A number of recent studies have suggested that Western economies are experiencing a `feminization' of poverty, as women are increasingly over-represented among the poor, while most empirical measures of poverty simply assume that households' resources are equally distributed among their members. Classifying individuals as `poor' if they belong to households with total incomes below a chosen `poverty line' may therefore mask such higher poverty among women. In Discussion Paper No. 913, Jeanette Findlay and Research Fellow Robert Wright use data from the Luxembourg Income Study to examine the effects of unequal sharing within households for Italy and the US, whose distributions of poverty by gender are markedly different under the assumption of equal sharing. They consider three measures of poverty: its incidence (a head-count measure) and average and relative deprivation (the average income level of the poor and its distribution among them).

Findlay and Wright calculate a range of estimates for each of these measures under different assumptions about income-sharing. If women transfer 20% of their `equal share' to men and children, the `gender gap' widens and women in Italy are shown to be over-represented among the poor as they are already known to be in the US. Repeating this exercise for a sub-sample of households consisting only of adult couples with children reveals that differences across household types have very little effect on the measurement of female poverty, which is determined rather by assumptions concerning their internal financial arrangements. Since these results highlight the importance of intra-household income distribution, on which no accurate data are available, Findlay and Wright suggest modelling them as a bargain between household members who control different resources and have different opportunities outside the household.

Gender Poverty and the Intra-household Distribution of Resources
Jeanette Findlay and Robert E Wright

Discussion Paper No. 913, February 1994 (HR)