Prizes Are an Incentive to Innovate – Work Well

We have examined one of the longest available datasets of awards for innovation to determine whether prizes spurred technological development. We find that prizes induced competitive entry and that the largest effects are for prestigious medals. Consistent with competitive entry, we find important effects of the prizes on counts of quality-adjusted patents, which cannot be explained by technology category substitution. Our quantitative evidence on the utility of prizes is supported qualitatively. The Scientific American remarked of the RASE prize system in 1867: ‘It is indisputable that these

[Table Omitted]

competitive trials have done, and are doing, much to raise agricultural engineering to the highest standards of efficiency and economy.’ With respect to steam engines, which had the largest impact on productivity growth of any technology in the mid-to-late nineteenth century (Crafts [2004]), the role of the RASE was again noted by the Scientific American in 1874: ‘An investigation of the results obtained from year to year shows a most extraordinary improvement in the engines, as regards economy and workmanship, and there is little doubt that the effect of these tests has been most beneficial to the users of steam power.’ An 1864 report by the Society of Arts noted: ‘Without the prize system, the manufacturers would not have been guided to the production of the class of implements really required.’

Equally, we are aware of caveats to our findings. We cannot determine how much of the boost to patents we observe is driven by shifts in the propensity to patent as inventors sought to avoid expropriation risk as a consequence of the RASE’s offering prizes and attracting inventors to the technology target areas. Furthermore, despite our best efforts to measure the effects of technology category switching, it is possible that some inventors may have strategically delayed technological development to synchronize their inventive efforts with the prizes, especially during the triennial rotation period when the technology categories eligible for awards could be predicted. This form of temporal substitution would upward bias our results if inventors delayed patenting to benefit from the advertising value of the prizes and maximize the returns from supracompetitive pricing during the patent term. The effects we observe may also be downward biased. We do not observe unpatented inventions that may also have been stimulated by the prizes, and our estimates do not include patents induced by the prizes that were filed in years other than the show year. If prizeinduced patented or unpatented inventions generated sequences of cumulative innovations, then the downward bias in our estimates will be large.

Our historical evidence on RASE prizes offers guidance for the design of current inducement prize contests. Mega prizes, such as those offered by the X-Prize Foundation, presuppose that inventors are incentivized by large pecuniary inducements, but R&D costs typically exceed the value of the prize. For example, 26 teams competed for the X-Prize for suborbital spaceflight and collectively spent in excess of $100 million for a ten million dollar prize. Our evidence suggests that non-pecuniary prizes can be particularly effective. They avoid the complex process of linking the magnitude of the prize to the value of a particular, technology and inventors are still able to appropriate by winning. The RASE contests offered free publicity and public approbation. Inventors could benefit from the seal of quality ascribed to the invention when selling or licensing their technologies. The RASE lowered administrative costs by using medals rather than financial awards.

One explanation for why the financial awards to inventors were relatively small (around one-third of the sale value of an invention according to Figure 2) is that the RASE prizes were complementary to patents. Intellectual property rights provided incentives for inventors to invest in useful knowledge because they could appropriate through proprietary pricing, which was augmented by the effects of the prizes. Although this means that the prizes may have magnified deadweight losses, they also realized benefits that the patent system could not. In particular, prizes facilitated the diffusion of technical information further than the disclosures required by patenting. Our evidence suggests that in agricultural technologies, the prizes encouraged innovation beyond the patent system alone. (690 LIAM BRUNT, JOSH LERNER AND TOM NICHOLAS)

© 2012 Blackwell Publishing Ltd and the Editorial Board of The Journal of Industrial Economics

Given the imperfections associated with patents, the literature on innovation incentives has attempted to evaluate the use of alternative, or complementary, mechanisms such as prizes. The theoretical literature is welldeveloped in this area, but empirical work has been lacking. This is particularly problematic because uncertainty about the cost-benefit tradeoff associated with prizes acts as a major barrier to changing innovation promoting policies (Kremer [1998], pp. 1162–1165; NRC [2007]). Insofar as policy changes require supporting empirical evidence, our findings suggest that inducement prizes for innovation can work.