[Patents] Patnews on Pierret

[Bernard Lang <Bernard.Lang@inria.fr> (Bernard Lang)]

Daniel

Thanks for this very clear statement, and for the excellent remark at
the end about novelty (the hitch being that I am not sure it could be
implemented for the patent system, but at least you show that there is
a real difficulty).

  One remark though.  The pro-patent lobby is standing its ground on
the following arguments:

  - patents help find venture capital to get new companies started, so
as to bring more innovation on the market.  Venture capitalists like
to see what looks like tangible asset.
  Note:
   This is of course purely psychological (in my opinion), since
people and experience matter more than patents (and it runs contrary
to studies on SMEs).  But psychology of economic actors, however
misguided or insane, is still to be taken into account by economic
analysis.

  - patents are necessary to give value to company stock, so they they
stand stronger on the stock market, or when mergers occur.  So a
country without patents would weaken its companies, or encourage them
to migrate where patents are available.
  Note:
   This argument, in my opinion does not hold water too well, since
studies show that SMEs (on the average) do not use and do not like
patents.  But it may hold for larger corporations.  On the other hand,
large corporation are transantional, and can therefore patent where
that is possible... like anyone else.  And larger corporation are not
the software innovators (nor do they create many jobs).

  - patents encourage researchers to consider applications and get
more out of their labs, and bring more things to market.
  Note:
  I have no idea whether that is corroborated by any available data.
The free software approach of the 70s-80s was quite effective in
having academic work put to use.
   ... and the main motivation of the free software people is just
that: have their work used.  In fact, software may be the only field
where there is a direct connection from the producer to the consumer,
without real need for intermediaries.... and that creates immediate
recognition rewards.
  Furthermore this pro-patent view of research has to be balanced with
the fact that free software communication, which was the rule among
researcher (with exceptions of course), will be subject to
infringement risk, and that may slow research, not to mention other
factors due to the fact that patenting reinforces monopolies.


   There are no other economic arguments that I have been able to
identify or remember (maybe I should reread some of the memos I
accumulated:-) ... only the usual lawyer junk-talk.
   But I welcome any suggestion of other economic pro-patent arguments.


Bernard

PS
  Greg states that there should be lines ... but not where they stand.


