JOB MARKET PAPER
Research on the effects of police presence tends to focus on the impact such policies have on crime rates. Less is known about how much individuals value place-based policing strategies. This paper fills this gap by estimating the willingness-to-pay (WTP) to avoid crime using housing market data in the three largest cities of Colombia. Specifically, I study the effects of 100 newly constructed police stations on crime and property values using an instrumented difference-in-differences design. I find that police deter violent crime by 12 percent and property crime by 22 percent in the immediate vicinity of the newly constructed police stations, with no crime displacement nearby. Using a hedonic pricing model, I find that the opening of new police stations leads to a 5 percent increase in property values —a gain of $3.5 million for households directly affected. While hedonic regressions identify the effect of crime on housing values for the marginal buyer, I estimate a correlated random coefficients model and show that the welfare effects of crime are homogeneously distributed in the population. I conclude that the average marginal WTP to avoid crime due to the local effects of the intervention is $4,500 per household. These results suggest that cities under-provide policing and target high crime neighborhoods while the benefits are widespread.
WORK IN PROGRESS
Police Militarization, Public Opinion and Willingness to Pay to Avoid Crime (with Justin Holz and Taeho Kim)
This paper studies the effect of police militarization on crime, public opinion and individual welfare. Previous studies of the causal effects of providing military equipment to local police on crime have shown mixed results in the United States. We investigate the broader consequences of local police acquiring military-grade equipment that goes beyond deterring crime. We start estimating the effect of police militarization on social media usage from Twitter as well as data on the Black Lives Matter (BLM) movement. We complement this analysis with a welfare calculation in terms of the willingness-to-pay (WTP) to avoid crime. To obtain our estimates, we pull together county-year panel data on spending on military equipment from the 1033 program, reported crime, social media usage, and housing values. We study how the acquisition of tactical weapons, optics, and vehicles affects public opinion and engagement in protests. Then, using the hedonic pricing method, we calculate the average WTP to avoid crime using variation in housing values. For causal identification, we use temporal variations in equipment availability and between-counties variations in the likelihood of receiving a positive amount of military aid to build a shift-share instrument. Then, we explore the effects of the variation of the acquisition of military equipment on public opinion data and housing values.
Location Sorting, Consideration Sets, and Sparsity under Household Preferences Heterogeneity (with Olivia Bordeu and Santiago Franco)
We adapt recent developments on dimension reduction in structural economic models to urban location sorting frameworks. We propose a modified version of the two-step group fixed-effects estimator of Bonhomme, Lamadon, and Manresa (2019) to estimate a household location problem within a city. Our estimator differs from the existing methods in two ways. First, we use household characteristics to define the household consideration set. This reduces the dimension of this set and allows us to incorporate heterogeneous budget constraints across households. Second, our method allows for rich non-linear heterogeneity in household preferences under sparsity. Preliminary estimations using rich Colombian micro-data show that our method delivers a better out of sample fit with respect to standard discrete choice models of residential location.
Forced Migration and Social Tensions: The Syrian Civil War Spillovers in Lebanon (with Kara Ross Camarena and Nils Hägerdal)
Do rebel groups use violence strategically during elections? The Role of Natural Resources