In the 1970's both the Route 128 complex of Boston and the Silicon
Valley were centers of high technology industry, but by the 1980's
the Route 128 area was stagnating while the Silicon Valley, after
experiencing economics shocks, was moving ahead to become the unchallenged
global leader in high technology. The difference in the two areas was not
in resources or location but in their commercial culture. Route 128 firms
tended to be insular and proprietary, whereas the Silicon Valley firms were
open and linked by social and economic networks which enabled them to adjust
to the vissitudes of market shifts. Route 128 was more highly dependent upon
federal government contracts than Silicon Valley and, when government
contracts declined with the moderation and end of the Cold War, Route
128 found it hard to adapt to the civilian market.
The Origin of the High Technology Centers
During World War II the Federal sought the development of high technology
weaponry at top universities such as the Massachusetts Institute of Technology
(MIT), Stanford University and the University of California at Berkeley.
There was a desire on the part of both the Federal Government and these
universities to continue the relationship. In some cases the research
during the war was carried out in research organizations set up organizationally
within the universities but physically separated from the campuses for
security reasons; e.g., Lincoln Laboratory at MIT. After the War these
research laboratories were made more independent of their universities
and began to function as businesses. They were however businesses that
were primarily dependent upon government contracts.
The Boston area had a long tradition of technology and the new research laboratories had to fit into the business environment of the area. Santa Clara County was still primarily agricultural at the end of the war and was not constrained by the institutional arrangements of business the way the new businesses were in the Boston area.
Some small comments on the history of the two areas are appropriate at this point. Ms Saxenian mistakenly asserts that the Silcon Valley area was most famous for "its apricot and walnut orchards," rather than its orchards of French plums which were used to make prunes.
The topgraphy of Boston with its river channels and bays make local travel difficult. In additional to the natural difficulties of the terrain Boston has a terrible, hap-hazard street pattern rather than anything approaching a rectangular grid. It is alleged that this street pattern arose because the early residents of Boston paved the meandering cowpaths to make roads. Therefore it was a major undertaking to travel from one side of the metropolitan area to the other. To remedy this situation the traffic planners decided to build a peripheral roadway, called Route 128, that would allow travelers to skirt the dense, difficult traffic conditions of Boston. Businesses quickly realized the advantages of locating close to Route 128. New businesses, particularly the the high technology companies, chose location near Route 128. Branches of major corporations such as Sylvania and RCA were located on Route 128. But startup companies became the most significant factor in the economy of the area. Ratheon rose to prominence in the Route 128 area. By 1970 Route 128 was the major center of electronics development in the U.S., but most of that development was financed by Federal Government contracts.
In the 1970's, as the space race and the Vietnam War wound to a close, military contracts for the Route 128 area dropped significantly. This decline in contracts produced a severe recession in the Route 128 area. The unemployment in high technology industry rose to 20 percent in the early 1970's.
The development of minicomputers saved Route 128. Ken Olsen, who had been working at Lincoln Laboratory left Lincoln with two other engineers in 1957 to found Digital Electrons Corporation (DEC). They worked on ways to reduce the size of computers and the result of their efforts was the minicomputer, a computer the size of desk instead of the size of a room. By 1977 DEC had 41 percent of the world's sales of minicomputers.
Some of the other computer firm operating in the Route 128 area were:
Data General Corporation (DG), a startup founded in 1968 by Edson DeCastro who had previouly worked at DEC
Prime Computers, founded in 1972 by William Poduska who had previously worked for the minicomputer division of Honeywell
Computervision, founded by Philippe Villers to manufacture minicomputers as components for Computer-Aided-Design (CAD) and Computer-Aided-Manufacturing (CAM) systems.
Although the first (DEC) and third largest manufacturers of minimcomputers in the world in the 1970's were located in the Route 128 area, the second largest, Hewlett-Packard (HP) was located in the Palo Alto area in California, in the area that later became known as "Silicon Valley."
Many of the founders of companies in the Silicon Valley originally came from the Midwest. Although they may have gone to college and later worked on the East Coast they did not really accept the East Coast formality and stuffiness. They found the casualness of California more to their liking. They also felt freer to experiment with new institutional arrangements in California.
The eight engineers that left William Shockley's firm to create Fairchild Semiconductor were the crucial catalyst in the development of the Silicon Valley. Fairchild Semiconductor became the training center for technological entrepreneurs. There was a high degree of cooperation and sharing of experience among the entrepreneurs and professional in the Silicon Valley. Some of this comraderie may have been a continuation of relationships that had originated as students in the top technical universities such as Stanford.
