SHIFTING SECTORS
The flurry of southern and suburban migration of businesses masqueraded plant closing during the 1970s. The tendency of mature profit cycle companies sectors to disperse to lower-cost geographies, however, was again “discovered” during the seventies, mostly by Policy World economists. Loss of jobs, plant closedowns were quickly linked with several not-so-new emerging agglomerations that made their appearance in the Policy World debate that raged during the seventies. Called the Great Reindustrialization Debate, it mostly included the (business and social science) Policy World, unions, and corporate management. Political leaders also kept a watchful eye.
The seeds for many of the paradigmatic strategies that characterize twenty-firstcentury contemporary economic development were being sowed in these years. We will discuss the Great Reindustrialization Debate in more detail in Chapter 19. In this section the history isolates two sectors as deserving special attention. These sectors served as the backdrop to much of the Great Reindustrialization debate. First, a “new” southern/Midwestern auto alley incrementally developed; second, a “technology” industry, distinct from manufacturing, appeared almost out of nowhere.
The Auto Alley was fueled by Japanese and European foreign direct investment (FDI). The emergence of the Auto Alley disrupted the already troubled traditional American auto industry, and it injected foreign competition into American domestic auto production—a competition that badly hurt American industry. New auto agglomerations in the South soon became a significant element of regional change and added to the perception the South was “stealing” much-needed manufacturing from Big City states. The shift of auto-related production to the southern alley, combined with the weakness of American auto manufacturers, was, of course, a major factor in the catastrophic decline of Great Lakes and Michigan’s jurisdictional economic bases— creating the infamous “Rust Bowl” (Klier and Rubenstein, 2013).
The southern Auto Alley also became linked to Big Three (American) downsizing due to lost sales to foreign competition. Detroit simply could not produce competitive high-quality vehicles attractive to consumers. That corporate weakness triggered new discussion within auto manufacturing that permeated university business management departments—spreading its application across many other sectors as well. Productivity, corporate management strategy, and worker management joined entered into the Great Reindustrialization Debate as well.
Auto Alley and Regional Change
Sunbelt auto alley expansion was driven by investment decisions made by foreign automakers.13 Internal auto industry dynamics, microeconomic costs and production efficiencies—and corporate strategy—were incorporated into foreign automakers’ location decisions. Consequently, there were many factors that entered into a location decision—not just the ED deal-jockeying described below. Foreign manufacturers arrived at America’s shores not because of recruitment or ED attraction, but for their own reasons. Japanese auto manufacturers and parts suppliers had embraced new techniques such as just-in-time (JIT), new process technologies (statistical process controls-batch processing) and logistical/distribution innovations—which seriously affected their search for American locational assets.
Japanese manufacturers, long protected by high tariffs that shielded them from foreign competition, were confronted with the 1973 expiration of the Bretton Woods trade agreements (upon which postwar global trade revolved). From that point on “floating” currency rates became the key single factor driving global trade and comparative advantage as import/export ceased its reliance on the gold standard. Japanese auto manufacturers could no longer compete behind high tariffs, so they “innovated” and developed a new business model based on low costs, cutting-edge processes, composites materials, production efficiencies and decentralized management/team-decision-making. Floating currency rates drove their decision to establish production facilities in high-demand markets such as the USA, but they required a low-cost business climate to accommodate their business model (Howes, 1993, pp. 61–2). Low-cost American locations had a compelling advantage in attracting these facilities.
Our tale of auto alley expansion starts with Honda’s 1977 first USA investment in Marysville Ohio—not usually thought of as a Sunbelt state. The Marysville plant was initially intended to assemble 60,000 motorcycles; but in 1980 Honda purchased adjacent land, and by 1982 an automobile assembly plant was in production also. By 1990 Honda sold 800,000 Accords in the USA—half were from Marysville. The 1982 assembly plant was followed by a series of additional Honda investments in Ohio (all with tax and other incentives), so that by 1990 Honda had invested over $2 billion and hired over 8000 worker-associates in Ohio automotive-related facilities (Marvel and Shkurti, 1993, pp. 52–3).
