PERCEPTIONS OF CRIME AND SUBJECTIVE WELL-BEING: URBAN-RURAL DIFFERENCES IN SOUTH AFRICA.

Reducing crime is a fundamental goal of all policymakers, because its costs are both pecuniary and non-pecuniary. A number of studies have shown that individuals who are victims of crime report lower levels of life satisfaction or subjective well-being (SWB) as compared to nonvictims (Powdthavee, 2005; Cohen, 2008; Davies & Hinks, 2010). But, what is not as clear is how this nexus may differ based on household geo-type, or put differently, whether a household is located in an urban area versus a rural area. The rationale as to why geographical location is likely to play a role is as follows: (a) people migrate from rural to urban areas in hopes of finding a job or better standard of living. Given that urban migration is associated with "false expectations" (Mulcahy & Kollamparambil, 2016:1357), it may lead some to turn to crime as a means of coping. Therefore, one expects to see more crime in cities as opposed to farm areas. (b) In rural areas, people share a sense of community or 'Ubuntu'1. This means that should someone fall victim to crime, others will ensure that its effects on well-being are minimized, by assisting where they can. (c) Social disorganization theory explains differences in urban and rural crime rates, as it suggests that "a person's residential location is more significant than the person's characteristics when predicting criminal activity" (Bond, 2015: para 4). This paper investigates the impact of perceived crime on individual well-being in urban and rural areas of South Africa using a nationally representative panel dataset from 2008 to 2017. By making use of a fixed-effect model, a finding was that perceived crime negatively influences the SWB of individuals living in rural and urban areas in South Africa. Unlike the urban coefficient, the rural coefficient is not statistically significant. These results suggest two patterns: firstly, perceived crime plays a stronger role in predicting SWB in urban areas as compared to rural areas. This is likely because urbanization is linked to a higher rate of crime (Park, Burgess & McKenzie, 1925). Secondly, it provides empirical support that there exists a geographically differentiated relationship between perceived crime and the SWB nexus in South Africa. The estimates are robust to the inclusion of the inequality variable as well as alternative models (specifically random ordered probit model). [ABSTRACT FROM AUTHOR]