The sense of community that existed among the technical people of the Silicon Valley was not just a pleasant social phenomenon. It enable Silicon Valley firms to solve technical problems more easily and rapidly than technical people who were limited to contacts with other employees of their company. This flexibility and adaptibility in the long run gave Silicon Valley an adaptibility and flexibility that was more important to the survival of the industry than any possible loss of trade secrets. Saxenian quotes Wilf Corrigan, the founder of LSI Logic, who expresses it in terms of people thinking of themselves as working for Silicon Valley rather than a particular company.
The frequent changes of jobs in the Silicon Valley necessitated and re-enforced the community of relationships that existed. In contrast, the formality of business relationships in the Route 128 resulted in technical people being reluctant to change jobs.
The success of technical people who left career jobs to become entrepreneurs made it easier for others to take the risk of starting their own companies. There was also more of a willingness to invest in startup companies. Often those providing the venture capital were the successful entrepreneurs of the past. The office complexes on Sand Hill Road near the Stanford campus became a major center of venture capital.
The end result of the ease with which companies could be formed resulted in a large number of small companies. By 1980 there were about three thousand electronics firms in the Silicon Valley, 85 percent of which had less than 100 employees and 70 percent had less than 10 employees. The community that existed among the employees and entrepreneurs of Silicon Valley was extended to a community of interest among the companies. Cross-licensing arrangements were common.
The fact that there was a diverse technical workforce and an abundant supply of technical services and parts also contributed to the ease with which entrepreneurs could startup companies. This is the power of the agglomeration of Silicon Valley.
Saxenian notes that, in addition to Stanford and the University of California at Berkeley, San Jose State University has been a major supplier of trained technical personnel for the Silicon Valley.
It is interesting that a high technology complex developed around Stanford but not around UC-Berkeley. It may be a result of the anti-business political climate of Berkeley.
Intel, which was founded by Robert Noyce and others from Fairchild Semiconductor, had a organizational arrangement similar to HP. HP emphasized the notion of a corporate family whereas Intel, while promoting decentralization, still encouraged competitive striving for excellence.
In HP and Intel professional employees were given stock options to contribute to their sense of the unity of the sucess of their company and themselves.
Ken Olsen, the founder of DEC, attributes the closedness of business firms in New England to puritanism. Stability and frugality were highly valued. Risk-taking was looked down upon and failure was an ineradicable blemish on a person reputation. In contrast, in the Silicon Valley risk-taking was admired and failure was a temporary setback but not a calamity. And when the Silicon Valley entrepreneurs made a fortune they spent on luxuries, sometime ostentatious luxuries. Consequently a fortune was more valuable to the entrepreneurs in the Silicon Valley; it meant a definite improvement in lifestyle. In the Route 128 area getting rich did not mean a change in one's standard of living. It is no wonder there was less risk-taking along Route 128. Also the spending of the fortune-makers in the Silicon Valley contributed to the prosperity of the local economy.
Ms Saxenian quotes the founder of Convergent Technology about his experience in the Silicon Valley after having worked for eleven years at DEC:
There is no way I could have started Convergent in the Boston area....When I started Convergent, I got commitments for $2.5 million in 20 minutes from three people over lunch who saw me write the business plan on the back of a napkin....In Boston, you can't do that. It's much more formal. People in New England would rather invest in a tennis court than high technology.
DEC's operational procedures were a modification of the traditional New England business. The founder, Ken Olsen, down played hierarchy and formality. Work was carried out in project teams. In this way DEC was more like the Silicon Valley firms. DEC emphasized very strongly loyalty. In return, DEC had an unofficial policy of no layoffs. But, as Ms Saxenian points out, such a policy has the effect of making success within the company depend more on relationships with managerial staff than solving technical problems or dealing with the outside world. Ultimately the key decisions were made by Ken Olsen and the top elite.
When Edson DeCastro left DEC to found Data General (DG) there was a bitterness between the companies that has lasted for decades. DEC threatened to sue DG over the theft of proprietary technology. DG has sued other companies over such issues. The net result is that each firm in the Route 128 area aspires to vertical integration and insularity. They value security more strongly than opportunity.
The technology in electronics began to change so rapidly that there was not much advantage to being an established business in the industry. Being an old firm often meant being committed to an obsolete technology. For example, the Philco Corporation created an automated line for manufacturing transistors in 1958 but by 1963 its technology was obsolete and the investment was not recoverable. Philco left the industry.
The Route 128 firms sought to produce their semiconductor devices within the company so the area lost the economies of scale advantages that accrued to the Silicon Valley economy of having such devices produced by specialized firms.