The traditional Great Lakes auto alley evolved during the 1920s from a highly agglomerated, oligopolistic automotive complex (Markusen, 1985, pp. 163–75) around Detroit, its northern suburban counties, across the bridge to Windsor Canada to Toronto and following Interstate Route 90 through Cleveland, Buffalo and all the way to Syracuse and Schenectady NY. The Big Three subsequently decentralized automotive assembly/parts production to regional hubs in key market areas: San Francisco, Los Angeles, Dallas, Atlanta, St. Louis, Kansas City, New York/New Jersey/Baltimore and Springfield Massachusetts (Markusen, 1985, p. 170). The core “traditional auto alley” was less OEM assembly plants than a huge automobile parts and automobile-related plant nexus that remained heavily concentrated within the Great Lakes auto complex. The Marysville investment was (just) outside the “traditional” auto alley (Rubenstein, 1992). The Marysville Honda facility marked the abrupt and remarkably rapid inauguration of a second auto alley.
Prompted not only by internal industry transformation but also by threats of American protectionism and unfavorable currency imbalances, Japanese automakers (including one Subaru facility) quickly poured new investment into the south central states. The new auto alley concentrated along Route I-65 (Indiana through Alabama) and I-75 (Michigan to Georgia) with an East–West dimension: a series or “rungs” along I-20, I-40, I-64, I-70 and I-80. By 1979, the US had 55 assembly plants, 34 in the (new) auto alley.
The seven southern states of Alabama, Georgia, Kentucky, Mississippi, North Carolina, South Carolina and Tennessee together had 7 percent of transportation sector employment in 1972. Thirty years later the region’s share had grown to 16 percent … The number of assembly plants in the South increased from 5 to 13 between 1979 and 2008. In addition, 67 percent of all parts plants in the South were opened between 1980 and 2006, compared with only 40 percent in the rest of the United States. (Klier and Rubenstein, 2008, p. 3)
By 2008, the number of assembly plants in the new auto alley had increased to 43; elsewhere their number declined to seven. During the 1950s, three-quarters of all parts were made in, or near, Michigan, whereas in 2008 the state retained only one-quarter. The new auto alley even after a three-decade run shows no sign of abating (Platzer and Harrison, 2009, p. 5, Table 2).
The 1980s’ Auto Alley expansion was viewed by many as one more example of the “usual” southern piracy through incentives, promotion, a cheap nonunion workforce, community college business assistance programs and business climate advantages (right to work, low taxes, pro-business). Attraction to the Auto Alley by foreign investors certainly was affected by these factors, especially right to work. The aggressive role of southern governors in promotion also frequently created an edge. Governor Lamar Alexander’s Tennessee was a prime example of a state attraction effort that successfully landed in 1986 a General Motors-Saturn plant facility at Spring Hill. That decision, however generated intense competition between states: New York allegedly put a $1.2 billion package into play; 18 states competed. That Asian automakers/parts plants intended to be nonunion from the get-go provided a prohibitive advantage to right to work states.
Lest we appear as an apologist for southern behavior, I argue, as does Stuart Rosenfeld, that foreign automakers were only one of many industry sectors investing in the South at this time (Rosenfeld, 1992, p. 70, Tables 6–8). FDI was a cornerstone of the post-1960 southern manufacturing renaissance. Great Britain (not a competitor in Auto Alley), for example, in 1989 owned most southern foreign-owned facilities; the Federal Republic of Germany was second. Canada and France combined owned more southern manufacturing facilities than third place Japan—the chief Auto Alley foreign investor. During the 1980s Japanese-owned plants in Southern Growth Policies Board States increased from 63 to 359, an annual compound growth rate of 21 percent.
LONG LIVE TECHNOLOGY
Technology firm location was grossly uneven (in fact political economists often referred to “uneven development” in their commentary). Almost all were concentrated in a few jurisdictional agglomerations. On the other hand, the shedding of mature manufacturing firms was widespread and seemingly on the rise. New technology agglomerations created Route 128 and Silicon Valley (and smaller agglomerations in Texas, Los Angeles and Minnesota) that figured prominently in the next half-century. Technology’s rise was strongly tied to federal government war production that had evolved over many years but now caught the attention of the media and policy-makers. Agglomerations and uneven development also entered into the Great Reindustrialization Debate, that two decades later would emerge as clusters and the knowledge-based-innovation paradigm.
Early on the “technology revolution” diffused to three principal geographies: Route 128, Silicon Valley and Dallas. A second wave based on new technologies brought it to several other locations such as Seattle/Redmond. By the mid-1980s high technology had become the “economic Holy Grail … hailed as the antidote to the decline in the smokestack industries” (Markusen et al., 1986, p. 1). The original home of this incredible gazelle sector was Boston/Cambridge, nearly a half-century earlier. The rise of the technology in Boston/Cambridge was no whim or accident.
