The effects of changes in family policy on gender differences in employment: a quantitative comparative investigation of 20 OECD countries over the years 1985-2010

This thesis provides a quantitative comparative investigation into the effects of changes in family policy on gender differences in employment outcomes. Existing comparative research offers a sophisticated account of the influence of family policy on gender equality in employment – certain leave and childcare policies are found to promote female economic activity, especially among less educated women, but may at the same time also damage women’s occupational attainment with harmful effects that appear particularly severe for women with high levels of education. Yet, this existing literature does suffer limitations, particularly in the extent to which many studies rely on cross-sectional data only. The contribution of this thesis is to test whether established associations hold when looking at changes within countries over time. It examines how changes in family policies affect measures of gender differences in economic activity and occupational attainment, both at the overall level and when data are broken down by levels of education. Data are measured at the country-level and take time-series cross-section form, with a sample that stretches across 20 OECD countries and the years 1985-2010. Analysis is conducted primarily through fixed effects multiple linear regression. Results suggest that increases in certain leave and childcare policies may help close gender gaps in labour force participation rates, particularly but not exclusively among men and women with low education. Simultaneously, increases in leave entitlements for mothers may moderate gains in women’s relative access to managerial positions and top quintile wages. Notably, though, the damaging effects of changes in leave do not appear to apply to highly educated women. Thus, results are consistent with the broader argument that family policies may have contradictory effects on gender differences in employment, but provide less support for the more recent suggestion that any such ‘policy paradox’ operates neatly along class lines.