Alcohol availability and crime: a robust approach

Abstract

The relationship between alcohol availability and crime is investigated in this study. It first considers common parametric specifications that have been used in the literature. After applying a powerful consistent conditional moment test for correct specification, it is found that these common parametric specifications are rejected by the data. The study then proceeds with a robust nonparametric method that can have a rate of convergence close to that for a correctly specified parametric model when the underlying relationship is somewhat linear. The application of nonparametric methods reveals structure present in the data that would remain undetected when applying common parametric specifications, but more importantly reveals that the impact of alcohol availability is considerably higher than one might believe on the basis of the misspecified parametric model. It is also found that the marginal effect of alcohol availability on crime changes with the level of alcohol availability.

Acknowledgements

We thank Frank Chaloupka, Philip Cook, Michael Grossman, Steven Levitt, Sara Markowitz, and Samuel Myers for comments on an earlier draft. We alone are solely responsible for any remaining errors.

Notes

¹ This study uses the terms ‘supply of offense function’, ‘crime equations’, ‘crime generating function’ and ‘supply of crime equations’, interchangeably.

² The model is not developed here for space considerations. Readers are referred to Gyimah-Brempong (2001) for further details.

³ It is noted that using the unemployment rate or poverty rate does not affect the results.

⁴ We tried to obtain data on police patrol patterns at the census tract level from the Detroit Police Department without success.

⁵ The inclusion of racial minorities does not imply that race per se causes crime. Race is only a proxy for some unobserved variables that may be highly correlated with race. For more on the correlation between race and crime, see Gyimah-Brempong (1997).

⁶ Other socioeconomic variables that have been included in crime generation equations are the unemployment rate, poverty rate, and per cent on public assistance. These variables are, however, all highly correlated with income and education. In the interest of parsimony, we do not include these variables in our model.

⁷ Those interested mainly in the resulting estimates may skip this section and proceed directly to Section 4.

⁸ For other estimators that have similar properties, see Racine and Li (2004) and Li and Racine (2003).

⁹ In the interest of brevity the details are not presented; the interested reader is referred to Li and Racine (2004).

¹⁰ For the sake of brevity, no attempt is made to summarize all marginal effects in this paper; however, all results are available upon request from the authors.

¹¹ The elasticities are calculated at the means of the variables.

¹² It is noted that the estimates are not directly comparable to those of Gyimah-Brempong (2001) since he used an IV estimator and his equations were estimated in log-linear form.

¹³ Details of this test are presented in Appendix A.

¹⁴ There is no adjustment for the number of explanatory variables in the nonparametric R ² figures. Hence, strictly speaking, they are comparable only to the unadjusted parametric figures. However, the differences between the adjusted and unadjusted parametric figures are negligible.