There are plenty of excellent works that detail the technology revolution.14 Our task is to understand the timing of Route 128 and Silicon Valley growth—and why. Technology has a longer history than many realize, and that should be understood to better appreciate the process underlying innovation and clusters. Our version of technology’s rise emphasizes the role played by the federal government, driven by threat of or actual war (World War II, Korea followed by the Cold War). Sectors and industries can emerge without federal government involvement, of course, but the rise of the technology that appeared in the seventies is closely tied to Defense Department/ Space Program demand and investment. Also, the period’s technology agglomeration was strongly affected by idiosyncratic roles of entrepreneurs and differing business cultures/practices of each agglomeration. These are likely to be essential ingredients for new emerging sectors cluster or agglomeration (cluster and agglomeration are not identical, by the way).
Route 128
Route 128, more correctly pre-Route 128 Harvard-MIT, was technology’s birthplace. Boston-Harvard-MIT grasped an initial advantage over other regions by dint of its well-established university relationships with the Defense Department and Washington DC. These relationships developed over two centuries and they crystallized around technology’s rise immediately previous to WWII. WWII holds the key to our understanding of pre-Route 128 new emerging sectors.
Harvard was the nation’s prime intellectual leader and “generator” of governmental/ corporate elites. By the twentieth century Harvard was Washington DC on the Charles River. War/defense policy-makers had more links to Harvard, it seems, than to West Point or Annapolis. The Harvard Business School played a similar role for America’s top corporate leadership. Harvard trained the leadership that sent government contracts to the Boston area and founded new industry. Harvard had been doing this for a long time; Edwin Land, for example, had spun off Polaroid (1937) from Harvard. Harvard’s other key sector-building contribution was to create the nation’s first venture capital firm (1946). With capital from the local insurance industry and the university, Georges Doriot and former MIT president Harvey Compton formed the American Research and Development Corporation, New England’s (and probably the nation’s) first formal venture capital firm.
MIT had unique advantages also. Founded in 1861 to pursue “its uniquely practical focus … the orientation toward the needs and interest of the industrial community” (Lampe, 1988, p. 3) rendered MIT a university whose mission dovetailed with private innovation and applied commercial R&D. MIT also had a longstanding history of spinning off companies—as early as 1889 (Arthur D. Little) and Edison’s Bell Telephone (1877). MIT proved to be the single most important launching pad for Route 128 firms, claiming that during the 1950s and 1960s it spun off over 100 startups from its research labs. Lincoln Labs, for example, spun off Data General, Raytheon and Digital Equipment. By 1986 MIT’s accelerators had spun off 400 firms since 1950, with sales in excess of $29 billion and employment of 175,000 (Lampe, 1988, p. 12).
Interestingly, neither Harvard nor MIT wanted research on the campus itself. Each set up separate off-campus research laboratories. Nor did the universities directly finance or invest in research. MIT (1955) concluded that “investing in start-up companies was too risky and not consistent with how ‘men of prudence, discretion, and intelligence manage their own affairs’ … [So] In spite of the university’s commitment to commercially relevant research, it kept firms at arm’s length” (Lampe, 1988, p. 4). Accordingly, MIT spun off-campus Lincoln Laboratory for Air Force-related research, followed by MITRE Corp. and Draper Laboratories—university accelerators in today’s parlance. Given that Route 128’s first span opened only in 1951, the separation of the university physically from the spin-off laboratory accelerators was a defining characteristic of Route 128 experience.
Central to this spinoff phenomenon is a “platform” company.” University research labs produced first-generation platform companies that spun off second-generation companies using the motherships’ technologies and processes. Platform companies spin off firms like a popcorn-maker pops corn. A platform product can potentially be used in a number of industries/sectors, not all of which actually exist when the product is first introduced. The organizational model that developed from Route 128’s platform firms’ business culture retained control over second-generation spin offs, keeping them within its corporate structure as a subsidiary or “department.” This transformed the firstgeneration company into a vertically integrated “holding corporation”—a technology conglomerate (Saxenian, 1996). Those second-generation firms outside the conglomerate found it difficult to acquire scale-up financing other than from their corporate parent—which imposed its products and technology as a condition of financing on the second-generation firm. Entrepreneurs within technology behemoths may have had little recourse than to leave the area to start a firm outside the conglomerate structure. Silicon Valley platform firms followed a different model with starkly different results—including “birthing” new sectors adjacent to the platform company.
