Do police cause crime : a replication
Date
1993
Authors
MacKinnon, Kelly
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
What is the relationship between police strength and reported crime rates? Are more police authorized in response to increased rates of crime or are increased crime rates a response to increased police strength? A correlational analysis by Koenig (1991) focusing on national level data from 1962 to 1988 suggests a tendency for increased police strength to precede increased crime rates. Using data from the ten Canadian provinces, this study replicates the approach taken by Koenig. Sworn police strength is correlated with selected crime rates from 1962 to 1988 to explore whether a relationship exists at the provincial level between police strength and crime rates. Temporal associations are examined through the use of lagged correlations. The crime rates examined include robbery, all violent crimes, break and enter, all property crimes, criminal code offences, and total offences. In order to control for variations in the age and sex structure of the population, each of these crime rates are expressed per 1,000 population male and per 1,000 population male of various ages. Controls are also introduced for unemployment, inflation (in the form of the consumer price index), and time.
The analysis of the provincial data is relatively consistent with the national level which it is seeking to replicate. There is a congruence in terms of the three major findings. First, both analyses reveal a zero order positive association between crime rates and police strength, and police strength and crime rates, both in lagged and unlagged correlations. Thus, it appears that per capita police strength and crime rates have been moving in tandem over the 27 year period of study.
Second, the provincial analysis reveals a general pattern of positive associations between per capita police strength and subsequent levels of selected crimes when unemployment, consumer price index and time are controlled. Thus, it appears that, in most provinces, increased police strength tends to precede increased reported rates of crime when controls are introduced for unemployment, consumer price index, and time. Increased police strength per capita has typically been associated with subsequent increases in the reported rates of selected crimes (i.e. robbery, break and enter, all property offences, criminal code offences and total offences).
Third, both analyses find that violent crime has a dynamic of its own. Consistent with Koenig's analysis, this analysis reveals a negative association between lagged police strength and subsequent reported rates of violence at the provincial level. While reported levels of violence continue to increase, this analysis suggests that provincially reported rates of violence have at least been moderated by increased police strength.
While the provincial level analysis is, for the most part, supportive of Koenig's general findings, the nature of the associations between crime rates and police strength were not uniformly consistent with the national level analysis. In terms of the tendency for positive associations between police strength and subsequent reported levels of crime, the data from the three most westerly provinces, British Columbia, Alberta, and Saskatchewan, are not consistent with the majority analysis. In addition, the absence of any consistent relationship between lagged police strength and subsequent rates of violence in Quebec is a deviation from the national findings.
Possible explanations for selected provincial deviations from the national level analysis are deficiencies in the provincial data and factors at the provincial level may influence the relationship between reported crime and police strength.
Description
Keywords
UN SDG 8: Decent Work and Economic Growth