Abstract

Individual welfare can be assessed from two directions: either directly by looking at the standard of living the individual has actually attained or indirectly by studying the resources available to the individual. Direct and indirect welfare indicators are afflicted by specific conceptual and empirical problems. Therefore, instead of (arbitrarily) selecting one single indicator for poverty analysis, it is argued that different indicators should be thought of as complementary rather than competitive. As a step towards a better understanding the article investigates systematically the measurement characteristics of both direct and indirect poverty indicators. Poverty indicators derived from data on expenditures as well as on living conditions and participation in social activities are compared to the usual income statistics and, as expected, the indicator based on expenditures (the money equivalent of the standard of living) shows a much higher compatibility with the income data than the proposed measure of deprivation from a socially accepted „style of living". The latter „agrees" only in the lower ranks of the distribution with the income indicator. Therefore, the proposed measure of deprivation seems to be a good indicator to identify individuals that have a low standard of living because of few resources, especially because of low incomes. Additionally, significant differences in the socio-demographic profiles of the individuals classed as poor by each of the three poverty indicators are observed. Household size shows the largest differences, but to a large extent this can be attributed to different assumptions on economies of scale implied by each indicator used in the analysis. The expected differences with respect to age, specific occupational groups and household types can only partially be confirmed and need more research, preferably using longitudinal data.