used by the old to protect themselves against the threat of the new. This pro-
tection is not necessary; there is nothing in our tradition that compels it. But
it is pushed not by those with the most to lose, but by those without the most
to win. And I have argued that we should be skeptical about just this sort of
protectionism.

But now I want to describe a second form of protectionism -- perhaps
more threatening to the promise of the Internet's future. This threat too is
the product of state intervention into Internet space. And this intervention
is even harder to justify.[11-61]

the issue here is patent law.[11-62] A patent is a form of governmental regula-
tion. It is a state-backed monopoly granting exclusive rights to an "inventor"
for an invention deemed useful, novel, and nonobvious.

The argument favoring patents is as old as the hills. If an inventor can't
get a patent, then he will have less incentive to invent. Without a patent, his
idea could simply be taken. If his idea could simply be taken, then others
could benefit from his invention without the cost. They could, in other
words, free-ride off the work of the inventor. If people could so easily free-
ride, fewer would be inventors. And if fewer were inventors, then we would
have less progress in "science and useful arts."

Getting more progress is the constitutional aim of patents. So the ques-
tion that must always be asked of any patent regime is whether we have good
reason to believe that patents have that effect. As Harvard law professor
Stephen Shavell has written, "there is no necessity to marry the incentive to
innovate to conferral of monopoly power in innovations."[11-63] So is there any
evidence that it does any good?

In some cases, the evidence is good.[11-64] For some kinds of innovations,
patents are extremely likely to induce more innovation. In particular, in
theory, where innovation is independent, or noncumulative (meaning one
invention is essentially separate from another), then economists predict that
patents will clearly benefit innovation.[11-65] Likewise, even where innovation is
cumulative, if the use of the patent is clear, then in principle, the original
patent holder will have a strong incentive to license a patent to follow-on in-
novators.[11-66] But here, economists have an important qualification: If we
don't know which direction an improvement is likely to take, then licensing
may not occur, and patents here may actually do harm.[11-67] Thus, for econo-
mists, at least, the theory suggests contexts in which innovation will be
helped by patents as well as contexts where it will be harmed.[11-68]

[[205]]

p204 _ -chap- _ toc-1 _ p205w _ toc-2 _ +chap+ _ p206

v?