatively standardized, mass production contrast, many states have virtually no venture cap- predominates, skill requirements are relatively low, ital funds. and little or no innovation is taking place. Competi- While an established high-tech employment base tion is primarily a function of cost. gives an area a decided advantage in new firm formation, relatively little is known about the initial generation of local start-ups. The initial "confluence of technological opportunity," or the appearance of The potential across states the first entrepreneurs, appears to be due to the availability of start-up financing and the existence of for programs to enhance informal (noninstitutional) personal and local con- productivity and competitiveness tacts supportive of new, unproven entrepreneurs through revitalization (U.S. Congress, Office of Technology Assessment 1984). Small firms, that is, those with fewer than 100 of established businesses employees, are the major source of entrepreneurs, is extensive. although a significant number of founders do origi- nate from large firms. It is important to differentiate among small firms in fashioning a high-tech development strategy. Most Considerable diversity exists among traditional small businesses create no jobs after the first few industries in terms of their organizational structures, years and many, particularly in the service sector, occupations, wage rates, and skill requirements. generate lots of relatively low-paying, dead-end jobs Within industries and even firms, mature segments conducive neither to innovation nor to entrepreneur- often coexist with high-tech segments, as well as with ship. Relatively few small firms have the potential for activities that involve products and technologies growth and expansion and act as "seed beds" for along the mid-range of the development spectrum. future jobs. Such firms are dominated by innovative, Effective revitalization strategies for these industries nonstandardized activities. will take a variety of forms, including integration of A high-tech job development strategy will be new technologies, better utilization of mature tech- extremely difficult, if not impossible, for relatively nologies, development of specialized product niches, small areas that lack universities, existing technology- and reorganization of the workplace. based companies, and skilled labor. Areas dominated In contrast to the recruitment and high-tech job by relatively mature industrial bases and technologies creation strategies, the potential across states for are also unlikely to be able to implement an effective programs to enhance productivity and competitive- economic development strategy around technology- ness through revitalization of established businesses based entrepreneurial firms. is extensive. There are two main reasons for this. Empirical evidence confirms that most research First, the dynamics of technological and industrial parks fail (Eisinger 1988). Some are unable to attract change accentuate the ongoing need for upgrading of tenants; others fail to generate spin-offs; almost all fail human resources and facilities to maintain competi- to stimulate technology transfer. With respect to tiveness. Second, states have only just begun to tap venture capital, most state programs are quite small the opportunities available to them regarding busi- and probably will not prove effective in establishing ness modernization strategies. the "critical mass" of high-tech firms needed to The introduction of new technologies across a generate a self-sustained growth environment. variety of established industries can benefit states by

May/June 1994 New England Economic Review 25 fostering product and process innovations that lead processes and highly skilled labor also facilitate adop- to new and improved products and new markets. tion of more advanced technologies (Doeringer, States need not have high-tech firms located within Terkla, and Topakian 1987). their boundaries in order to benefit from such a Organizational and managerial changes are often strategy. While still a strong competitor in terms of necessary to fully exploit the potential productivity R&D and innovation, the United States continues to gains of new technologies and corporate restructur- fare poorly with respect to the transmission of "best ing. U.S. managers have been criticized, however, for practice" technologies throughout the industrial several shortcomings in this area: (1) failure to eval- structure. U.S. rates of adoption of robotics, comput- uate effectively both the short-term and the long- erized numerical control devices, and other advanced term costs and benefits of technological adoptions; technologies continue to fall behind those of our (2) inadequate development of human resources to industrial competitors. Moreover, even when adop- meet changing needs; (3) insufficient development of tion rates are similar, U.S. firms have been found to organizational structures that can fully exploit the be less efficient in their implementation (Osterman productivity gains associated with new technologies; 1988; Dertouzos, Lester, and Solow 1989; U.S. Con- and (4) failure to establish fruitful cooperative rela- gress, Office of Technology Assessment 1990a, 1990b). tionships with workers (Hayes and Abernathy 1980; Cyert and Mowery 1987; Drucker 1988; Hayes and Jaikumar 1988). Small firms, in particular, have difficulties with The failure of firms to remain technological adoptions because of costs, skill and retraining requirements, and the need to keep up-to- technologically competitive date. State industrial extension and training efforts, contributes more to worker however, reach relatively few small firms. State