Fabian Sinn
I am a PhD student in economics at the Institute for International Economic Studies (IIES) in Stockholm.
My research focuses on questions in labor, health, and public economics.
CV Email: fabian.sinn@iies.su.se
Working Papers
Does who we work with matter? The Effect of Coworkers on Income
I study the role of coworkers by estimating an individual-level measure of coworker quality for the universe of workers in Sweden. High-quality coworkers increase the earnings of the individuals they work with. I develop a novel method combining penalized estimates with group fixed effects in an AKM framework to identify individual and coworker quality by observing workers' movement across plants. Coworker quality explains around 10% of the variations in individuals' earnings or around ⅔ of the variation otherwise explained by firms. This finding has positive and normative implications. In an application, I estimate the social returns to education by coupling my measure of coworker quality with an instrumental variable strategy, which exploits the exogenous variation in the distance to the opening of new colleges. I find that receiving a college degree boosts the earnings of coworkers by approximately 8%, implying that the social returns to a college degree are 20% higher than implied only by individual returns. This finding implies a substantial externality of college education and suggests individuals underinvest in their human capital.
HERSTORY: The rise of self-made women with Arash Nekoei
We document the evolution of women's status across the globe and throughout recorded history. We first construct a new database of seven million notable individuals (Human Biographical Record). We then measure women's status as women's share among the most prominent fraction of population that allows comparison across time and space. The records show no long-run trend in women's share in recorded history. Historically, women's power has been a side-effect of nepotism: the more important family connections, the higher the women's share. But self-made women began to rise among the writers in the 17th century before a broader take off started with the 1800 birth cohort: first among artists and scholars, followed by elected politicians, and finally appointed politicians. The first wave among writers emerged when informal humanist education and new public spheres shaped a supply of literary women, who met the demand of a new female reading public. A strong writer wave predicts a stronger takeoff of self-made women in the 19th century. This effect has persisted and created cross-country divergence.
Social Inclusion: Definition and Measurement, with Arash Nekoei
We define social inclusion (SI) as the equality of success rate conditional on circumstances, i.e., inherent characteristics and family background. We measure SI using an aggregator, social inclusion function, e.g., mutual information or the share of success variance unexplained by circumstances. Another interpretation of SI is the unpredictability of success using circumstances. We discuss the role of out-of-sample prediction and machine learning in estimating SI, which we then illustrate in three complementary applications. The first two applications document global SI using a small set of circumstances: the inclusiveness of the elite in long-run history and contemporary SI of education. The first application measures the inclusiveness of the elite in the long-run history and reveals the exceptional high SI in the 20th century. The second application estimates contemporary SI of literacy across 68 countries and shows that country of birth is an obstacle to SI as much as family background. The third application estimate SI in Sweden today using a rich set of circumstances from administrative data. In order: Gender, nationality of parents, and their income/education restrict inclusion whereas the place of birth plays almost no role.
Human Biographical Record (HBR), with Arash Nekoei
We construct a new dataset of more than seven million notable individuals across recorded human history, the Human Biographical Record (HBR). With Wikidata as the backbone, HBR adds further information from various digital sources, including Wikipedia in all 292 languages. Machine learning and text analysis combine the sources and extract information on date and place of birth and death, gender, occupation, education, and family background. This paper discusses HBR's construction and its completeness, coverage, accuracy, and also its strength and weakness relative to prior datasets. HBR is the first part of a larger project, the human record project that we briefly introduce.