The critical fuel behind Route 128, the Silicon Valley and Texas Instruments growth was federal research contracts and a world war. Federal contracts over decades paid for applied research to develop and construct specific prototype models (radar systems, for instance)—not basic research. Basic research had been developed decades previous (in some cases four decades previous). Like a fine wine, basic research must be carefully aged until a specific product is needed for which someone is willing to pay. Federal spending for basic research will likely employ your grandchildren. Researchers, previous to 1940, had already developed important breakthroughs and concepts, some equipment and specialized skills/processes and knowledge-mentoring relationships. In the 1920s, for instance, MIT researchers started work on the minicomputer, and in 1953 constructed Whirlwind, the world’s first reliable real-time electronic digital computer. It was waiting to be used when a market developed.
Design was one thing, production was another. Commercialization and scale-up production required yet another infrastructure to be readily available. The 65 colleges and universities in the Boston area “provided a critical source of professional labor, including physicians, managers, and lawyers as well as engineers and scientists—the labor force and middle management needed to grow production from the innovation. The Boston area could provide both. The results of all this were pretty dramatic stuff: “By the late 1960s high technology has taken firm root in Massachusetts, accounting for nearly 10 percent of total employment” (Lampe, 1988, pp. 5–6).
Route 128 itself is a limited-access, circumferential highway that opened in 1951, allegedly the nation’s first. As early as 1955, Business Week referred to the highway in an article, “New England Highway Upsets Old Way of Life,” and called it the “Magic Semicircle.” By 1958, 99 firms and 17,000 workers had located along Route 128; by 1961 there were 169 firms employing 24,000 directly on the highway, and as many very close by. In 1965 MIT researchers counted 574 companies, and the number more than doubled in eight years. In 1972 there were 1212 firms (Lampe, 1988, p. 16). By the 1970s the “Massachusetts miracle” had created the nation’s leading center of electronics innovation. The reader might remember this happened while New England’s textile industry was collapsing.
Sustained Innovation: A Second Wind?
Military contracts to the region, however, fell precipitously between 1967 and 1972— about 40 percent decline in real terms. The state lost 112,000 manufacturing jobs, 15 percent from high-tech firms. Close to 30,000 defense-related jobs were lost between 1970 and 1972 (Lampe, 1988, p. 17). Route 128 firms forced to wean themselves away from federal contracts shifted into production/sales to commercial consumer markets. The key to 128’s transition was the minicomputer. A second Massachusetts miracle began in the 1970s as advances in engineering design, discovered during the sixties, reduced computer size and cost while expanding capabilities Advances opened up new markets and innovated commercial applications, creating demand for products Route 128 firms could supply. DEC, Wang, Raytheon and Honeywell scrambled to take advantage.
DEC (a 1957 spin off from Lincoln Lab) was the platform company. Its entrepreneur, Kenneth Olsen, had developed a microcomputer Spun off in 1957 from Lincoln Lab, Olsen located DEC in an abandoned textile mill (not on Route 128). First on the market, DEC’s sales exploded. Fortune magazine (1986) declared Olsen America’s most successful entrepreneur. DEC could have spun off lots of entrepreneurs and small firms, and to a certain extent it did. Data General spun off, becoming a DEC competitor. So did Silicon Valley’s Hewlett-Packard. But for the most part, DEC retained innovations within its corporate structure, became more rigid and, for various reasons, less responsive to consumer demand. DEC tried to shape, if not compel, consumer demand to suit its research designs and products. That didn’t work— consumers wanted what consumers wanted. Data General, using DEC technology, was the first to fail and was sold to EMC in 1998. The writing was on the wall. In 1998 DEC also was acquired by Compaq (a Texas Instruments spin off whose chief competitor in the PC market was another Texas firm, Dell). In 2002 Compaq was acquired by Hewlett-Packard. The minicomputer had lost the battle to the PC. Technology had moved too fast for DEC. Massachusetts had lost its firms to Texas and California.
Silicon Valley
As late as the 1940s Silicon Valley was an agricultural region. During WWII military bases triggered in-migration and provided federal contracts to Santa Clara Valley firms. Silicon Valley developed from spin offs generated by Stanford University. But its real founder was Professor Frederick Terman (1924 MIT electrical engineering). Terman, whose father was a Stanford professor, returned home after graduating from MIT in 1925. By 1940 Terman was a national leader in electronic/radio engineering (vacuum tubes). An energetic business developer, he did whatever was needed to encourage startups, including a small tech firm pioneered by two of his students—a Mr. Hewlett and a Mr. Packard. Terman personally lent H&P their startup financing, secured a bank loan for them and found them a shop.
