[Patents] economic studies

[PILCH Hartmut phm@a2e.de]

On Sat, 6 Jan 2001, Gregory Aharonian wrote:

> Bernard cites:
> >   The theoretical and other economic literature does not demonstrate,
> >   indeed casts doubt, on whether economic efficiency, i.e. increased
> >   overall welfare, is achieved by having or making computer program
> >   related inventions patentable.
> 
> Again, whoopee.  

Again, you fail to answer the challenge.
The rest is excuses, reiterated for the umpteenth time.

> Without seeing the economic assumptions, and simplifications made in
> these studies, as well as knowing the patent world experience of the
> economists involved in these studies, I do not put much faith in their
> conclusions.  

Let me resume:
You know no economic studies that seriously argue for a positive impact of
patents on software innovation.
You know that economic studies that argue for the contrary exist, but,
without having read them, you don't put much faith in them.
Good.
 
> For example, I have heard a variety of economists talk
> about measuring the flow of information between companies by measuring
> how they cite each others patents. Yet many of these economists are
> unfamiliar with how patent searching is done (often not by the
> inventors, meaning the citations don't measure information flows) to
> the extent that it discredits their conclusions.

There have been a lot of studies on the value of patent statistics,
including those pertaining to citations of other patents in patents.
The people in this field are not as naive as you portray them.
Everybody knows that these data are only very indirect indicators.
And the studies we cited don't heavily depend on such statistics.
Anyway, finding fault with studies you don't like is easy.
The point was not whether the ones we cited have some faults, but whether
you can cite some that claim swpat are beneficial for promoting progress
and systematically argue the case for this claim.
On that point your implicit answer was "no".
Your method of refusing to make your answers explicit and instead shifting
the focus discredits you as a serious dialogue partner.
 
> That there are few economic studies to support software patenting means
> little as well. 

OK, this sounds a little more direct, although your answer is nested in a
"that ..." clause.
Let me repeat.
"There are few economic studies to support software patenting."
So please name at least one of these few studies.
Preferably the one with the most systematic argumentation for the case.

> First, it is hard to measure the impact of patenting for any field of
> technology.  Second, assuming patents have some value in general, and
> that software patents are as much a part of the patent process as any
> other field of technology, there is little interest to study anything
> other than the general impact of patents.

There is little interest to study the "patent process", and even less
interest in metaphysical assumptions about "patents in general".

Most existing studies focus on the effect of the patent system on
individual chosen fields (such as drugs, semiconductors, software,
business methods).

OK, your answer is that "all existing studies can be ignored, because it
is wrong to differentiate between the impact on different fields".

> So the reckless citing in this forum of economic studies whose assumptions
> and qualifications is unknown proves nothing.  And relying on such studies
> seems a good way of losing an argument.

Our argument didn't rely on any studies.
But denying the value of those questions on which you lose is a way of
doubly losing your argument.
 
> So if Europe in general is serious about this issue, it might want to
> get the EC to fund a comprehensive study of the impact of patents,
> past and future, for all areas of technology.  Until then, randon
> citation of random studies proves nothing.  My bet, and I would be
> glad to put my money with my mouth is, is that the impacts of patents
> in a variety of fields, software, hardware, etc., in terms of losses
> of freedom and providing incentives to innovate, will be relatively
> the same.

Such comparisons already exist.
It is commonplace to claim beneficial effects for drugs and bad effects
for software innovation.
The reasoning is very simple, and it is pervasive in studies and even in
statements of pro-patent lobby people.

Now your tactics again is demanding that all existing studies be ignored
and instead one new irrealistically huge project be started from scratch.
What I called "selective perfectionism".
By pointing out one or two weaknesses of the other side, you may be able
to rally those partisans behind you, who are blind on the same eye on
which you are blind.  But any somewhat philosophically skilled third party
will easily judge that you lost the argument.  And in the long run you do.

> And for the person who doesn't read English, I mentioned earlier that
> IEEE publications and software publications in general are straining
> the storage space and organizing time of librarians to simply point
> out in terms of publication volume, that software patents are having
> zero impact on the development and publications of new software ideas.
> Zero impact.

Didn't you say just before that information flow is very hard to measure?
Now you gave a beautiful example of your selective perfectionism.
Apparently, while patent statistics may not be used as indicators of
information flow between companies, IEEE papers so exactly represent the
information flow that they sustain even the boldest conclusions.  Provided
that these conclusions are in line with what you want to argue.

-phm