Mauricio Caceres Bravo

We develop a new approach to estimating flexible demand models with exogenous supply-side shocks. Our approach avoids conventional assumptions of exogenous product characteristics, putting no restrictions on product entry, despite using instrumental variables that incorporate characteristic variation. The proposed instruments are model-predicted responses of endogenous variables to the exogenous shocks, recentered to avoid bias from endogenous characteristics. We illustrate the approach in a series of Monte Carlo simulations.

 Inmates in the US are assigned to different government-run prisons to serve their sentences and can face highly heterogeneous environments. I study how being assigned to prisons with different levels of inmate misconduct affects their outcomes. Using data from the Pennsylvania Department of Corrections, I estimate the effects of prisons on inmate misconduct while incarcerated by controlling for a rich set of sentencing and assessment variables used to assign inmates to prisons. I test for bias in my estimates in two ways. First, I show balance across inmate demographics. Second, I leverage inmate transfers between prisons in a “movers” design to demonstrate that misconduct effects accurately reflect causal prison treatment effects. Being assigned to a prison in the highest vs. lowest decile of misconduct effects approximately doubles the inmate’s misconduct, increases additional months in prison by 9%, and increases prison reentry from serious crime by 11%. Overcrowding and the criminality of peers are predictive of misconduct effects. A policy that assigns 20% of new inmates to the prisons that most reduce misconduct can decrease these inmates’ misconduct by up to 40%, time in prison by 4%, and reentry from serious crime by 5%.