Unemployment, business cycles, and crime specialization: Canadian provinces, 1981–2009
| Date | 01 September 2016 |
| DOI | 10.1177/0004865815575395 |
| Published date | 01 September 2016 |
| Subject Matter | Articles |
Australian & New Zealand
Journal of Criminology
2016, Vol. 49(3) 332–350
!The Author(s) 2015
Reprints and permissions:
sagepub.co.uk/journalsPermissions.nav
DOI: 10.1177/0004865815575395
anj.sagepub.com
Article
Unemployment, business
cycles, and crime specialization:
Canadian provinces, 1981–2009
Martin A Andresen
School of Criminology, Institute for Canadian Urban Research Studies,
Simon Fraser University, Burnaby, Canada
Shannon J Linning
School of Criminology, Simon Fraser University, Burnaby, Canada
Abstract
The relationship between unemployment and crime is complex, consisting of two independ-
ent and counteracting effects: motivation and guardianship. The Cantor and Land model
integrated these two effects leading to a new literature investigating the relationship between
unemployment and crime. However, this literature always considers the impact of unemploy-
ment (or some other measure of the economy) on the volume or rate of crime. In this paper,
we investigate the role unemployment plays in crime specialization on the Canadian
provinces, 1981–2009. Using panel data and a hybrid modeling technique we find that
unemployment impacts crime specialization, but this impact varies in magnitude and by
crime type.
Keywords
Crime specialization, hybrid modeling, location quotient, unemployment
Introduction
There exists a long-standing inquiry into the relationship between economic fluctuations
and crime in the criminological literature. Theoretical perspectives such as social disor-
ganization and strain theories have considered the impact of economic conditions on
crime and delinquency (see Cloward & Ohlin, 1960; Merton, 1938; Shaw & McKay,
1942). Cantor and Land (1985) argued that these approaches, however, only accounted
for the presence of motivated offenders and not for opportunity and/or guardianship
(pp. 318–319). Their work set out to address this overlooked component and utilized
unemployment as a measure of economic conditions. Results suggested that the rela-
tionship between unemployment and crime varied by crime type and whether focus was
Corresponding author:
Martin A Andresen, School of Criminology, Institute for Canadian Urban Research Studies, Simon Fraser University,
8888 University Drive, Burnaby, BC V5A 1S6 Canada.
Email: andresen@sfu.ca
placed on criminal motivation vs. opportunity (p. 330). Overall, the opportunity effect
was more prevalent than the motivation effect (Cantor & Land, 1985). Their work
inspired a surge of empirical work that persists to this day. Contemporary research
has lobbied for the use of multiple economic measures, more specific crime data and
disaggregate units of analysis (Andresen, 2012, 2013a; Arvanites & DeFina, 2006;
Phillips & Land, 2012).
The works of Cantor and Land (1985) and others have compared economic meas-
urements to traditional crime measures such as counts or rates. For many decades,
however, empirical research has cautioned researchers about potential inaccuracies
inherent in such crime measures (Andresen & Jenion, 2010). The location quotient
(LQC), a statistic introduced to the criminological literature by Brantingham and
Brantingham (1993) addresses some of these potential inaccuracies and is adaptable to
criminological research. LQC calculations provide a comparison of the occurrence of
crime in a particular location to crime relative to the entire area of study, specialization
(Andresen, 2007). In other words, this technique allows researchers to determine whether
a particular area specializes in a specific crime type relative to the entire study area
(Block, Clarke, Maxfield, & Petrossian, 2011).
This paper aims to further the research established by Cantor and Land (1985) by
employing alternate measures of the unemployment–crime relationship. First, multiple
economic indicators are included to provide a better understanding of the short- and
long-term economic effects on crime—this has been shown to be important in recent
research (Andresen, 2015). These variables include: unemployment, gross provincial
product (GPP), GPP per capita, and low income. Second, panel data from the 10
Canadian provinces, 1981–2009, are utilized in a hybrid modeling technique to better
characterize the motivation and opportunity relationship (Andresen, 2013a). Finally,
and most importantly, the LQC is applied to the analyses as an alternate measure of
crime to identify crime specialization across provinces and crime types. As stated above,
all previous research in this area have used traditional measures of criminal activity such
as counts and rates. However, understanding the impact of unemployment of crime
specialization has not been undertaken and is shown to be instructive, below.
The relationship between unemployment and crime
Cantor and Land (1985) laid the framework for decades of research and debate over the
unemployment–crime relationship. While much of the previous literature had explored
the notion that declining economic conditions lead to an increase in criminal activity,
these authors considered the principles of routine activity theory: motivated offenders,
suitable targets and a lack of a capable guardian (Cohen & Felson, 1979, p. 589). As
such, Cantor and Land (1985) explain that while much of the traditional research
focused primarily on the first element, focus should also be paid to the notion of guard-
ianship. They argue that when the unemployment rate increases, individuals are not in
places of employment and return to the relatively protective environment of the home,
increasing guardianship of themselves and property, decreasing opportunity (Cantor &
Land, 1985).
Incorporating this new dimension into the unemployment–crime research, two overall
phenomena became the focus of Cantor and Land (1985): motivation and guardianship/
Andresen and Linning 333
To continue reading
Request your trial