But there was one product which did become standardized and Silicon Valley firms thought could be produced as a commodity. That product was computer memory chips. Intel introduced 1 K Dynamic Random Access Memory (DRAM) chips in 1970. Many other firms soon entered the market and price competition became fierce. In 1974 the size of the chips was increased to 4K and to 16 K in 1975. The innovations in size were made by Intel but other firms quickly matched these. By 1979 there were 16 firms in the 16K DRAM market; five of them were Japanese. The size and price competition continued to escalate. By 1984 Japanese firms introduced the 256K DRAM chips and when U.S. producers tried to match the Japanese prices they suffered substantial losses and by 1986 had dropped out of the market. There was a loss of 25 thousand jobs in the Silicon Valley.
It is easy to get paranoid about unfair competition from Japan. There is first of the problem of Japan, Inc., the network of government and private industry that turns economic competition into a political equivalent of war and tries to find ways to negate the rules of the market place. Political manipulation of interest rates and the exchange rate to produce an undervalued currency can give an insurmountable advantage to a country's producers. Lifetime employment makes labor costs a fixed cost and results in a firm being willing to continue sales at price levels that would cause other firms to drop out of production. Control of access to domestic markets can enable a firm to sell at a lower cost to foreign buyers than to domestic buyers. All of these could have been involved in the lower prices for Japanese memory chips. But the key to the Japanese success in the memory chip competion was simpler and more innocent.
Integrated circuit devices are produced by creating many copies of their circuitry on a silicon wafer. Production costs depend upon the number of wafers processed. Some of the chips in a wafer may be defected so the output of the process depends upon the proportion of the chips on a wafer that are good, the yield rate. Cost per unit are thus inversely proportional to the yield rate. Japanese producers gave greater attention to quality control and achieved substantially higher yield rates than the American producers and consequently the cost per unit device produced was about half that of the American producers.
But initially the American producers did not understand the nature of the problem. At the height of the price competion the Silicon Valley abandoned their tradition of collaboration with suppliers and customers and tried to push the cost cutting off onto the suppliers. The antagonisms that developed interfered with the solution of design problems and thus made things worse rather than better. The American producers tried to rely upon high volume and the economies of scale and this approach did not work. Silicon Valley firm's strength was in their creativity and agility in finding new products and improving them ahead of the competition rather than in the brute force economics of the production of commodity items.
Silicon Valley firms left the field of DRAM chips but Intel had created a new product that became the basis for an entirely new industy. That product was the microprocessor. This was a complex integrated circuit that could be programmed. Thus a customer who wanted a specialized device did not have to have an integrated circuit custom designed and produced in a small batch. Instead the microprocessor could be programmed to produce the same result and the microprocessors were produced in quantities in which some economies of scale could be achieved. However Intel did not stop with one microprocessor. They constantly redesigned and improved their microprocessor. The first was the 8080 followed by the 8086, then the 80286 and 80386.
The microprocessor was not invented with the personal computer industry in mind but quickly some realized that the micropressor was, in effect, a computer on a chip. The history of the personal computer is told elsewhere.
DEC's achievements in the minicomputer field were outstanding, but they have been forgotten in the wake of the even more spectacular developments in personal computers. In 1965 DEC introduced the PDP-8 (Programmed Data Processor) that sold for only $18,000 when the price tag for computers had recently been in the hundred thousand dollar price range. The PDP-8 was four times faster than its rivals. But in 1969 Data General offered its NOVA with double the speed and memory capacity of the PDP-8. DEC came back in 1977 with its VAX-11/780 super-minicomouter that had the power of a mainframe computer at a fraction of its cost. Both DEC and DG strived and to a large degree achieved vertical integration, but this made them vulnerable to technical breakthrough elsewhere in the same way that their development of minicomputers weakened IBM.
The VAX line of DEC and the NOVA of DG both had proprietary operating systems which limited access to programs developed by the general programming community. At the time the development of proprietary systems seemed the natural approach. It was only later with the proliferation of the personal computer that people began to understand the power and importance of open systems; i.e., the development of standards such as operating systems that enabled users to share their work. Even when customers began to show a preference for open systems the Route 128 companies stuck with their proprietary system approach.
The expansion of the minicomputer industry in the 1970's created a boom in the economy of Massachusetts. The boom had ended by the mid-1980's and in the late 1980's 50 thousand jobs were lost by the Route 128 firms. The major competition for the Route 128 firms was the Silicon Valley, but the real enemy of the Route 128 area firms was their organizational structure that was inappropriate for the dynamic field of computer technology.
Silicon Valley firms, relying upon components and services available in the market, were able to develop new models and even new product lines far, far faster than the Route 128 firms which insisted upon developing everything in-house and effectively had to re-invent the wheel.
The Route 128 firms not only failed to communicate effectively with the market they often failed to communicate internally. Saxenian cites the case of DEC cutting its scheduled production run on a personal computer from 250,000 to 100,000 but the divisions producing components for this computer continuing to produce 250,000.
In 1985 DEC set up a research laboratory in Palo Alto but largely ignored the information and insights this operation gained by being in the Silicon Valley.
