Skin in the Game (with Erich Battistin and Michele De Nadai)
In fantasy sports, participants manage virtual teams of real-life athletes under a shared set of rules. We focus on Fantasy Premier League (FPL), the world’s most popular online fantasy football game, to study if the race of athletes, proxied by their skin tone, affects manager decisions. This low-stakes, anonymous setting provides a unique environment to test for discrimination: fantasy managers form rosters through non-rivalrous choices in one global market with common player prices and budget constraints, facing minimal strategic disadvantages from potential biases. The app features an intuitive interface that prominently displays player photographs alongside a wealth of performance statistics. Simple and transparent rules make player productivity observable and quantifiable to managers. Using an administrative panel of 8 million managers worldwide, we develop a structural model of manager decision-making grounded in expected payoff maximization and incorporating manager-specific preferences, subjective beliefs about player performance, and collective demand factors. We estimate the model using data on manager choices, player performance, and player skin tone, which we measure from a machine learning algorithm that converts photographs into quantifiable skin color classifications.
Neighborhood change and local economic activity (with Hector Blanco)
Draft available upon request.
The impact of gentrification on the economic opportunities of local residents is still unclear. We study how urban renewal programs affect the distribution of local firms and employment opportunities. We use the ‘regeneration’ of public housing developments in London (UK) into mixed-income housing as a natural experiment that increased the proportion of high-income households living in the affected neighborhoods. Over a nine-year period, we show that regenerations resulted in a spatial reshuffling of firms located within 1.2km from the site, without affecting their total number. Local firms experience an increase in employment of about 2-4%. Such increase is temporary, however, suggesting that housing regenerations did not generate persistent employment opportunities for incumbent residents.
Closing the Books: Public Library Closures and Adult Cognitive and Health Outcomes (with Agnese Romiti)
Libraries serve as learning hubs and gathering spaces. As several countries experience rising social isolation and political disengagement, understanding libraries' role is essential for public policy, especially in an ageing society where isolation impacts cognitive health for the elderly. Using a Difference-in-Differences design, we study how public library closures in England affect cognitive and health outcomes of adult individuals. In provisional results, we document that adult individuals experiencing a library closure in their neighbourhood exhibit lower cognitive and health outcomes following the closure.
The Effects of Temporary Confiscation of Vacant Housing (with Josep Amer-Mestre)
High vacancy rates in tight housing markets are perceived by policymakers and society at large as a market failure that calls for government intervention. We study the effect of a policy passed in the Balearic Islands (Spain) that dictated the temporary confiscation of vacant properties, unless they are put on the market. Using a synthetic difference-in-differences strategy, we show that following the policy announcement, house prices increase while rents temporarily decrease before reverting back to pre-policy levels. This suggests that while supply effects dominate in the rental market, the sale market experiences an increase in prices. This is potentially due to a combination of an 'amenity effect' - i.e., the removal of the disamenity imposed by vacant properties on nearby housing - and a regulation effect, whereby imposing an implicit tax on vacant housing dampens private housing investments.
Price responses to affordable housing under mandatory inclusionary zoning (with Hector Blanco)
Many housing assistance programs incentivize policies that involve a combination of affordable and market-rate units in the same building, i.e., mixed-income housing. Although prior research suggests that households in the private market are willing to pay to live in areas with higher-income neighbors, little is known about the magnitude of these preferences within a given building. This paper estimates the price response of market-rate units in mixed-income buildings to the share of affordable housing. We exploit variation from London’s mandatory inclusionary zoning program, under which new residential buildings are required to maintain a minimum amount of affordable housing set by each of the 33 city boroughs and that can vary on a case-by-case basis. Using an instrumental variables approach that leverages the discretionary part of the decision process, our preliminary results suggest that an increase in affordable housing in a mixed-income building significantly decreases the sale and rental prices of market-rate units in the same building.
Disruption in the Classroom
Many countries support the education of children with learning and behavioural difficulties in mainstream schools. However, there is little evidence on the effects of such inclusion policies on the academic achievement of targeted children and their classmates. I address this question by considering public schools in England, where a Maimonides' rule for class size formation yields quasi-experimental variation in the number of students with severe special educational needs (SEN). Instrumental variables estimates show no sizeable impact of the number of SEN students on the academic performance of their classmates. This result is not driven by additional resources that may be carried by SEN students, such as larger funding or additional teachers.