Science of Science Day.

The NBER's Program on Productivity held a "Science of Science Day" when it met on March 14. Program Director Ernst R. Berndt and NBER Research Associate Pierre Azoulay, both of MIT, organized the meeting. These papers were discussed:

Shulamit Kahn, Boston University, and Megan J. MacGarvie, Boston University and NBER, "How Important is Location for Research Productivity in Science?"

Discussant: Richard B. Freeman, Harvard University and NBER

Pierre Azoulay; Joshua Graft Zivin, Columbia University and NBER; and Jialan Wang, MIT, "Superstar Extinction"

Discussant: Lynne G. Zucker, University of California, Los Angeles and NBER

Liam Brunt, University of Lausanne; Josh Lerner, Harvard University and NBER; and Tom Nicholas, Harvard University, "Inducement Prize and Innovation"

Discussant: Michael Kremer, Harvard University and NBER

Laure Turner, CREST-ENSAE; and Jacques Mairesse, CREST-ENSAE and NBER, "Individual Productivity Differences in Scientific Research: An Econometric Exploration of Publications by French Physicists"

Discussant: lain Cockburn, Boston University and NBER

David Popp, Syracuse University and NBER, and Richard Newell, Duke University, "Where Does Energy R and D Come From? A First Look at Crowding Out from Environmentally-Friendly R and D"

Discussant: Rebecca Henderson, MIT and NBER

Kahn and MacGarvie ask whether scientists located outside the United States are at a disadvantage when it comes to research productivity, collaboration, and knowledge diffusion. The principal difficulties of comparing scientists inside the United States with those outside the United States arise from unobserved heterogeneity among scientists and the endogeneity of location choices. This paper uses a new and unique dataset of foreign-born U.S.-educated scientists that allows its authors to exploit exogenous variation in post-Ph.D, location induced by visa status. They thus are able to compare students who were required by law to leave the United States upon the completion of their studies with similar students who were allowed to remain in the United States. The researchers assess whether students who left the United States have more or fewer publications, patents, citations, and collaborators when compared with a control student with the same advisor. They also ask whether these students have more or fewer international collaborations.

Azoulay and his co-authors estimate the magnitude of spillovers generated by 137 academic "superstars" in the life sciences onto their coauthors' research productivity. These researchers died while still being actively engaged in science, thus providing an exogenous source of variation in the structure of their collaborators' coauthorship networks. Following the death of a superstar, coauthors suffer a lasting 8 to 18 percent decline in their quality-adjusted publication output. These findings are surprisingly homogenous across a wide range of coauthor and coauthor/superstar dyad characteristics. Together, they suggest that part of the scientific field embodied in the "invisible college" of coauthors working in that area dies along with the star--a genuine and irreplaceable loss of human capital.

Brunt and his co-authors examine prizes as an inducement for innovation, using a novel dataset of awards for inventiveness offered by the Royal Agricultural Society of England from 1839 to 1939. At annual shows, the RASE held competitive trials and awarded medals and monetary prizes (exceeding one million pounds in current prices) to spur technological development. This paper uncovers large effects of the prizes on contest entries, especially for the Society's gold medal. Matching award and patent data, the authors also detect large effects of the prizes on the quality of contemporaneous inventions. These results hold even during the period when prize categories were determined by a strict rotation scheme, thus overcoming the potential confound that awards were offered in "hot" technology sectors. The evidence suggests that prize awards can be a powerful mechanism for encouraging competition and that prestigious non-pecuniary prizes can be a particularly effective inducement for innovation.

An empirical regularity has often been observed in economics of science: productivity differences among researchers are extremely large and persistent, and a prolific minority of scientists produces most of the publications and accounts for most of citations in almost all research fields. Turner and Mairesse investigate to what extent such dispersion and persistence can be accounted for by three types of factors in so far as they can measure them: individual variables, mainly age and gender, career stage variables, and laboratory variables. Does individual scientific productivity significantly drop off as scientists become older and more or less advanced in their careers? Is it strongly related or not to career promotions and to the productivity and quality of the laboratories in which scientists work? Even if these factors prove quite significant, is it nonetheless the case that individual productivity differences remain largely unaccounted by them, and that they have to be mostly imputed to unobserved individual circumstances and characteristics, or so called "individual effects"? To answer such questions, the authors have put together a 12- year panel database for 465 condensed matter physicists working in the French public research organization CNRS, and have specified and estimated a simple econometric model for both "productivity" and "quality" measured respectively by the number of publications per scientist per year and by the corresponding average citation impact of these publications.

Recent efforts to endogenize technological change in climate policy models demonstrate the importance of accounting for the opportunity cost of climate R and D investments. Because the social returns to R and D investments are typically higher than the social returns to other types of investment, any new climate mitigation R and D that comes at the expense of other R and D investment may dampen the overall gains from induced technological change. Unfortunately, there has been little empirical work to guide modelers as to the potential magnitude of such crowding out effects. Popp and Newell attempt to address this question. They consider the private opportunity costs of climate R and D, asking whether an increase in climate R and D represents new R and D spending, or whether some (or all) of the additional climate R and D comes at the expense of other R and D. They begin at the industry level, using sectoral data on R and D expenditures to ask whether increases in energy R and D spending crowd out other R and D spending in various industries. Preliminary results show some evidence of crowding out in sectors active in energy R and D, but not in sectors that do not perform energy R and D. This suggests that funds for energy R and D do not come from other sectors, but may come from a redistribution of research funds in sectors that are likely to perform energy R and D. Given this, the authors proceed with a detailed look at climate R and D in two sectors: automotive manufacturing and alternative energy. Linking patent data and financial data by firm, they examine patenting behavior of large firms. For these firms, they are able to separately identify patents pertaining to innovative energy technology (for example, wind energy and fuel cells) from all other patents. They ask whether an increase in these energy patents leads to a decrease in other types of patenting activity, which would suggest R and D crowding out. Preliminary results show no evidence of crowding out within these firms. Finally, the authors use patent citation data to ask where the most valuable patents in these sectors come from. They find that firms specializing in alternative energy research provide more valuable patents than larger firms for which alternative energy R and D is just part of a broader portfolio.