Publications
Measuring Science: Performance Metrics and the Allocation of Talent
(with Carlo Schwarz and Fabian Waldinger)
American Economic Review, 2024, 114 (12): 4052–90.
[Abstract]
We study how performance metrics affect the allocation of talent by exploiting the introduction of the first citation database in science. For technical reasons, it only covered citations from certain journals and years, creating quasi-random variation: some citations became visible, while others remained invisible. We identify the effects of citation metrics by comparing the predictiveness of visible to invisible citations. Citation metrics increased assortative matching between scientists and departments by reducing information frictions over geographic and intellectual distance. Highly cited scientists from lower-ranked departments (“hidden stars”) and from minorities benefited more. Citation metrics also affected promotions and NSF grants, suggesting Matthew effects.
Working papers
Dictators, Democracies, and Discoveries: Political Institutions and the Creation of Knowledge
(with Felix Radde, Carlo Schwarz and Fabian Waldinger)
[Abstract]
We investigate the role of political institutions in shaping global knowledge production. Using a newly assembled global dataset covering universities, scientists, publications, Nobel Prize-winning discoveries, and technological breakthroughs from 1900 to the present, we show that institutional quality is a key determinant of scientific production. Countries with high-quality institutions build larger academic sectors and produce more scientific output, including frontier research and Nobel Prize-winning discoveries. Event-study evidence from sharp institutional improvements and deteriorations confirms that changes in political institutions lead to sustained shifts in scientific production. These effects operate through both a larger research workforce and higher productivity per researcher. Institutions also shape the breadth of inquiry: autocracies concentrate research in a narrower set of fields, achieving excellence in some areas but lacking the broad exploration of ideas that characterizes democracies. Finally, countries with better institutions not only produce more published research but also generate more path-breaking technological discoveries, including those not published in academic journals.
Can Know-How Be Transferred? Evidence from the Allied Investigation of German Industry after World War II
(with Marcel Schlepper and Timo Wochner)
[Abstract]
We study whether and how large-scale knowledge transfers can shape innovation. After the Second World War, the United Kingdom and the United States engaged in a large-scale program of “intellectual reparations.” British and U.S. experts investigated firms and production plants, copied recipes and drawings, and wrote reports that were distributed publicly. We build the first systematic database of these investigations from about 150,000 pages of reports held at the Deutsches Museum and the Imperial War Museum. It covers 3,873 reports and more than 20,000 visits to German firms. We establish novel facts on the program by documenting which German firms, industries, and technologies were investigated. Moreover, we link the investigated German firms to the investigating U.S. firms to directly observe knowledge transfers. In current work in progress, we estimate the effect of these reports and investigations on U.S. innovation.
Speaking of Gender: Superiors’ Gender Attitudes and Women’s Careers
[Abstract]
This paper investigates superiors’ revealed preference for diversity. I study the effect of ministers who differ in their gender attitudes on the career trajectories of women in German ministries. I develop a novel measure of gender attitudes: the extent to which a person speaks gender-inclusively. Based on thirty years of parliamentary speeches, I construct a politician-level score measuring the degree of gender-inclusive language. I calculate each minister’s score and relate it to self-collected employee-level data from German ministries. Superiors’ gender attitudes matter for women’s career outcomes: ministers with one standard deviation higher gender attitudes increase women’s probability of being promoted. This effect is driven by both male and female ministers.
Work in progress
Leadership, Language and Loyalty: Ethnic Diversity in the Austro-Hungarian Army
(with Emilio Esguerra and Eric Golson)
[Abstract]
We ask whether commanders can lead soldiers of a different ethnicity and language as effectively as soldiers who share their own. We study the multi-ethnic Austro-Hungarian army in World War I to ask whether ethnic and linguistic divisions in command weaken unit cohesion, increase losses and erode loyalty. Based on newly digitzed data on the organizational structure of the army, we analyze changes of commander within a regiment . Comparing ethnic groups within the same unit at the same time allows us to estimate whether sharing the commander’s ethnicity lowers an ethnic group’s share of the dead and wounded. Beyond losses, we examine loyalty, measured by surrender and desertion. We test whether in-group commanders reduce these defections and whether this effect weakened as the empire disintegrated in 1917–18.
Following the Citations: Performance Metrics and the Direction of Scientific Research
(with Carlo Schwarz and Fabian Waldinger)
[Abstract]
We study how performance metrics affect the direction of scientific research. The introduction of the first citation database in the 1960s made it possible, for the first time, to see which topics attracted more citations than others. This may have given scientists an incentive to be more strategic in choosing what to work on. We examine whether scientists herded into highly cited topics, and whether this response was stronger for early-career scientists and others facing steep career incentives. We also study whether the topics scientists moved into went on to have higher long-run impact, or whether herding pushed effort into already crowded areas.