William Poduska created the Route 128 firm of Prime Computers. After Prime was well launched Poduska left Prime to start Apollo Computer. Apollo introduced the workstation computer to the world in 1980. Sun Microsystems of the Silicon Valley entered the workstation field in 1982. Despite this two-year lag Sun Microsystem, the archtype of the aggressive and agile Silicon Valley startup, won the workstation market away from Apollo.
The early 1980's began the era of the personal computers based upon the microprocessor pioneered by Intel. Route 128 firms remained committed to the minicomputer architecture with custom-made integrated circuits for central processing units. They could not except the grim reality that technological innovations can make past technology, no matter how wonderful it once was, obsolete and as dead as yesterday's newspaper.
In contrast the Silicon Valley, although running away with the new technology of the microprocessor, did not seem to be wedded to any particular technology or product. As times changed Silicon Valley firms seemed to be able to adapt, experiment and recombine endlessly. It was a much healthier institutionally than Route 128. Saxenian cites a very eloquent statement by Tom Hayes, an executive of Applied Materials and a founder of Joint Venture. Hayes said:
Our aim is to build a comparative advantage for the Silicon Valley by building a collaborative advantage...to transform Silicon Valley from a valley of entrepreneurs into an entrepreneurial valley.
Although...Route 128's independent-firm-based system had provided economic scale and organizational stability that were valuable in an earlier era, by the 1980's they served primarily to discourage adaptation. The committment of local companies to vertical integration meant that technical capabilities and know-how ...remained locked up within large firms. The paucity of horizontal communications stifled opportunities for experimentation and learning while traditional corporate structures limited the development of managerial initiative and skill....This may have imposed a minor inconvenience to large firms, it bacame a significant disadvantage for start-ups and small firms that were unable to learn about or acquire state-of-the art components or services as rapidly as their West Coast counterparts.
It did not take the talented people long to realize the land of opportunity for them was not along Route 128 but instead in Silicon Valley.
Although Silicon Valley lost the price war on memory chips to the Japanese producers there were still important markets left. There reappeared a market for custom-designed chips, or as they were now called Application-Specific Integrated Circuits (ASIC's). The aggregate market for these amounted to something comparable to the memory chip market in revenue. Instead of seeking profits in terms of economies of scale the custom chip makers gained profits from the differentiation of their products. The new chip startups produced one to two hundred different types on "mini-fab" production lines with runs of ten to ten thousand chips in constrast to commodity product "mega-fab" runs of millions of the same chip.
In the 1980's there was a new generation of startups in the computer industry of the Silicon Valley. Some of these were:
| Company | Specialty |
|---|---|
| Sun Microsystems | workstations |
| Silicon Graphics | 3D graphic workstations |
| MIPS Computer Systems | RISC architecture computers |
| MasPar | massively parallel computers |
| Tandem | |
| Pyramid Technology |
There were also major startups in the Silicon Valley in the 1980's in fields such as computer peripherals and software.
Silicon Valley firms frequently limit their purchases from any one supplier to keep themselves from becoming excessively dependent upon one source of a crucial input. Likewise firms also try to avoid becoming excessive dependent upon any one customer.
Saxenian gives an interesting example of firm-supplier symbiosis in the case of electronics assembly. Some firms such as Flextronics began to do contract assembly work, what was called "board stuffing." In the 1970's these board stuffing specialists were small and low tech. The client provided the components and the directions. By the 1980's firms like Flextronics had developed special equipment and could provide expert guidance in the selection of components. When a state of confident trust developed between the electronics firm and the board stuffers the firm could turnover much of the selection and procurement of components to board stuffing specialist. The board stuffing firm might also assume responsibility for testing of the finished devices. This type of operation involving component selection, procurement and testing is called "turnkey." Flextronics ultimately played a role in design of printed circuit boards the company "stuffed." Some technological advances were also made in assembly. The traditional assembly operation was the soldering of wires that passed through holes in the circuit boards. A new method was developed, called "surface mount technology," (SMT) which involved the fastening of wires to the boards with epoxy cement. This method allowed for mounting components on both sides of a board. This method was pioneered by the Silicon Valley firm of Solectron. This technology involved a large capital investment and would not likely have been developed in vertically integrated firms.
Another example of firm-supplier synergy is the relationship between HP and Weitek. Weitek designed ultra-high-speed chips for faster numerical computation. HP purchased Weitek chips for its computers. But Weitek achievements were being limited by the state of its chip fabricating operation, its foundry. HP discerned that Weitek using HP higher quality foundry could produce better chips for HP. HP opened its foundry to Weitek to use to produce chips not only for HP but also for other Weitek customers. This was a highly beneficial arrangement for both Weitek and HP.