Assessing the benefits of robbery reduction: The case of a large Brazilian city

Abstract

We estimate willingness to pay (WTP) for a first-order stochastic reduction on the risk of robbery at a large city in Brazil. We develop a structural choice model that nests a process of contingent valuation among non-degenerate lotteries and estimate it by both maximum likelihood and geographically weighted regression (GWR) using a dataset from the city of Fortaleza that contains a precise measurement of individual's subjective probability of victimization among detailed socioeconomic measurement, willingness to pay, and police presence variables. Expected loss, gender, age, education, and perception of patrolling explain WTP. Our global model estimated a mean WTP of R$ 19.23 (U$ 10.33) per month. Our local model, estimated by GWR, suggests that there is a reasonable amount of spatial heterogeneity that follows the city's socioeconomic spatial distribution profile. Although the city's northwest periphery presents higher WTP, as long as we go inwards, there is plenty of heterogeneity on its spatial distribution. Our results support a theory of crime with an active role for victim's (costly) precautions influenced by socioeconomic spatial heterogeneity.