It’s key to innovation.

The ex ante / ex post dilemma arises for government encouragement of positive externalities, too. Over the past thirty years or so, questions of how best to design systems for encouraging innovation and the arts have risen to become perhaps the most important debates in intellectual property.223 The core of the debate is whether prizes---basically, grants awarded for achieving particular outcomes---should supplement or even replace our existing patent regime.224

For a time, commentators assumed that prizes had to be offered ex ante, while patents are essentially an ex post device whose value is determined by the behavior of consumers after the invention is completed.225 In that context, a key point of contention was whether the informational benefits of waiting outweighed the lower social cost of prizes.226

My work here suggests that we might view prizes more favorably. As in the negative externality context, commentators seem to assume that government needs enough information to match the value of its incentives fairly closely to the social benefits of the invention.227 Again, my claim is that this degree of accuracy is usually not worthwhile. A system of research grants that offered only a few different reward structures, and matched them with decent accuracy to projects, could eliminate most of the deadweight loss involved in a patentless system.228

Similarly, critics of the research & development tax credit might argue that it is fairly unconnected to the ultimate social value of an innovation.229 In fact, though, the tax system does a reasonable, if rough, job of tailoring the value of the credit, since at least in the startup context the credit is only useful to firms that end up being profitable.230 This duplicates the patent system’s method of valuing inventions: it pays more to those that sell. If desirable, this targeting could be extended to older firms by limiting the use of the credit to offsetting profits from inventions funded through recent R&D.

Commentators have also suggested that prizes, too, might be shifted ex post in order to allow the government to have more information at its disposal when the prize is awarded.231 My analysis suggests some reasons to be cautious about this move. As we saw earlier, the gain in information would be at least partially offset by increased credit and agency costs. In other words, entrepreneurs would have to raise outside capital on the promise of a future prize, with all the costs that such a separation of ownership and control usually entails. That the separation occurs in a setting with especially high information asymmetry---the entrepreneur knows far more than the investors about her invention and her future effort in developing it---makes the contracting situation especially difficult.232

What this perhaps suggests is that the ideal innovation system would in some way blend ex ante and ex post incentives.233 The high agency costs of an ex post system may tend to reduce the returns to innovating, discouraging some inventors. We could think of an ex ante prize system, such as government research grants, as a second price we use to further fine-tune the ex post system. By offering startup capital free of outside investors, the grant helps restore positive externalities to the level they would achieve in the absence of financing costs.234