Its precise definition depends on two subsidiary definitions: of what is poverty and what is feminization. Poverty is a deprivation of resources, capabilities or freedoms which are commonly called the dimensions or spaces of poverty. The term feminization can be applied to indicate a gender biased change in any of this dimensions or spaces. Feminization is an action, a process of becoming more feminine. It necessarily involves changes over time or populations (comparing geographical areas, for example). Feminine, in this case, is used to mean 'more common or intense among women or female headed households'.
Because it implies changes, the feminization of poverty should not be confused with the existence of higher levels of poverty among women or female headed households. Feminization is a process; ‘higher poverty’ is a state. It is also a relative concept based on a women-men (or female-male/couple headed households) comparison, where what matters are the differences (or ratios, depending on the way it is measured) between women and men at each moment. Since the concept is relative, the feminization does not necessarily imply an absolute worsening in poverty among women or female headed households: if poverty in a society is sharply reduced among men and is only slightly reduced among women, there would still be a feminization of poverty.
Measures of poverty "among female headed households" and "among women" are not indicators of the same phenomena. Both capture a gender dimension of poverty but in distinct ways. They differ by the unit of analysis and by the population included in each group, and obviously have different meanings. There are reasons to consider both. The goal of headship-based indicators is to represent what happens to specific vulnerable groups of women and their families, therefore their unit of analysis is the household and the population considered includes both men and women (and children) living in these households, but excludes women and men living in other household formations. Indicators of poverty among females, by their turn, make a complete separation of men and women as individuals, counting or not children as a gendered group in their aggregations. Interpreting the results based on individual-based measures of poverty is affected by the fact that poverty is usually measured at the household level and therefore male poverty is intrinsically associated with female poverty and vice-versa.
The overrepresentation of women among the income poor at a given moment seems to be a much more common phenomena than the process of the feminization of income poverty. Actually, despite the enormous political controversy around the issue, there is little support to the claim that there is a systematic feminization of income poverty in the world. It seems that feminization occurred in the United States between the 1950s and the mid-1970s, although some reject the hypothesis for the years after 1970 and the 1980s. Data from the United Kingdom indicates that from the late 1960s to the mid 1980s there was no evidence of a feminization of poverty. In Canada, it was found a feminization of poverty between 1973 and 1990 when ‘feminization’ was understood as an ‘increase among female headed households’, but not when the ‘increase among women’ definition was used. Little is known about developing countries, but the existing research reports no evidence of a feminization of income poverty after the 1990s in Latin America .
The feminization of poverty, among many other factors, may be caused by changes in:Family composition