Against Intellectual Monopoly

[B]ecause of concern
over global price regulations ... innovative pharmaceuticals may actually become
available to Indian consumers more slowly.23
    Small firms prefer other mechanisms (e.g. secrecy) to protect their innovation or
distrust patents, maybe because of the large costs involved in defending a patent.
Another explanation of this result would be that small firms - on average - are more
engaged in incremental innovation which does not fulfill the novelty requirement of
patents. Moreover, large firms more probably apply for patent due to institutional
requirements.... In addition, firms apply for patents because patents are used in
cross-licensing agreements with other firms.24
    National patent protection alone does not stimulate domestic innovation, as
estimated by changes in citation-weighted U.S. patent awards, domestic R&D,
and pharmaceutical industry exports. However, domestic innovation accelerates in
countries with higher levels of economic development, educational attainment, and
economic freedom. Additionally, there appears to be an optimal level of intellectual
property rights regulation above which further enhancement reduces innovative
activities 25
    However, econometric analysis using both Japanese and U.S. patent data on 307
Japanese firms finds no evidence of an increase in either R&D spending or innovative
output that could plausibly be attributed to patent reform.26
Route 128 and Silicon Valley
    We now take up the tale, not of traditional intellectual monopoly such as
patents and copyright, but of restrictive noncompete labor contract clauses.
Although these are not, strictly speaking, tantamount to patents and copyright, they serve a similar purpose - that of maintaining monopoly over an
innovation - and they are often used in place of patents where the latter
are not legally allowed or easily enforceable. As the remark by Gary Becker
we quoted in the previous chapter makes clear, legally preventing workers
from spreading the knowledge they acquired in previous occupations is an
inefficient way to internalize knowledge spillovers, something that could
much more efficiently be achieved by using prices and wages. Noncompete
clauses in labor contracts are a very common example of such inefficient and
monopoly-inducing legal means. Testing their impact on the rate of innovation should help our understanding of the extent to which intellectual
monopoly serves a beneficial social purpose.
    You have probably heard of Silicon Valley. Perhaps you have not heard of
Route 128. Yet Route 128 in Boston has been a high-technology district since
the 1940s, long before farmers were displaced from Santa Clara Valley, as Silicon Valley was then known, to make space for computer firms. In 1965,
both Silicon Valley and Route 128 were centers of technology employment
of equal importance, and with similar potentials and aspirations for further
growth:

    Route 128 began the race well ahead. In 1965, total technology employment in the
Route 128 area was roughly triple that of Silicon Valley. By 1975, Silicon Valley
employment had increased fivefold, but it had not quite doubled in Route 128,
putting Silicon Valley about fifteen percent ahead in total technology employment.
Between 1975 and 1990, the gap substantially widened. Over this period, Silicon
Valley created three times the number of new technology-related jobs as Route 128.
By 1990, Silicon Valley exported twice the amount of electronic products as Route
128, a comparison that excludes fields like software and multimedia, in which Silicon
Valley's growth has been strongest. In 1995, Silicon Valley reported the highest gains
in export sales of any metropolitan area in the United States, an increase of thirtyfive percent over 1994; the Boston area, which includes Route 128, was not in the
top five.27
    What explains this radical difference in growth of the two areas? Certainly
both had access to important universities, which are instrumental in the
computer revolution - Harvard and MIT in the case of Route 128 and
Stanford in the case of Silicon Valley. A careful analysis by Ronald J. Gilson
shows that the only significant difference between the two areas lay in a
small but significant difference between Massachusetts and California labor
laws. According to Gilson:
    A postemployment covenant not to compete prevents knowledge spillover of an
employer's proprietary knowledge not, as does trade secret law, by