Seftiembre de 1«»»
ÜESARROLLO Y S
Another point to take into aceount is the possibility for part of the
population to buy private protection through more or less sophisticated
alarm systems, private guards and strict residential segregation improving
the capacity to spot intruders and would-be criminals. If the social class
which can afford this type of security has some control over political
decisions, one may very well imagine a situation where nothing substantial
is done to increase public security outside these residential areas despite
mounting criminality.
An important implication of private protection against crime is that it
may drastically modify the relationship between poverty, inequality and
crime. In effect, the possibility of self-insurance against erime logically
lessens the relationship between the rate of crime and poverty. This works
as follows. Potential victims anticipate that more poverty and inequality
associated with an unbalanced process of urbanization increase crime risks.
They buy additional protection and this reduces the actual change in
criminality. The marginal social cost of poverty and inequality going directly
or indirectly through crime remain the same, however. In the expression
(5) of this loss, the effect of a marginal increase in the rate of crime c at
constant ‘police’ expenditures, P, is simply replaced by a change in the
‘private protection part’ off?
The last argument of the general crime function (4) is certainly the most
difficult to discuss for an economist. For the sake of simplicity, it was
referred to simply as an ‘honesty’ parameter. But, actually, one should
include in it all the variables which may explain that given some cost-
benefit ratio of crime and some characteristics of the justice and police
systems, the degree of crime may vary drastically from a city or a country
to another. These include ethnicity, religion, family structures, residential
segregation, etc. Some of these factors may clearly be related to economic
phenomena. The increase in the proportion of single mothers in urban
areas is probably not foreign to conditions on the labor market —see, for
instance. Burtless and Karoly (1995) for the case of the l'S It is also often
singled out as a powerful sociological factor of violence’9. Likewasc1
residential segregation has been analyzed as a mechanism to reproduce
14 See dɪɪulio (1996} More generally sec the implications of out of wedlock childbearing and ‘men
without children' iɪt Λkcrlof (1998).
77