This is the type of paper we ought to be looking for, although the
result is in the "wrong" direction (I've only briefly looked over it,
but it appears to be a classic correlation->causation argument, and they
say things like "In terms patent activity, we observe that non-resident
patent registrations are weakly negatively related to female self-
employment rates", which is a large WTF to my mind).
An obvious problem that stands out is that they're using data sets from
the World Bank, music industry, etc. etc. - all data which I'm pretty
sure will be intrinsically biased. We know that the music industry
regularly peg the "cost of piracy" in the billions of Euros and bring
ridiculous value lawsuits against infringers; if those economic values
are being brought into the calculation it seems clear to me that the
argument would weight totally in one direction.
Their finding "Moreover, that a commitment to international IPR laws has
a positive impact on self-employment rates" is a tremendously bizarre
statement to make given that they're not looking at historic data
(unless I'm misreading this).
There's a nice bibliography at the back that would probably deserve some