Other Terman-inspired spinoffs located around the Stanford area, creating a budding cluster of prewar technology firms. One was Litton Engineering Laboratory, found in 1932 to produce glass vacuum tubes. By war’s end Litton was the nation’s leading producer of glass-making machinery. Another was Varian Associates (lent $100 by Stanford and given free use of Stanford labs in exchange for 50 percent interest in any resulting patents). During World War II Terman relocated to Harvard’s Radio Research Laboratory to work on radar-jamming/anti-aircraft defense technologies. Returning to Stanford in 1946 he was committed to developing a strong West Coast technology industry capable of competing with the East. To accomplish this goal Terman recruited top-notch technical scholars, in effect making Stanford the West Coast’s MIT. By 1950 Stanford awarded as many doctorates in electrical engineering as MIT. Terman used his East Coast contacts and experience to leverage Korean War defense contracts, founding the Stanford Research Institute in 1950 (defense-related research) and establishing the Honors Cooperative Program (engineers took graduate courses at their workplace)—by 1961, 400 engineers were enrolled (Saxenian, 1996, p. 22).
Terman became Dean of the Electrical Engineering School in 1951 and, true to form, developed it into one of the nation’s best. Terman’s Stanford Industrial Park (1951)—a short walk from Stanford classrooms—recruited its first tenant, Varian’s R&D unit. A slew of corporations followed. Another classic Terman deal was Lockheed. Terman convinced Lockheed to locate its new missile and space division in nearby Sunnyvale (1957). Stanford agreed to provide faculty members to advise and train employees, while Lockheed in turn was instrumental in rebuilding Stanford’s aeronautical engineering department (Saxenian, 1996, p. 24). In 1955 the park encompassed 220 acres; in 1961 652 acres (25 companies and 11,000 employees) (Saxenian, 1996, p. 22).
East Coast firms, including Raytheon, established branches in Silicon Valley; so did the forerunner of NASA. Diversity created a critical mass of firms, engineering talent, suppliers and startups in areas such as lasers, microwaves and medical instrumentation. By the late 1960s, Santa Clara County was recognized as a center of aerospace and electronics activity. In comparison, Route 128 was more specialized. SV attracted several platform technologies spread across a number of sectors and industries—tied mostly to consumer-driven products, never totally dependent on the Defense Department.
Silicon Valley’s most explosive growth was driven by an innovation that did not exist until 1951: the semiconductor, whose chief element gave the place its name in the early 1970s. By 1970 semiconductor manufacture was the largest and most dynamic sector of Santa Clara County, and the nation’s leading center of semiconductor innovation. The industry started in 1955 in Palo Alto with Shockley Transistor, SV’s equivalent of Ken Olsen. William Shockley, an MIT grad and inventor of the transistor, worked for decades at Bell Labs in NYC and New Jersey. In 1951 he was elected to the National Academy of Sciences. Having failed in a startup transistor firm with Raytheon (1956), he returned to Palo Alto (to care for his ailing mother) and started Shockley Transistor—the first firm to develop a silicon chip.
Shockley hired the best engineering talent; but, similar to the Caine Mutiny’s Captain Queeg, Shockley inspired a mutiny and most of his staff, the “traitorous eight,” left to form a competing venture, Fairchild Semiconductor. While still in a garage, Fairchild Semiconductor got an order for 100 mesa silicon transistors, followed by Air Force and NASA contracts. By 1963 sales were $130 million. Fairchild spawned ten spin-offs in its first eight years, one of which was the Intel Corporation (1961). Thirty-one semiconductor firms started in SV during the 1960s, the majority tracing their lineage to Fairchild. Only five of the 45 independent semiconductor firms started in the US between 1959 and 1976 were located outside of SV (Saxenian, 1996, pp. 25–7).
From its inception, Silicon Valley was diversified, financed less by government contracts than venture capital (by 1974 more than 150 venture capitalists operated there) and aggressively competed in consumer-driven markets. By 1975 Silicon Valley surpassed Route 128’s employment level. Interestingly, the period in which Route 128 and Silicon Valley took off was the fifties and sixties; but it is in the seventies—when both underwent transition—that they finally entered into our ED history.
Ron W. Coan 