Sounds like it will be a good paper. On the title... Hmmm... Keep in mind that many more people will read the title than the first sentence, so the first sentence should be subordinate to the title, not vice versa. Imagine someone skimming a table of contents that she has just been emailed: can she glean your argument from your title? I might modify the title to mention pharmaceuticals directly, and to try to make your argument. Also, your title and first sentence suggest that you are interested in market size and innovation generally--but then in sentence two you introduce pharmaceuticals. The transition from general to specific strikes me as slightly abrupt. Again, mentioning pharmaceuticals in the title could help signal what the paper is really about. Finally, I am not sure I understand the core contribution. Past research has used ostensibly exogenous demographic variation to estimate how changing market sizes impact drug innovation. But where exactly does drug usage come in--do such techniques need to make assumptions about usage rates? I wasn't quite clear on that. And a note for future work: on the issue of sensitivity to exogeneity, you might check out Jonathan Wand's short article in the Political Methodologist from a few years ago, it provides an algorithm for estimating results' sensitivity to exogeneity assumptions. Best, Dan On Tue, 9 May 2006, aram harrow wrote: _______________________________________________ gov2001-l mailing list gov2001-l(a)lists.fas.harvard.edu http://lists.fas.harvard.edu/mailman/listinfo/gov2001-l

thanks for the comments! how about the more boring but descriptive: Pharmaceutical market size and innovation: Can we estimate causal effects?

Attempts to directly measure the effect of market size on innovation---a key economic question---typically run into severe problems of endogeneity and omitted variable bias. One possible solution (due to Acemoglu and Linn, 2004) is to use exogenous demographic trends to estimate the market size for different

categories of drugs. However, after controlling for category and time fixed effects to remove endogeneity, the residual variation in market size becomes so small that robust estimation of its effect on innovation becomes impossible.

For example, a change of less than 0.2% in drug usage patterns could be chosen to either eliminate or triple the estimated effects of market size on new drug approvals. We are left with a dilemma: either to base estimates on large variations in market size that are vulnerable to endogeneity, or to use small fluctuations that are hopelessly sensitive to minor specification errors. On 5/10/06, Gary King <king(a)harvard.edu> wrote: