The Impact of Geographical Scale in Identifying Areas as Possible Sites for Area-Based Interventions to Tackle Poverty: The Case of Montreal

Many studies in geography have demonstrated that results can vary according to scale and configuration of spatial units. However, implications of the MAUP-Modifiable Area Unit Problem-have received little attention in urban planning and policy studies. The first objective of this paper is to demonstrate how identification of poor areas (residential areas with 40% or more of low-income population) is sensitive to changes in scale. The second objective is to measure the completeness (to what degree targeting only poor areas is adequate to reach all the poor population of a metropolis) and the efficiency (to what degree the population benefitting from targeted interventions is poor) of area-based interventions. The methodology to identify poverty areas is based on three spatial levels of analysis: dissemination areas (DAs), census tracts (CTs) and districts, namely micro, meso and macro levels. Hierarchical linear models are used to analyze the variance partitioned between these levels. Results of multilevel analyses demonstrate that CTs are more heterogeneous on the basis of DAs than Zones according to CTs. Independently of the scale of analysis, results show that the majority of low income population is living in areas with no poverty concentration in Montreal (completeness dimension). Moreover, results show that about half of the population living in zones of poverty concentration is not poor even when micro-zones are used for measurement (efficiency dimension). These results illustrate the importance of scale in identifying poverty areas which also have implications for social policy interventions.