offi- cials indicate that it is hard to find small companies, displacement and job loss assess their needs, and spend enough time with them than does the adoption to make a difference. of new technologies. The fact that industrial extension programs are rarely integrated with state training efforts highlights other missed opportunities. Neither technology nor training in isolation from systemwide support will Some observers express concern that adoption of effectively increase productivity and jobs. The recent new technology causes permanent job loss. In fact, trend, albeit small, to link training with capital invest- however, the failure of firms to remain technologi- ments is a good step in promoting industrial compet- cally competitive contributes more to worker dis- itiveness. placement and job loss than does the adoption of new The shift in some state-financed training pro- technologies (U.S. Government Accounting Office grams away from recruitment and toward the more 1986; Cyert and Mowery 1987; OECD 1988). Adop- efficient use of existing state resources and firms also tion of technologies in their relatively early phases of has the potential to enhance competitiveness and development has primarily positive impacts such as long-term economic growth. However, while mod- upgrading and job enlargement. In contrast, the ernization efforts generally require flexible and more preponderance of negative impacts such as mass broadly trained workers, most state-financed training layoffs, unemployment, and job downgrading relate programs continue to provide relatively short-term to adoptions of relatively mature technologies or to training for individual firms (Creticos and Sheets the failure of firms to adapt at all. 1990). In-plant training provided by state-financed An alternative to the technology-based approach training programs has not been assessed on a sus- for enhancing the competitiveness of established tained basis; skills corporations, too, have had few firms involves a shift toward customization and mar- evaluations. Furthermore, the firms accepting public ket niches. Flexible manufacturing systems that make funds might have provided the training anyway. shorter production runs economical and encourage Matching requirements help to limit the degree of product differentiation have promoted a trend to- substitution taking place; questions remain, how- ward greater use of small-batch production of rela- ever, about the transferability of the skills being tively specialized products. More flexible production provided.

26 May/June 1994 New England Economic Review IV. Development of State Strategies State Employment Assessment The life-cycle perspective on competitive strate- ¯ How does the state’s industrial structure com- gies is useful to states for several reasons. First, states pare with the national economy? How has can use it to assess where their economies are in this been changing over time? terms of emerging, evolving, and maturing employ- ment opportunities, and thus what economic devel- ¯ How does the state’s occupational mix in its opment needs might be. Second, states can use it to major industries compare with the national guide their determination of where they might want averages in those industries? to be, the feasibility of their aspirations, and the ¯ Describe the extent of various kinds of pro- economic development issues that must be addressed duction activities located in the state (the mix to move in that direction. Third, states can use it to of branch plants, headquarters, and R&D fa- determine the relevance of the experiences of other cilities). Is there a trend in recent years? states to their own competitiveness strategies. ¯ Make a grid classifying the state’s major in- dustries and employers by development stage Tailoring Competitiveness Strategies (emerging, growing, stabilizing, declining). to Individual States ¯ What is the birth rate of new firms in the Most states will select a mix of strategies (recruit- state? How does this compare with the na- ment, job creation, retention) to promote competi- tional average? tiveness and long-term economic development. A ¯ What are the characteristics (industries, firms, state’s economic development goals should reflect its products, technologies) of the state’s high- competitive strengths and opportunities. In addition, tech employment? the selection and design of strategies and particular ¯ What are the characteristics (industries, firms, programs should be linked to the state’s employment products, technologies) of the state’s major base and resource mix. traditional employers? States will differ with respect to composition of employers, characteristics of the work force, institu- ¯ What is the extent of entrepreneurial small tional capabilities, and other resources. Goals and firms within the state? Identify potential high- strategies, therefore, are expected to vary from state growth areas. to state. In tailoring their strategies, states should ¯ What industries have been the primary assess their existing employment base, the charac- sources of plant closings, layoffs, and unem- teristics and potential of state resources, and the ployment in the state in recent years? What strengths on which they can build competitive advan- were the reasons for these events? tage. What are the needs (skills, technological, fi- Initially, states should analyze the nature and nancial) of the state’s traditional employers? mix of their employers and jobs. This analysis re- quires looking beyond industry aggregates and iden- tifying the types of production activities (for example, R&D, standardized assembly), types of employers, implement a high-technology job creation strategy or