NBER Summer Institute 2018: Innovation

Last week I was a discussant at the Innovation section of the 2018 NBER Summer Institute (full schedule here), which I highly recommend to scholars interested in the economics of innovation. The quality of the papers and the discussion was pretty uniformly high. There were a few examples of the insularity of economics, such as remarks about topics that "no one has studied" that have been studied by legal scholars, but I think this just illustrates the benefits of having scholars familiar with different literatures at disciplinary conferences.

Here are links and brief summaries of the innovation-related papers. (There was also a great panel discussion on gender and academic credit, which I might post about separately at some point.)

Janet Freilich, Prophetic Patents – This was the paper I was a discussant for, and I'll devote a stand-alone post to it soon, so for now I'll just note that it does a great job highlighting the problem of fictitious data in patents and demonstrating how this confuses scientists.

Joshua Krieger, Danielle Li & Dimitris Papanikolaou, Developing Novel Drugs – Using a new measure of a drug's novelty relative to already-approved drugs, the authors show that novel drugs are less likely to be approved by the FDA but, conditional on approval, generate higher private and social returns. Additionally, firms respond to an increased cashflow with investment in more novel drugs, suggesting that "on the margin, firms perceive novel drugs to be more valuable ex-ante investments, but that financial frictions may hinder their willingness to invest in these riskier candidates." Lots of clever empirics, though figuring out the welfare effects is difficult.

Yifei Mao & Jessie Jiaxu Wang, Labor Scarcity, Finance, and Innovation: Evidence from Antebellum America – Using the staggered passage of free backing laws across states from 1837 to 1860 and differences in labor scarcity between slave and free states, they find that greater access to finance "encouraged technological innovation that substituted for free labor, but discouraged technological innovation that substituted for slave labor." The economic historians in the audience had questions about some of the data sources.

Michaël Bikard & Matt Marx, Hubs as Lampposts: Academic Location and Firms’ Attention to Science – Interesting results on knowledge flows: Academic publications are more likely to be cited by firms when they emerge from "hubs"—geographic concentrations of patenting by firms in the same specialized technical field as the paper—even when the citing firm is not in the hub. Academics in hubs do not seem to engage in more applied research; rather, the increased attention to hub-based research seems driven by its higher quality and the use of hubs as search heuristics (in that firms are more likely to cite a hub-based paper than a "twin" paper making the simultaneous finding outside the hub).

Lily Fang, Josh Lerner, Chaopeng Wu & Qi Zhang, Corruption, Government Subsidies, and Innovation: Evidence from China – Allocation of R&D subsidies in China was influenced equally by firms' innovative capabilities and corruption (as measured by anomalously high "Entertainment and Travel Costs"), and the 2012 anti-corruption campaign was effective at reducing the influence of corruption and increasing the effect of subsidies on future innovation. Seems like an important input for the choice between government-set and market-set innovation incentives.

Michael J. Andrews, The Role of Universities in Local Invention: Evidence from the Establishment of U.S. Colleges – Comparing cities that received new colleges 1839-1954 with runner-up sites suggests that new colleges caused 40% more patents per year. This effect seems primarily driven by migration: establishment of other institutions such as prisons or mental institutions had a similar effect; colleges had no independent effect after controlling for population; and most patents are by migrants to the college county rather than alumni or faculty. I thought this was a great paper that represented a tremendous amount of work, though I'm not sure that it tells us anything about the role of colleges in promoting local innovation today—a lot has changed since 1954.

Martin Watzinger, Lukas Treber & Monika Schnitzer, Universities and Science-Based Innovation in the Private Sector – Another interesting paper about knowledge flows. Comparing newly hired professors for a German university with runner-up candidates shows the effect of professors on the local private sector: local firms start to cite the newly hired professor's articles more and local patents become more similar to her articles. The effect seems primarily driven by PhD graduates working in the private sector.

Enrico Berkes & Ruben Gaetani, Income Segregation and Rise of the Knowledge Economy – Using a novel instrument related to the network of patent citations, they find that "innovation intensity is responsible for 14% of the overall increase in urban segregation between 1990 and 2010." The relationship between innovation and inequality is a growing area of scholarly interest (see, e.g., the interesting work by Raj Chetty et al.), and this is an important contribution to the field.

Hugo Hopenhayn & Francesco Squintani, On the Direction of Innovation – Using a theoretical model and plausible assumptions, they show that "the competitive market allocates excessive innovative efforts into high returns areas." This result is different from standard rent dissipation through patent races and from the point that private and social value can be misaligned. Instead, their point is that there is differential rent dissipation stemming from externalities imposed by marginal entrants to others focused on the same research goal. There are three sources of bias: (1) a static misallocation to high-value areas that depends on the assumption that the elasticity of discovery (i.e., the probability of solving the problem) is decreasing with the number of researchers; (2) a dynamic misallocation due to the externality a successful researcher imposes on remaining researchers who must bear the costs of switching to different research lines (which increases with the number of researchers); and (3) a second dynamic misallocation caused by individual researchers not taking into account the future option value of an unsolved problem (which is higher for high-value problems). The authors also claim that IP regimes cannot be the solution "as they grant property rights over the solutions and not the original problems," though there are lots of ways in which patent rights are often granted over ideas rather than complete solutions—see, e.g., Janet Freilich's paper above, or my review of some of this literature here.

Eugenie Dugoua, International Environmental Agreements and Directed Technological Change: Evidence from the Ozone Regime – The Montreal Protocol drove a large increase in R&D on alternatives to ozone-depleting molecules. Careful empirics, though the result isn't that surprising, and I'm not sure that this tells us much about international environmental agreements more generally. As scholars such as Cass Sunstein have explained, Montreal was much more successful than the Kyoto Protocol for a host of institutional reasons.

Ruchir Agarwal, Talent Matters: Evidence from Mathematics – Performance on the International Mathematics Olympiad (IMO) is correlated with a number of outcomes related to mathematical success, but IMO participants from low- and middle-income countries have less mathematical success than equally talented participants from high-income countries. This reminded me of the work on Lost Einsteins, though I wasn't sure what to take away from the results.