Publications
Income and Democracy: Evidence from System GMM Estimates, with Benedikt Heid and Mario Larch. Economic Letters 116(2): 166-169, 2013.
Work in Progress
Justice Study: The Effects of Short-Term Incarceration on a Human Life, with Andreas Beerli, April Faith-Slaker, James Greiner, and Michel Maréchal
Incarcerated populations have skyrocketed in most parts of the world over the past decades. Rigorous scientific evidence on the effects of incarceration is scant. Identifying the causal effects of incarceration has proven challenging because of the many unobservable external factors (such as pre-existing differences in personality and criminal attitudes) that threaten the validity of a comparison between incarcerated and non-incarcerated people.
We study the psychological, social, economic, and health consequences of short-term incarceration using a randomized controlled trial (RCT). We build on the US cash bail system, which in theory provides a financial incentive for court appearance but in practice may impose a high burden for subjects with lower socio-economic status. The study offers recently arrested people who were assigned bail they most likely can’t pay themselves, and who want to participate in the study, a 50-50 chance of being placed in an “Extra-Chance” group or a “Regular Pre-Trial” group. For people randomized to the Extra-Chance group, a local non-profit organization posts their bail, so that they can be free from jail while they await their trial. For people randomized to the Regular Pre-Trial group, the local non-profit does not post their bail, but participants can still be released if they can post their own bail, if someone else posts their bail for them, or if the judge removes their bail. In other words, people in the Regular Pre-Trial group experience the criminal justice system as though there were no study. We will use administrative data sources and administer surveys to study the impact of being released from pre-trial detention on participant’s lives.
Narratives in the Great Depression
Did stories affect the severity and duration of the Great Depression? Recent research suggests that economic booms and busts cannot be fully explained by economic fundamentals but are related to the prevalence and vividness of certain stories. This project tries to empirically test this claim. I hand-collect a data set of 300.000 title and editorial pages from 150 national daily newspapers in the United States for the years 1929-1933. Exploiting methods from deep learning and computer vision, I am able to create the first large-scale news and commentary database for the Great Depression. I then use methods from natural language processing and media monitoring to cluster these news into larger stories to answer three questions. First, exploiting the historical segmentation of media markets and the comprehensive nature of the database, I analyze which stories go viral and why they do so. Secondly, I use daily data on bank suspensions to analyze which stories are related to big shifts in people’s confidence. Third, conditioning on the specific story cluster, I analyze whether information or interpretation matter more for these confidence shifts.
Financial Scandals and Extremism: Evidence from the Dreyfus Affair, with Joachim Voth and Noam Yuchtman
Technology Transfer and Economic Growth, with Raffaele Blasone and Bruno Caprettini
Reason in Revolt? – Enlightenment and the French Revolution
Did the Enlightenment bring about the French Revolution? Starting from 1789, historians and commentators have argued that the political revolution was preceded by a revolution of the mind. In this project, I provide the first quantitative test of this hypothesis. To this purpose, I construct a novel city-level database of measures for the presence of Enlightenment knowledge elites and revolutionary ideology as well as a new database of deputies’ biographies, roll-call votes and speeches for both the early years of revolution (1789-1792) and the Reign of Terror (1792-1794). Preliminary results suggest that cities with a larger presence of knowledge elites demanded more radical state reforms in the Cahiers de Doléances, have a higher fraction of revolutionary first names, are more likely to change their name, and are more likely to have a Jacobin club in the early years of the French revolution. Turning to policy-makers, deputies from cities with a high presence of knowledge elites employ more radical political rhetoric and are more likely to vote for the execution of the king.