|
Abstract
AIDS and crime rates are often treated in isolation from each other. This study, the first to empirically test whether HIV prevalence could influence crime rates, employs an interdisciplinary model of crime to fit the complex socio-demographic context of South Africa. I construct a provincial Crime-AIDS panel dataset for the period 1994 to 2003 and use different econometric models (OLS, Fixed-Effects, Instrumental Variables and a combined I.V plus F.E specification) to test the short-run relationship between HIV prevalence and crime rates in South Africa. The fixed-effects specifications are found to best capture the relationship between crime and HIV prevalence in South Africa. I find strong evidence that HIV prevalence is positively correlated with some types of monetary/property-related crimes. This study also assesses some of the behavioral mechanisms that could drive this positive relationship between AIDS and crime. I find evidence that those infected with HIV are also more likely to engage in other risky behaviors such as illegal drug use. Those infected with HIV may be more likely to engage in risky behaviors such as drug use and crime because their reduced life expectancy leads them to discount the future. I find no evidence of HIV/AIDS crowding-out police/law enforcement expenditure in favor of other government agency spending for the study period. I find weak evidence of a negative correlation between HIV/AIDS and some types of crime that are of a violent nature, possibly because HIV positive people become too sick to commit these types of crime. This study has important policy implications with regard to the planning, coordination, budgeting and execution of programs aimed at reducing crime and AIDS in South Africa. For example, a direct implication of the results presented in this study is that policies that decrease the incidence of HIV/AIDS, or increase the life expectancy of those with HIV/AIDS, will also contribute to reducing some types of property-related crimes.
Footnote:
-
This paper is part of a more comprehensive doctoral dissertation of the same title, recently submitted to the South African Journal of Economics.
|
|
Related documents
- 379Kb ~ 2 min (34 pages)
