Research

How can a nation become a global leader in science and technology? Investment in education can boost human capital, which is crucial for economic development. This project examines whether large-scale education reforms, exemplified by the Prussian education reforms of the early 19th century, can cultivate upper-tail human capital (advanced skills and knowledge). After its defeat by Napoleon, Prussia initiated nation-wide reforms to modernize the state. A core component of these reforms were education reforms that focused on natural sciences and technology. We exploit natural variation in cities’ exposure to the education reforms, comparing Prussian cities affected by the reforms with cities in other German-speaking states that were unaffected. We approximate upper-tail human capital by analyzing the births and deaths of prominent scientists at the city level. Using a Synthetic Control empirical framework, we estimate the effects of the reforms on upper-tail human capital formation. Our project contributes to two areas of economic literature: the origins of (upper-tail) human capital and the study of notable individuals. To the best of our knowledge, we are the first to systematically explore the causal impact of national education reforms on upper-tail human capital formation, focusing on Prussia’s experience. This study sheds light on the origins of the highly skilled population that contributed to Germany’s scientific and industrial success in the late 19th century.

This paper investigates the impact of scientific break-throughs on scientist productivity by examining the unexpected arrival of quantum mechanics between 1918 and 1932. Using historical correspondence data of quantum physicists, I construct a social network to proxy knowledge flow and measure scientists' exposure to the breakthrough based on their social proximity to key founders of quantum mechanics. I complement the social network data with publication and citation data of scientists from Clarivate’s Web of Science. A difference-in-differences with inverse propensity score weighting quantifies the effect of exposure to the scientific revolution. Results show that scientists with greater exposure to quantum mechanics ideas –measured by network distance to founders – published more papers and received more citations post-1925. The identification strategy leverages the unexpected and rapid development of quantum mechanics, which was unpredictable even to its founders, ensuring quasi-random variation in exposure. The findings highlight the role of social networks in disseminating break-through ideas and their causal effect on individual productivity.

This paper reconstructs and analyzes the social structure of physics in the first half of the twentieth century using the Inventory of Sources for History of Twentieth-Century Physics (ISHTCP), a large archival index of scientific correspondence. I digitize more than 12,600 pages of microfilmed records covering approximately 5,600 physicists active between 1896 and 1952 and convert them into a longitudinal correspondence network. Using period-specific graphs, centrality and community measures, and the graph Laplacian, I describe changes in the network’s size, connectivity, centralization, structure, and most prominent actors. I then compare the correspondence network with complementary networks based on nominator–nominee relationships, institutional and geographic affiliations, conference participation, coauthorship, and cocitation in seven leading physics journals. Analytical and visual measures of similarity and overlap assess whether these networks capture the same relationships, communities, and central individuals. Finally, I address measurement error arising both from incompleteness and inaccuracies in the original inventory and from vision-language-model-assisted digitization, documenting the extraction and validation procedures and testing the sensitivity of key network statistics to uncertain, missing, and misclassified ties. The resulting multilayer analysis provides a large-scale, uncertainty-aware account of how communication, collaboration, proximity, professional encounters, and scientific recognition intersected and evolved in twentieth-century physics.

This paper investigates physicists’ connections beyond their discipline during the first half of the twentieth century using the Inventory of Sources for History of Twentieth-Century Physics (ISHTCP), a large archival index of scientific correspondence. I digitize records of letters sent and received by approximately 5,600 physicists active between 1896 and 1952 across the major fields of pure physics, focusing on their ties to a broader pool of about 75,000 correspondents and identifying interlocutors outside physics. I construct weighted, time-varying networks from the numbers of letters exchanged during successive intervals and classify non-physicist correspondents by professional and social domain. I analyze the prevalence, intensity, and evolution of these ties, differences across fields and individuals, and the physicists and non-physicists who connected otherwise separate communities. Rather than treating correspondence solely as evidence of scientific collaboration, I use it to map physicists’ relationships with political, cultural, artistic, literary, industrial, and other social worlds. The study provides a systematic account of the breadth and structure of physicists’ extra-disciplinary relationships and examines how professional scientific life intersected with wider social affiliations and interests.