|
| |
| Sarah A. Frederick, James J. Jozefowicz and Zackary T. Nelson |
| |
| ''A dynamic panel data study of the unemployment-crime relationship: the case of Pennsylvania'' |
| ( 2016, Vol. 36 No.3 ) |
| |
| |
| The present study analyzes the unemployment-crime (U-C) relationship in Pennsylvania using a balanced panel data set
of the 67 counties over the period from 1990 to 2009. A dynamic panel data model is estimated by Generalized
Method of Moments to account for endogeneity, measurement error, heteroskedasticity, and serial correlation. This
estimation methodology overcomes several econometric problems ignored in previous analyses of the U-C relationship.
Explicitly accounting for the dynamics of crime isolates criminal inertia from potential criminal motivation effects. The
results suggest a statistically significant impact of previous criminal activity on future crimes, but a statistically
insignificant relationship between the unemployment rate and the crime rate. Although these results run counter to the
Cantor and Land (1985) hypothesis, they indicate that specifying a dynamic model of crime and addressing the
econometric shortcomings of OLS regression analysis may yield more precise results. |
| |
| |
| Keywords: crime, unemployment, panel data, Pennsylvania, county |
JEL: K4 - Legal Procedure, the Legal System, and Illegal Behavior: General
R1 - General Regional Economics (includes Regional Data) |
| |
| Manuscript Received : Aug 22 2015 | | Manuscript Accepted : Aug 03 2016 |
|
