Against Intellectual Monopoly

were priced, provided that information
transfer from nearby firms is less costly than it is from distant firms. Did
Silicon Valley form so that employees might overhear valuable ideas in bars,
or because it made it relatively easy for firms to interact with one another
contractually? Certainly, evidence supporting the idea that large unpriced
spillovers take place among innovating firms is scarce at best. Ellison and
Glaeser provide the most careful analysis, finding only very weak evidence
that agglomeration is due to spillovers.14 Other studies find even weaker
or no evidence for the allegedly pervasive unpriced spillovers. Acemoglu
and Angrist, for example, estimate average schooling externalities at the
U.S. state level and find no evidence of significant externalities." Ciccone
and Peri examine local labor markets to test whether productivity increases
with the average human capital of the workforce in the area where firms are
located; their data reject the hypothesis.16 Castiglionesi and Ornaghi look
carefully for external effects in a large panel of Spanish manufacturing firms
data and conclude that they cannot find any.17 Most anecdotal evidence
about industrial agglomeration, from Silicon Valley to the greenhouses
of Almeria, suggests that firms do price informational and technological
spillovers into the wages of their employees.
    If unpriced spillovers are indeed important, it must be that ideas are
so inexpensive to transmit that mere observation is enough to convey the
essential core of the idea. Here the evidence is overwhelmingly against:
there is a large literature on technology transfer strongly indicating that -
even with the active help of the innovator - ideas are difficult and costly to
transmit. Several examples of technology diffusion illustrate the point.
    One of the earliest known examples of the diffusion of technology is the
spread of agriculture during the Neolithic period. Work by Cavalli-Sforza
and others has documented that the average speed of diffusion of agriculture
was of about one kilometer a year, over a period of many thousands of years.
Transportation available at the time - walking - could carry the ideas many
thousands of kilometers per year, so there is a difference of three orders of magnitude between the rate at which ideas could physically move from one
location to another and the rate at which the idea actually was transmitted
and became useful.18

    Of course, part of the reason for the slow diffusion of agriculture was the
need to adopt crop strains to local circumstances, not merely the need to get
"the idea" of agriculture. But the adaptation of ideas to local circumstances
is important for most ideas - books printed in English, for example, are not
of terrifically great value in China. As we argue elsewhere, competition, and
not monopoly, generally provides the collaborative advantage that speeds
diffusion. If copyright laws were enforced in China so that English books
could not easily be pirated into Chinese translations, is it likely that this
would increase the speed with which translations became available?
    Another good example is that of seventeenth-century silk production:
    In 1607 Vittorio Zonca published in Padova his Nuevo Teatro di Machine et Edificii,
which included, among numerous engraving of various contraptions, the description of an intricate water-powered machine for throwing silk in a large factory.
Zonca's book went into second edition in 1621 and a third in 1656.... G. N. Clark
has shown that a copy of the first edition of Zonca's book had been on the openaccess shelves of the Bodleian Library from at least as early as 1620.19
    Yet despite the fact that the blueprint for a silk factory was readily available,
it was not until one hundred years later that "the English succeeded in
building a mill for the throwing of silk." This occurred only after "John
Lombe, during two years of industrial espionage in Italy, found means to
see this engine so often that he made himself a master of the whole invention
and of all the different parts and motions."20
    Other examples from the past also show the difficulties involved in transferring knowledge. There are many cases of individuals migrating to find
out about technologies and inventions. To learn to work the dockyards, to
make the pendulum clock, or to make woolens, you moved to Holland. To
learn to cast ordnance, you moved to England. To make spectacles or