[misc.legal] Do patents do any good?

turpin@cs.utexas.edu (Russell Turpin) (04/13/90)

-----
Follow-ups directed to misc.legal only.

-----
brnstnd@stealth.acf.nyu.edu (Dan Bernstein) writes:
> For thousands of years ideas have been free.

In article <2148@mit-amt.MEDIA.MIT.EDU>, edit@mit-amt.MEDIA.MIT.EDU (Film/Video) writes:
> That may be so, but for hundreds of years inventions have been
> patentable.

Which ducks the question of whether patents are a good or bad
idea.  Obviously, to those individuals who are enriched by
patents, they seem a good idea indeed.  But are they good for the
industry concerned or the economy as a whole? 

The theory behind patents is that without them, people would be
less driven to create new things and processes, and companies
would be less willing to fund the research behind this.  The cost
to the economy of having patents is two-fold: (1) the inventions
and processes that are patented are less well utilized, and (2)
there is less incentive to produce improvements on things and
processes that are already patented. 

It is important to note that there remains great incentive to
creation even without patents.  There is considerable time
between coming up with an idea and bringing the results to
market, and this lead time plus the marketing advantage of being
there first provides a considerable edge to companies who
successfully do it.  

The software industry of the last three decades is an example of
a field without patents or their equivalent, since the
"processes" and "things" which drive it are ideas (algorithms,
interfaces, techniques, etc), which until recently were not
protected.  As Mr Bernstein writes, ideas were free.  But the
software industry can hardly be characterized as one that lay
fallow for want of sufficient legal protection of invention.
Despite the fact that Lotus had no protection from other
companies producing spreadsheets, IBM from other companies
producing relational databases, Hoare from people using Quicksort
without permission, etc, the software industry is as full of
people trying to do something new, different, and better as any
industry has ever been. 

I have read of a few economic studies that attempt to illuminate
the question of whether patents help or hurt the economy.  It is
a thorny problem, because it involves the counterfactual.  We
cannot directly compare two industries that are identical except
that one developed with patents and the other without; but
instead are limited to asking "what might have been?"  The
studies I have read about all put a dim light on patents, suggesting
that in almost all cases, the industry and economy as a whole was
hurt by them rather than helped.  (Does anyone know of
countervailing studies?)

Large and established companies like the idea of patents.  They
can more easily afford the research required to create them, are
in a better position to make use of an exclusive right, and have
legal staff to bend the patent laws to their best interests.  In
many cases, it does not matter to the large company whether their
own researches developed the work, or that they license it from
someone else.

For small companies, the addictional cost of licensing patents,
and even determining what patents need to be licensed can be
significant.  An inventor can create something new, only to be
prohibited from any application because the components or some
part of the process is already patented.  (The individual who
starts a new industry relying on a patented device is more
frequent in fiction than real life.  Most often, individuals with
a good idea must license their patent to existing companies in
order to see any profit from their endeavors, and quite
frequently, there is only one buyer with the other patents
required to make use of the inventor's improvement.)

It may be time to shift the focus of the debate.  Instead of
arguing whether or not patents are needed for algorithms and
interfaces in the software industry, we might do better to ask
whether or not patents should be eliminated in other fields. 

Russell