Development of the Canadian Marginalization Index: A New Tool for the Study of Inequality

Flora I. Matheson, James R. Dunn, Katherine L.W. Smith, Rahim Moineddin, Richard H. Glazier


OBJECTIVES: Area-based measures of socio-economic status are increasingly used in population health research. Based on previous research and theory, the Canadian Marginalization Index (CAN-Marg) was created to reflect four dimensions of marginalization: residential instability, material deprivation, dependency and ethnic concentration. The objective of this paper was threefold: to describe CAN-Marg; to illustrate its stability across geographic area and time; and to describe its association with health and behavioural problems.

METHODS: CAN-Marg was created at the dissemination area (DA) and census tract level for census years 2001 and 2006, using factor analysis. Descriptions of 18 health and behavioural problems were selected using individual-level data from the Canadian Community Health Survey (CCHS) 3.1 and 2007/08. CAN-Marg quintiles created at the DA level (2006) were assigned to individual CCHS records. Multilevel logistic regression modeling was conducted to examine associations between marginalization and CCHS health and behavioural problems.

RESULTS: The index demonstrated marked stability across time and geographic area. Each of the four dimensions showed strong and significant associations with the selected health and behavioural problems, and these associations differed depending on which of the dimensions of marginalization was examined.

CONCLUSION: CAN-Marg is a census-based, empirically derived and theoretically informed tool designed to reflect a broader conceptualization of Canadian marginalization.


socio-economic status; health; methods; marginalization; inequities