Introduction

The concept of risk has long since developed from an abstraction solely concerned with attempting to predict the consequences of modern society to one which is now used to articulate them (Douglas, 1992). This is no less true in the context of studies on homelessness where the concept of risk generally finds its expression in two related but ultimately distinct ways. The first sense concerns the risk of becoming homeless such as when we are told that the young are more at risk of homelessness than the old, that in England as a whole the risk of experiencing homelessness over the last 10 years is about 4.3% and that a black head of household is over three times more at risk from experiencing homelessness than a white head of household. The second concerns the risks inherent in and attendant on homelessness; exemplified in the finding that single homeless persons are at greater risk of poor health than the general population (see, for example, Bines, 1994; Burrows, 1997; Kemp, 1997; Pleace and Quilgars, 1997).

Risk, although rarely explicitly, is therefore an important and much used medium of ostensible explanation and articulation in the discussion of homelessness. What is more significant in the context of this discussion, however, is not the pervasive use of risk, but rather how the use of risk converges to at once produce and reinforce the status of the homeless population as a population of victims. More specifically, the homeless are primarily presented as suffering in two principal ways – first as a simple result of being homeless and then again for experiencing the problems that are attendant on homelessness, such as malnutrition and premature death (Keys and Kennedy, 1992; Grenier, 1996). This depiction of homeless people as victims, although predominant, is not, however, complete, but rather is subject to one important exception, namely those narratives that relate to homelessness and crime.

When this dimension of homelessness is typically represented the relationship between homelessness and the risk of victimisation is inverted (see, for example, Baron and Hartnagel, 1998; McCarthy and Hagan, 1991; Rothman, 1991). The homeless person as a victim, in other words, tends to disappear and instead the emphasis is placed on the risk to the ‘respectable’ public of criminal victimisation from the homeless. This is not to say that the depiction of the relationship between the homeless and criminal victimisation is monolithic.