occupational requirements, and skill needs. Business to recruit good jobs? The characteristics (age distribu- revitalization strategies, in particular, further accen- tion, education levels, occupations, wages) of the tuate the importance of understanding the existing state’s labor force should be compared with national employment base. While each state is likely to iden- averages to identify state strengths or potential prob- tify additional questions relevant to its particular lems. A state with a relatively old work force, for circumstances, the first box provides guidelines for instance, will face more replacement needs than conducting this employment assessment. others. A state with relatively high proportions of States should then develop an inventory of labor engineering and technical talent can have an advan- and other resources available (educational and train- tage over others in high-tech development possibili- ing institutions, R&D facilities, venture capital) that ties. A state with relatively low production wages can can influence competitiveness efforts. Does the state attract manufacturing plant production jobs. The have the types of resources necessary to effectively overall structure of a state’s education and training

May/June 1994 New England Econo~nic Review 27 tion, potential barriers and constraints to implement- State Resource Inventory ing strategies and programs should be identified. When policy options have been identified as particu- ¯ How does the state’s work force compare with larly appropriate for the state, the experiences of national statistics regarding demographic and other states in that regard may then prove particu- educational factors? What are the implications larly useful. What were the impacts of those pro- in terms of education and training needs? grams elsewhere, and what problems were encoun- ¯ What are the major R&D institutions in the tered? The third box provides guidelines for thinking state? strategically about the state’s economic development ¯ What are the extent and sources of venture policies and employment and work force needs and capital available to new firms? opportunities. ¯ Describe the "business culture," labor cli- mate, and status of labor relations in the state. "’Defensive" and "’Proactive’" State Actions Give examples. The life-cycle framework highlights the impor- ¯ What major skill shortages and surpluses tance of distinguishing between "defensive" and have occurred in recent years? How were "proactive" actions in seeking to bolster a state’s these imbalances resolved? competitive advantage and long-term economic de- Describe the evolution and current status of velopment. Defensive actions represent an expedient the state’s education and training network. way of improving competitive position by lowering What are the strengths and weaknesses of the costs. They do not, however, address issues of work various institutional components of this net- force quality and technological change that underlie work? business performance. In contrast, proactive or inno- vative adjustment mechanisms can lower costs by Which firms have used state-financed training increasing labor productivity, motivating workers, programs? Describe the extent and types of improving efficiency, and increasing the quality of skills provided. the work force (National Center on Education and the What relationships/partnerships exist be- Economy 1990; Doeringer and others 1991). tween education and training institutions and Classifying state actions as defensive or proactive employers (for example, co-op programs, ap- can be useful in understanding the impacts and prenticeships, advisory boards)? Have these trade-offs, both short-term and long-term, of various met expectations? policy options. Defensive state actions such as tax abatements or other financial incentives can quickly lower costs to potential employers and perhaps at- network should be identified. Further, the roles and tract relatively large numbers of jobs to some states in track records of the institutional components of the a short period of time. As discussed above, however, education and training network should be assessed in these mechanisms may undermine long-term eco- terms of skill generation and responsiveness to nomic growth, as the jobs recruited are often rela- changing labor market needs, in order to understand tively low-skilled and vulnerable to further relocation the capabilities of the system. The second box pro- to even lower-cost areas. Proactive strategies may vides guidelines for the development and assessment increase costs initially and will take longer to reduce of the state’s resource inventory. costs via productivity increases. However, the ulti- Lastly, competitiveness strategies and programs mate impacts on jobs and growth are likely to be should be assessed in light of the state’s employment more positive and longer-lasting. and resource bases. In which activities is state policy The defensive/proactive dichotomy highlights likely to be most effective in generating good jobs and the importance of having public policies focus on long-term economic development? In which indus- "good jobs" as opposed to "jobs" per se. Moreover, tries? In which types of firms? Assessments should be "output" should be viewed in addition to jobs in made of various recruitment, job creation, and busi- evaluating policy effectiveness, particularly with re- ness revitalization programs previously implemented spect to relatively mature industries where increasing in the state. Such assessments should include both competitiveness and long-term viability are often the short-term and the long-term impacts. In addi- achieved with lower employment levels.

