Re: [Patents] Re: Bhagwati on Trips

[Rufus Pollock <rufus.pollock@okfn.org>]

Laura Creighton wrote:
> > But in patents it is different, as Bhagwati m
> > entions: "we have here the economist’s tradeoff: more IPP means more knowl
> > edge but it also reduces diffusion. The social optimum lies somewhere in 
> > the middle, as often."
>
>
> I think that the assertion 'more IPP means more knowledge' is highly
> debatable.  It is trying to measure the knowledge gained by people
> snip ...

Yes and quite a bit of modern economic research backs this up. More IP does 
not /even/ mean more knowledge. And it guarantees /more monopolies/. The 
original 'crude' stuff (e.g. Arrow 1962 and Nordhaus 1969 and following) just 
assume:
1. only way to appropriate returns / exclude others from innovation/R&D is 
with IP
2. that innovation is 'one-shot' (so not cumulative)
3. no transaction cost issues or imperfect information issues (for example in 
the real world the scope of a patent will be never be perfectly specified)
4. ideas are completely non-rival / infinitely expansible. That is: there is 
precisely 0 cost of reproduction and reproduction can be done instantaneously
...

As I am sure some of you know recent work such as Bessen and Maskin 1999 
(sequential innovation, patents and innovation) show that altering assumption 
2 can lead to patents being a hindrance to innovation.

On an even stronger tack recent other work by Boldrin and Levine and others on 
changing assumption 4 entirely overturns the traditional analysis. It shows 
that even in a world with no IP rights at all innovation could and would be 
paid for. In fact when fixed costs of innovation are sufficiently small 
without IP you can obtain the 'Pareto' optimum (sorry for the economists jargon).

> What's worse, it cannot measure the people, who considered embarking
> on an investigation that is likely to produce knowledge, but stopped.
> They believed it was more likely that they would be the third, fourth,
> fifth or sixth discoverer, rather than the first.  Thus they feared 
> their effort would be wasted -- whatever they discovered would be
> owned by patent-holder #1.  So they never bothered to discover it all
> all.  Nobody did.  We still don't know it.

This isn't an argument for inefficiency however. there is a long literature on 
patent races addressing some of these precise issues (is there too much or too 
little innovation in a 'winner takes all' system)

> Since you cannot measure these things it is hard to conclude that the
> knowledge stronger IPP will inspire outnumbers the knowedge stronger
> IPP will inhibit.  It is in the realm of the unknowable.

Absolutely and this is a constant problem in this debate that the empirical 
evidence is so scant and so uncertain. However you can still try - there are 
certainly quite of a lot of anecdotal/case study work on the uses and values 
of IP in different industries.

Regards,

Rufus