On Thu, Jul 12, 2001 at 11:37:02AM +0200, Daniel Probst wrote:
> Greg,
> 
> I do not quite understand your point of view. (sorry for jumping in, I may
> have missed some clarifying mails)
> 
> Bias warning: as an economist I believe that the only valid argument for or
> against patents is economic (same as your constitution). Issues of elegance
> and even absolute consistency  in legal implementation are of secondary
> concern.
> 
> 1) Economic arguments for/against weak/strong protection of IP crucially
> depend on the nature of the industry .
> >From an economic perspective there is a priori absolutely nothing against,
> for example, allowing patents in the cheese industry and banning patents in
> the ice cream industry. Whether or not this is a good idea solely depends on
> the production/research technologies and the market structure in the
> respective industries. The fact  that it might be legally or logically
> difficult (or even impossible) to correctly evaluate whether cheese flavored
> ice cream, or frozen cheese belongs to one or the other category is of
> secondary importance.
> 
> 2) The balance of theoretical considerations as well as  empiricial
> corroboration indicates, that in industries with the market structure and
> production/research technology similar to the software industry, patents are
> not an efficient means of IP protection. (Contrary to for example in the
> pharmaceutical or chemical industry.)
> Given the current level of (economic) knowledge in this area, the question
> being debated is more "how bad are patents" in an industry like software
> and not "are they beneficial or bad" ).
> 
> I know of no empirical or theoretical study which indicates that patents in
> industries comparable to software  might conceivably be beneficial.
> 
> 3) I do no see any compelling reason for discussing patent issues "as a
> whole". Except of course at a level where the suggested "universal patent
> system" allows for the distinction between different industries, takes
> issues of complementarities/sequentiality into account, looks at market
> structure, looks at cost structure, takes potential entry into account,
> takes network externalities into account etc. etc. etc
> 
> 4) You frequently claim that many problems arise due to the bad
> implementation of the current patent system. With a properly working patent
> system most problems would go away, i.e. transaction costs would be so low
> that a patent system would be beneficial to a broad area of IP. To me you
> also seem to suggest that one should go ahead with stronger IP related
> property rights and solve the problems of implementation later (furthermore
> these problems are solvable).
> 
> The short list in 3) of relevant issues a good patent system must address,
> suggests to me  that this view is debatable, to say the least. Even given
> the assumption that novelty can correctly be assessed (whatever that means,
> see below ), there are still a multitude of market specific issues
> (competition issues, market entry issues etc.) which you never seem to
> address.
> 
> Regarding  novelty:
> 
> If you extend patentability to research which in many ways is strongly
> related to what is traditionally seen as academic research , let me ask
> following question:
> 
> Why does academia have such great difficulties in assessing novelty (Maybe
> its because we are all daft ;-). Why has such an ellaborate system evolved
> in academia:
> 
> 1) Researchers publish their research in "famous" journals and thereby gain
> reputation (future monetary income).
> 2) Journals choose referees who assess novelty. Journals have the incentive
> to choose competent referees who choose articles which will be judged novel
> by their audience.
> 3) Journals compete against other journals. The competetive pressure induces
> them to pick competent referees (who MUST also be well known researchers in
> the field).
> 
> Why is this system so complicated? Why don't we have property rights in e.g.
> economic theory or mathematics. I pay license fees to people whose research
> I am citing/basing my work on. I then bring my research to the a state paid
> university bureaucrat who judges the novelty of my research, and stamps
> novel on it. I can then require license fees from researchers building on my
> work (who might then pay a lawyer to try and bust my patent, claiming it is
> not sufficiently novel).
> 
> Does the fact that we do NOT choose this institutional arrangement, but
> instead the above mentional reputational system, indicate anything to you ?
> 
> dan probst
> 
> ----------
> daniel.probst@vwl.uni-mannheim.de
> Dr. Daniel Probst
> Economic Theory
> Mannheim University
> Germany
> ----- Original Message -----
> From: "Gregory Aharonian" <srctran@world.std.com>
> To: "Gregory Aharonian" <srctran@world.std.com>; "Don Marti"
> <dmarti@zgp.org>
> Cc: <patents@aful.org>
> Sent: Tuesday, July 10, 2001 7:40 PM
> Subject: Re: [Patents] Patnews on Pierret
> 
> 
> > Don writes:
> > >Greg, you don't seem to be willing to draw the line on what is
> > >patentable and what is not.  Do you even agree that there should
> > >be a line?  If there is a line, software was more in common with the
> > >activities on the "no patents" side of the line.
> >
> > There are lines, there should be lines, and software is on the patentable
> > side of the line.
> >
> > And to be clear, I am making my arguments mostly in the context of US
> > intellectual property law.  People in Europe and free to agree or
> > disagree from the context of European IP law.
> >
> > And for US IP law, the line is quite clear:  processes and methods,
> > tangible or intangible, are patentable, not copyrightable.  That is
> > a very clear line.  And it is back up with 100 years of court decisions
> > here in the US.  That is the legal rule.  The mathematics rule is
> > also quite clear - software (i.e. source code, user interfaces, system
> > designs, protocols, etc.) are processes and methods.  In fact, if I can
> > be presumptous and use a language I don't speak:
> >
> >                         Software ist Verfahren
> >
> > Don, wonder over to the Stanford mathematics library some afternoon, and
> > you will find thousands of articles proving/assuming this to be true.
> >
> > So combine the two, law and mathematics, and software is on the patentable
> > side of the line, especially in light of hardware/software codesign tools
> > and systems to map C code directly into FPGAs.
> >
> > That said, as I have said endlessly, the quality of software patents needs
> > to be drastically improved, so that all of the crappy software patents
> > that I and others hold up to ridicule the system, don't issue.
> >
> > This argument, made formal, may not be completely perfect, but it is close
> > enough that any argument against software patents is going to be pretty
> > much economics-based.  The problem there is that like software, the
> > economics of patenting in other areas of technology have their
> controversial
> > issues (said with drug patents in developing countries, etc.).
> >
> > These economic issues are worthy of discussion, and are being discussed,
> > but they apply to all areas of patentable technology, not just software.
> > And if you really concerned about the economics of IP, I would think that
> > you would be more interested in addressing the economics of the excessive
> > lifetimes of copyright, 100 years or more of protection, especially if
> > you want to rely on copyright protection for software (which I will argue
> > is illogical under US law, the increasingly de facto viewpoint of US IP
> > law, especially after the Apple and Lotus losses).
> >
> > Greg Aharonian
> >
> > _______________________________________________
> > Patents maillist  -  Patents@liberte.aful.org
> > http://liberte.aful.org/mailman/listinfo/patents
> >
> 
> 
> _______________________________________________
> Patents maillist  -  Patents@liberte.aful.org
> http://liberte.aful.org/mailman/listinfo/patents

-- 
         Non aux Brevets Logiciels  -  No to Software Patents
           SIGNEZ    http://petition.eurolinux.org/    SIGN

Bernard.Lang@inria.fr             ,_  /\o    \o/    Tel  +33 1 3963 5644
http://pauillac.inria.fr/~lang/  ^^^^^^^^^^^^^^^^^  Fax  +33 1 3963 5469
            INRIA / B.P. 105 / 78153 Le Chesnay CEDEX / France
         Je n'exprime que mon opinion - I express only my opinion
                 CAGED BEHIND WINDOWS or FREE WITH LINUX