28 May/June 1994 New England Economic Review instance, is indicative of this shift away from a pure Strategic Thinking about Economic cost orientation to one that emphasizes productivity Development Policy and Employment and technological competitiveness. The policy op- and Work Force Needs tions being used within these broader strategies have been evolving in a similar direction. More complex ¯ What are the areas in which the state has recruitment packages that include training grants for particular strengths, in light of the employ- upgrading and for relatively skilled positions, for ment and resource inventory assessments? example, can reduce labor costs through productivity ¯ What firms have moved into the area in recent gains~in contrast to tax abatements and other finan- years? Did they relocate from another state (if cial incentives. so, which)? Are they foreign-owned? What With respect to business revitalization, while are their major production activities and the efforts are still limited, states are experimenting with nature and extent of their jobs? a range of options with the potential to enhance ¯ What incentives have been used by the state productivity at the workplace. These include helping in recruiting firms? Did those firms that have older firms adopt new technologies or make more moved in take advantages of these? effective use of traditional technologies, and helping them develop new markets by customizing or export- ¯ To what extent have new, high-tech firms ing their products. This shift toward more proactive been created in the state in recent years? In approaches promises more highly skilled jobs. Proac- what fields? What was the source of venture tive approaches also should provide real cost savings capital? over time, whereas defensive ones threaten to be- ¯ What are examples of traditional industries come increasingly expensive. With respect to recruit- and firms in the state that have modernized ment strategies, for example, when the first few their workplaces in recent years? Were state- states began offering tax abatements and customized financed training programs involved? Were training, these incentives helped to differentiate one any education and training institutions di- state from another as they sought to attract new rectly involved? employers. Over time, more and more states have ¯ Has the state been able to leverage funds to found it necessary to follow suit or risk not being provide for training? To what extent? With considered a serious contender. Now virtually all which employers? states offer tax and financial incentives and custom- ized training, so states are incorporating additional ¯ What types of coordination and cooperation features into recruitment packages in order to distin- of education and training institutions appear guish themselves from the others. necessary to implement the programs that Proactive approaches have a further advantage: appear to meet best the state’s current and At the national level the likelihood is greater of real future employment and training needs? net employment gains, rather than just a reshuffling ¯ What barriers and constraints may inhibit the of jobs among states. Moreover, proactive ap- implementation of strategies and programs proaches have the potential to lead the way to an that appear to meet best the needs of the economic development outcome with relatively high state? wages, high skills, and high living standards, effec- tively bypassing low-wage, low-skill alternatives.

Note: The work reported herein was supported under the Educa- In recent years, state economic development tion Research and Development Center program, agreement num- strategies have begun to focus more on proactive ber R117Q00011-91, CFDA 84.117Q, as administered by the Office options and less on defensive responses. The trend of Educational Research and Improvement, U.S. Department of Education. The findings and opinions expressed in this report do away from an almost exclusive focus on recruitment not reflect the position or policies of the Office of Educational toward job creation and business revitalization, for Research and Improvement or the U.S. Department of Education.

May/June 1994 New England Economic Review 29 Applebome, Peter. 1993. "States Raise the Stakes in Fight for Business Reviao, vol. 66, no. 5 (September/October), pp. 77-85. Jobs." The New York Times, October 4. Hekman, John S. 1980. "The Future of High Technology Industry Browne, Lynn E. 1983. "Can High Tech Save the Great Lake in New England: A Case Study of Computers." New England States?" New England Economic Review, November/December, pp. Economic Review, January/February, pp. 5-17. 19-33. Malecki, Edward J. 1983. "Technology and Regional Development: Browning, E.S. and Helene Cooper. 1993. "States’ Bidding War A Survey." International Regional Science Reviao, vol. 8, no. 2, pp. Over Mercedes Plant Made for Costly Chase." The Wall Street 89-125. Journal, November 24. ¯ 1990. "New Firm Formation in the USA: Corporate Struc- Creticos, Peter A. and Robert G. Sheets. 