Gender, Poverty, and Social Policy Regimes: A Comparative Analysis of Twenty-Four Upper-Income Countries (Draft)

Although all industrialized countries have enacted public policies that place a floor under household resources and/or redistribute income from higher to lower income families, none have entirely eradicated income poverty. A substantial research literature on poverty in rich countries has reached two over-arching conclusions. One is that the prevalence and intensity of poverty varies markedly across relatively similar countries, due at least in part to variation in social policy designs. The second is that, within all countries, poverty outcomes vary extensively across subgroups. This paper draws on data from the Luxembourg Income Study (LIS), a cross-national microdata archive, to examine one widely recognized factor associated with poverty – that is, gender. Specifically, we focus on the question: How does gender as a poverty risk factor vary across a group of 24 upper-income countries?

A large body of research, much of it drawing on the LIS data, has established that, in many upper-income countries, women are more likely to be poor than are their male counterparts. That is true both before and after taxes and transfers are taken into account. The causes underlying women’s higher risk of economic insecurity are complex, overlapping, and cumulative. The most powerful factor is women’s weaker attachment to the labour market. On average, women command lower market income, including wages and occupational pensions, than do men and, as a result, they also receive lower employment-related social transfers. In addition, as a group, women still command lower pay than do men for each hour worked, partly due to their concentration in lower-paying occupations and partly due to pay discrimination based on gender. In turn, the main reason that women’s connection to paid work is weaker than men’s is their disproportionate engagement in caring for family members, especially young children. Largely because of their greater care-giving duties, women are less likely to be employed than are similarly-situated men and, if employed, they average fewer weekly work hours, including among those in full-time employment. Recent evidence indicates that being an active caregiver (independent of gender) further reduces many women’s hourly pay.

Furthermore, in all upper-income countries, a substantial number of parents are raising their children without partners, and everywhere single parents are overwhelmingly women. Single mothers, as a group, typically report worrisome levels of poverty – not surprisingly, as their solo care-giving responsibilities depress their own labour supply; their gender is associated with lower hourly earnings; and their homes typically lack a second earner. Finally, diverse households – young and old, female-headed and male-headed, with and without children – receive tax benefits and public income transfers. Among lower-income households, those transfers can make them less poor or lift them out of poverty altogether. In some countries, social benefits targeted on children are meagre compared to those granted to other demographic groups. As a result, families with children, which disproportionately include women, are more likely to be poor than are other family types. In many countries, these factors – both micro and macro – operate independently and interactively to raise women’s likelihood of poverty relative to men’s.

Against this broad portrait of commonality, this paper focuses on cross-national variation, in particular on variation that captures diversity in social policy designs. Although nearly all of these 24 countries are high-income countries – three are classified as upper-middle income countries – they are spread across diverse geographic regions, which largely correspond to equally diverse welfare state models. This study includes five Anglophone countries, seven Continental European countries, three Eastern European countries, four Nordic European countries, three Southern European countries, and two countries not easily classified, Israel and Mexico. The selection of countries – especially the limited inclusion of middle-income countries – was driven by data availability. Although the LIS archive will add a large number of middle-income countries over the next three to five years, unfortunately only a few are included at this time.