Social disorganisation, immigration and perceived crime in Spanish neighbourhoods

This dissertation adopts a quantitative approach to investigate the determinants of residents’ perceptions of neighbourhood crime, focusing specifically on a series of structural factors at the community level, in accordance with the social disorganisation model. Using different statistical models, including correlations, linear regression, multilevel models and spatial regression analyses, and several Spanish data sources, in particular the 2001 Population and Housing Census and a nationally representative survey conducted in 2006, the research confirms the relevance of its exogenous sources in explaining perceived neighbourhood crime. These include classical variables, such as neighbourhoods’ socioeconomic status, residential stability, ethnic diversity, family disruption and degree of urbanisation, but also other features related to the time, skills and resources deployed by residents in their residential areas such as commuting time to work, the number of working hours and the availability of a second home. For its part, other local conditions traditionally associated specifically with perceived neighbourhood crime, such as social incivilities and physical decay, act as mediators of other contextual effects, in particular of the number of retail shops and offices. The research also demonstrates the urban nature of the social disorganisation theory. That is, that the local conditions typically associated with social disorganisation, urban unease and the various social problems that can affect neighbourhoods, are better predictors of residents’ perceptions of crime in town and large cities than in rural areas, operationalized as municipalities of less than 5,000 inhabitants. Small municipalities seem particularly successful in controlling their younger residents for neither the proportion of adolescents and young adults, nor the number of children per family exert an important effect on residents’ perceptions of neighbourhood crime. Among these local conditions, special attention has been devoted to measures of diversity and immigration demonstrating that their effect on residents’ perceptions of neighbourhood crime, except for the positive impact of Asians, is not necessarily robust to different model specifications and statistical methods. This erratic immigrant effect is surprising given how consistent the belief in a crime-immigration nexus is among Spaniards. Precisely on this point, the dissertation has investigated why the belief in a crime-immigration nexus varies significantly between individuals and across communities. Three variables have been identified as determining factors: contextual parochialism, right-wing ideology and the media. In rural areas with high residential stability, a significant presence of elderly population and a low socioeconomic status, residents are more likely to unconsciously associate immigration and crime, even when individual attributes are adjusted for and, more importantly, even if few migrants live in the surroundings. Not surprisingly, right-wing residents are more likely to associate both phenomena yet, in contrast to many statements by scholars and pundits, the media in Spain seems to exert a moderator effect.