1990. Evaluating State- ture, Venture Capital, and Local Environment." Entrepreneurship Financed Workplace Based Retraining Programs: A Report on the and Regional Development, vol. 2, pp. 245-65. Feasibility of a Business Screening and Performance Outcome Evalua- ¯ 1991. Technology and Economic Development: The Dynamics tion System. Washington, DC: The National Governors’ Associ- of Local, Regional and National Change. New York: John Wiley and ation, Research Report 89-08. Sons, Inc. Cyert, Richard M. and David C. Mowery, eds. 1987. Technology and McCarthy, Michael. 1993. "Why German Firms Choose the Caro- Employment. Washington, DC: National Academy Press. linas to Build U.S. Plants." The Wall Street Journal, May 4. Dertouzos, Michael L., Richard K. Lester, and Robert M. Solow. National Center on Education and the Economy. 1990. America’s 1989. Made in America. Cambridge, MA: The MIT Press. Choice: High Skills or Low Wages. New York: National Center on Doeringer, Peter B. and others¯ 1991. Turbulence in the American Education and the Economy. Workplace. New York: Oxford University Press¯ Organisation for Economic Cooperation and Development Doeringer, Peter B., David G. Terkla, and Gregory C. Topakian. (OECD). 1988. Industrial Revival Through Technology. Paris: OECD. 1987. Invisible Factors in Local Economic Development. New York: Osborne, David. 1987. Economic Competitiveness: The States Take the Oxford University Press¯ Lead. Washington, DC: The Economic Policy Institute. Drucker, Peter F. 1988. "The Coming of the New Organization." Osterman, Paul. 1988. Employment Futures. Cambridge, MA: The Harvard Business Review, vol. 66, no. 1 (January/February), pp. MIT Press. 45-53. Rees, John and Howard Stafford. 1984. "High Technology Location Eisinger, Peter K. 1988. The Rise of the Entrepreneurial State. Madi- and Regional Development: The Theoretical Base." In U.S. son, WI: The University of Wisconsin Press¯ Congress, Office of Technology Assessment, Technology, Innova- Flynn, Patricia M. 1991. "The Life Cycle Model for Managing tion and Regional Economic Development. Washington, DC: Gov- Technological Change." In Peter B. Doeringer and others, Tur- ernment Printing Office. July. bulence in the American Workplace, pp. 105-123. New York: Oxford Rose, Robert L. and Alex Kotlowitz. 1991. "Once the ’Rust Belt’ University Press. Heartland Fares Better than Coastal States." The Wall Street ¯ 1993. Technology Life Cycles and Human Resources. Lanham, Journal, July 30. MD: University Press of America¯ (Originally published as Rosenfeld, Stuart A. 1992. Competitive Manufacturing. New Bruns- Facilitating Technological Change: The Human Resource Challenge. wick, NJ: Center for Urban Policy Research, Rutgers, The State Cambridge, MA: Ballinger Publishing Company, 1988.) University of New Jersey. Ford, David and Chris Ryan. 1981. "Taking Technology to Mar- Shanklin, William L. and John K. Ryans. 1984. Marketing High ket." Harvard Business Review, vol. 59, no. 2 (March/April), pp. Technology. Lexington, MA: Lexington Books. 117-26. Southern Growth Policies Board. 1988. Halflaay Ho~ne and a Long Fosler, R. Scott. 1988. The New Economic Role of American States. Way to Go. Research Triangle Park, NC: SGPB. New York: Oxford University Press. U.S. Congress, Office of Technology Assessment. 1984. Technology, Ganzglass, Evelyn and Maria Heidkamp. 1987. State Strategies to hmovation and Regional Economic Development. Washington, DC: Train a Competitive Workforce: The Emerging Role of State-Funded Job Government Printing Office, July. Training Programs. Washington, DC: The National Governors’ . 1990a. Making Things Better¯ Washington, D.C.: Govern- Association. ment Printing Office, February. Gittell, Ross J. and Patricia M. Flynn. 1994. Looking Forward: ¯ 1990b. Worker Training: Competing in the New International Massachusetts" Business Climate in Comparative and Historical Per- Economy¯ Washington, DC: Government Printing Office, Sep- spective. Boston, MA: Massachusetts Industrial Finance Agency. tember. Grubb, W. Norton and David Stern. 1988. "Separating the Wheat U.S. Government Accounting Office. 1986. Dislocated Workers: from the Chaff: The Role of Vocational Education in Economic Extent of Business Closures, Layoffs, and the Public and Private Development." Berkeley, CA: National Center for Research in Response. Washington, DC: Government Printing Office. Vocational Education, June. Vernon, Raymond. 1979. "The Product Cycle Hypothesis in a New Hayes, Robert H. and William J. Abernathy. 1980. "Managing Our International Environment." Oxford Bulletin of Economics and Way to Economic Decline." Harvard Business Review, vol. 58, no. Statistics, vol. 41, no. 4, pp. 255-67. 4 (July/August), pp. 67-77. Wells, Louis T., Jr. 1972. "International Trade: The Product Life Hayes, Robert H. and Ramchandran Jaikumar. 1988. "Manufactur- Cycle Approach." In Louis T. Wells, Jr., ed., The Product Life Cycle ing’s Crisis: New Technologies, Obsolete Organizations." Harvard and International Trade. Cambridge, MA: Harvard University Press.

30 May/June 1994 New England Econo~nic Review