Subject: Re: "incentive void" (was Re: A different patent covenant...)
From: <stephen@xemacs.org>
Date: Thu, 5 Oct 2006 19:43:26 +0900

simo writes:

 > One of the very bad things of the patent system is that it is a sort of
 > lottery, the first to file get the patent, all the others will just get
 > the research costs and no benefits at all.

You're absolutely correct.  This effect has been accepted for about 3
decades in economics.  (Specifically, Tom is only partially correct to
say they should figure the risks in; of course they will, but the
research is still more or less duplicative, and therefore a net
burning of money.)

However, it's hard to measure this effect empirically, there are a
number of theoretical countervailing effects, and even this theorem
depends on how close the research done is to being 100% duplicative
and whether there are joint products (ie, a leg up on a non-
duplicative innovation).

 > I think that a rule that states that in case two or more party file
 > a patent for essentially the same technology either no patent is
 > granted and the invention is considered obvious or all the parties
 > are granted a shared patent.

"No patent" has a problem, because the patent rarely gets you to a
marketable product.  But the implementation and marketing costs are
also typically fixed, and therefore are likely *not* to be recovered
under full competition; without any patent it increases the risk that
marginal inventions will suffer long delays before they get to market,
although much money may have been spent to develop them.

"Shared patent" under the usual assumptions in the economic literature
is irrelevant, because the parties are risk-neutral.  In the real
world, I would expect that SME developers would like it a lot (except
that I wonder if the probability is really so high that it affects the
incentives much).  However, the net effect on society should be
basically the same, because monopoly should be the highest profit
setup, and so you would expect that the "strongest" firm would buy out
the others.  This of course depends on antitrust posture, but at least
in the U.S. the main theory is that a patent is a monopoly granted by
the nation, so the FTC has no business interfering with its exercise.

Since the social calculus seems to be null, it's an interesting idea
for the SMEs, which are a good thing.