Do area-based markers of poverty accurately measure personal poverty?

Area-based markers of deprivation (e.g., postal codes) are commonly used to identify groups of people with low socioeconomic status (SES); the validity of this approach, however, remains unknown. In this study, we determined the accuracy of using income quintile groups calculated on the basis of the median family income of each forward sortation area (1996 Canadian census) to identify those living in poverty (i.e., annual family income of < $12,620). The sensitivity and specificity of using the lowest income quintile to capture those in poverty were 26% and 83%, respectively (likelihood ratio (LR) of 1.49; 99% CI, 1.49 to 1.50). Among those in non-metropolitan and metropolitan areas, the LRs were 1.26 (99% CI, 1.26 to 1.27) and 2.01 (99% CI, 2.01 to 2.02), respectively. The use of postal codes as the only marker to identify people with low SES may result in substantial misclassification of personal poverty, particularly in non-metropolitan areas.