Is crime Giffen?
| Pages | 80-85 |
| Published date | 02 January 2009 |
| Date | 02 January 2009 |
| DOI | https://doi.org/10.1108/13590790910924984 |
| Author | Michael Cain |
Is crime Giffen?
Michael Cain
School of Business, University of Wales, Bangor, UK
Abstract
Purpose – This paper aims to develop a utility maximising framework for criminal behaviour and to
consider a Markowitz-type utility function, centred at current wealth, for a representative criminal.
Design/methodology/approach – The paper takes the form of a conceptual discussion.
Findings – The Kahneman-Tversky value function form seems to be consistent with the current
empirical evidence, expressed mainly as the ratio of certain elasticities. The criminal’s optimal level of
crime is obtained and comparative statics derived to help suggest how crime might be controlled.
Research limitations/implications – Empirical work is encouraged to elicit the utility function of
a representative criminal in this analytical framework, to ascertain if crime is a favourable bet and if it
could indeed be Giffen.
Originality/value – The solution shown in the paper involves an upper limit on the odds against
detection of the criminal and it is shown that a distinct possibility is that crime is a Giffen commodity.
Keywords Criminals, Behaviour, Wealth
Paper type Research paper
1. Introduction
An expected utility maximising criminal commits an illegal act and, if he is not caught
and punished, his total wealth thereby increases by an amount x. His criminally
enhanced total wealth, wþx, will be greater than his current wealth, w. He is caught
and punished with probability pand the punishment consists of a fine, z, which is less
than or equal to his enhanced wealth, wþx. His personal assessment of any benefits to
him of his criminal activity is described by a utility of wealth function, U(.). Currently,
the utility of his wealth is U(w) but after the criminal activity, his utility will have
risen to U(wþx) or fallen to U(wþx2z). His expected utility in committing the
crime is:
E¼pUðwþx2zÞþð12pÞUðwþxÞ;ð1Þ
an average of his utilities according to whether or not he is caught and punished. This
is essentially the model of Becker (1968) who, noting that criminals are more
responsive to a change in the probability of detection, p, than to a change in the fine, z,
shows that within this framework a criminal must be risk-seeking; a risk-averse
expected utility maximising criminal cannot respond more to a change in the chance of
being caught than to a change in the severity of the punishment.
Instead of restricting attention to purely risk-averse or risk-seeking utility functions
as Becker did, this paper considers the more interesting case of a criminal who has a
Markowitz (1952) type utility function, centred at his current wealth, with the
particular form suggested by Kahneman and Tversky (1979). Contrast this approach
with that of Neilson and Winter (1997) who assume state-dependent utility or a
rank-dependent utility model to explain observed behaviour; in order to incorporate
The current issue and full text archive of this journal is available at
www.emeraldinsight.com/1359-0790.htm
JEL classification – D11, K42, D80
JFC
16,1
80
Journal of Financial Crime
Vol. 16 No. 1, 2009
pp. 80-85
qEmerald Group Publishing Limited
1359-0790
DOI 10.1108/13590790910924984
To continue reading
Request your trial