Crime os о Social Cost of Poverty
and Inequality: A Review Focusing
on Developing Countries
F. Bourguignon
existing economic inequalities -see, for instance, Benabou (1996). The
variable H in the general crime function (4) thus provides a third channel
through which economic cycles or the equalizing or unequalizing nature
of the urbanization process may affect crime and violence. The first one is
through the direct benefit and cost of crime. The second one goes through
public decision making in matters of crime deterrence. This third one
goes through the influence of economic conditions on some sociological
factors behind the propensity of individuals to commit crime20.
In any case, the main economic mechanism directly linked to the honesty
variable is probably the one alluded to above, that is the way this part of
the social capital may be eroded durably by an increase in a crime rate
whose causes lie in the economic sphere. In the presence of more crime
resulting from the adverse effects on poverty and equality of a long and
severe economic recession, moral and social structures are likely to be
weakened, which in turn may increase the prevalence of crime and
violence21.
There may be an objection to this and most of the preceding arguments
that they are taking too much an economist’ view of criminal behavior
and therefore that they may be misleading for policy. For instance, many
observers insist that violence in big metropolitan areas of developed and
developing countries is often not directed towards the property of others
but takes place internally within specific segments of society located in
the poorest districts22. Obvious examples of this is all the violence related
to conflicts related to the control of illicit activities like drug dealing,
drug trafficking, and different types of gambling or prostitution. In many
violent parts of today's metropolises this, rather than more conventional
property crime like burglary or robbery, seems to be the single dominant
cause for the development of violence and the surge in homicides. Another
departure from the canonical model might lie in the very low probability
of crime detection and sanction noted in many studies of erime and violence
21-1 For a ⅛eneral analysis of these factors with a framework similar to the present one. see Ha⅛an
(1994).
21 On this relationship between societal values and crime, see Λker!of and Yellen ( 1994). See also
Vferdierand Bisin (1997)
22 .See, for instance. the taxonomy of violence proposed Irj .Moscr (1997).
78