Working Paper: NBER ID: w18731
Authors: Hope Corman; Naci H. Mocan
Abstract: This paper investigates the relationship between alcohol consumption, deterrence, and crime for New York City. We examine high-frequency time-series data from 1983 to 2001 for one specific location to examine the impacts of variations in both alcohol consumption and deterrence on seven "index" crimes. We tackle the endogeneity of arrests and the police force by exploiting the temporal independence of crime and deterrence in these high-frequency data, and we address the endogeneity of alcohol by using instrumental variables where alcohol sales are instrumented with city and state alcohol taxes and minimum drinking age. We find that alcohol consumption is positively related to assault, rape, and larceny crimes but not murder, robbery, burglary, or motor vehicle theft. We find strong deterrence for all crimes except assault and rape. Generally, deterrence effects are stronger than alcohol effects.
Keywords: Alcohol Consumption; Crime; Deterrence; New York City
JEL Codes: I10; K00
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| Alcohol consumption (L66) | Assault, Rape, Larceny crimes (K42) |
| Deterrence (arrests) (K42) | All crimes (except Assault, Rape) (K42) |
| Alcohol consumption (L66) | Murder, Robbery, Burglary, Motor vehicle theft (K42) |
| Lagged arrests (K40) | Current crime rates (K42) |