Consumption patterns of poor households

Abstract

The essential purpose of this study was to identify typical consumption patterns of poor households. In particular, this study explored consumption pattern dimensions and consumption pattern types, and examined two hypotheses: (a) if differences exist in consumption patterns between poor and nonpoor households; and (b) if differences exist in consumption patterns among poor households. Factor analysis was used to identify consumption pattern dimensions, and cluster analysis was used to identify consumption pattern types. Multiple logit analysis was employed to test the hypotheses.

The data were taken from the BLS 1987 Consumer Expenditure Survey, and the sample consisted of 1,775 households.

This study found three consumption pattern dimensions, i.e., the Private Transportation/Shelter dimension, the Social Consumption/Shelter dimension, and the Health & Home Necessities/Private Transportation dimension. Based on the variation in the dimensions, five consumption pattern types were identified, i.e., the Shelter-Dominated Pattern, the Balanced Moderate Pattern, the Private-Transportation Dominated Pattern, the Health & Home Necessities Emphasis Pattern, and the Social Emphasis Pattern. Findings from the logit analysis showed that: Holding other factors constant, poor households were more likely than nonpoor households to have the Shelter-Dominated Pattern or the Health & Home Necessities Emphasis Pattern, and less likely to have the Balanced Moderate Pattern, the Private-Transportation Dominated Pattern or the Social Emphasis Pattern. Membership in a given consumption pattern of poor households varied with number of earners, homeownership, education, age, race, marital status, and household size.

The findings of this study suggest that the poor are not homogeneous in their consumption patterns. They also showed that some socioeconomic and demographic characteristics of poor households, such as age, race, education, marital status, household size, number of earners, and housing tenure, did help to predict their consumption patterns.