Crime as a Social Cost of Poverty
and Inequolrty: A Review Focusing
on Developing Countries
F. Bourguignon
instruments. This is a rather delicate task. Some work has been done on
whether ‘prisons pay’ in the US -see Piehl and diɪulio (1995)-. The
desirability of anti-crime policies like ‘zero tolerance’, ‘three strikes and
you’re out’, ‘fixing broken windows’, which have been very" much debated
in the US, would also have to be analyzed within such a framework17.
However, in most countries, and particularly in developing countries, the
data base necessary to make the corresponding calculation is hardly
available.
From a positive point of view, the endogeneity of expenditures on criminal
justice and police leads to an analysis of crime and its evolution partly
based on the public-decision process behind these expenditures. In times
of rapid urban expansion, all public infrastructures, including those linked
to crime prevention and sanction, tend to lag behind the needs of the
population, which should imply some increase in criminality, other things
being the same. An increased demand for crime deterrence is then
expressed by the civil society, part of which will be effectively met. How
much essentially depends on public-spending decision mechanisms and
possibly on the social structure of urban society. It is interesting to note
that through this political economy of spending on crime deterrence,
economic and social inequality may in effect play an indirect role on crime,
on top of the direct incentives they represent for criminals. It is not clear,
however, whether more inequality should lead to a larger anti-crime budget
or the opposite. Not only the structure of society and the political weight
of the various classes is important here, but also the social geography of
the city and the technology of crime prevention. One may well imagine
instances where the public-decision mechanism about spending on crime
deterrence lead to rich neighborhoods and business districts being heavily
protected and relatively little being spent in poor neighborhoods and on
more general crime disincentives. That in many societies not only criminals
but also victims are found predominantly among disadvantaged social
groups maybe explained by such a mechanism. Ofeoursc, a less inequitable
outcome may also be possible depending on the social characteristics of
the city18.
17 See Kclling and Coles (1996). See also the recent book review by Massing (1998).
1*i This is partly studied in Bourguignon (1998),
76