Gender policy and gender equality in a public health perspective : investigating morbidity and mortality in Sweden and 22 OECD countries

Abstract

The aim of this thesis is to investigate gender policy and gender equality as determinants of health. Data at individual, municipal and country level were employed, and the settings were Sweden and 22 OECD countries. The studies span the time period 1973-2008.

In Study I, gender equality at municipal level (N=289) was measured using indicators of political participation, division of labour in the private and public spheres, and the distribution of economic resources. The outcomes were life expectancy and number of days of compensation for sickness absence and disability. The data were cross-sectional (2000-2004) and linear regression was used. Gender equality was correlated with lower life expectancy and higher levels of sickness and disability for both men and women. In Study II, a cohort of parents (N of approx. 75,000) were followed from 1980 to 1999.

The aim was to assess possible interaction between gender equality at couple (between partners) and local level. For this purpose, a combinatorial index was constructed. The outcome was number of days compensated from sickness insurance during 1986-1999, with a cut-off point at the 85 % percentile. Logistic regression was used. Among fathers, those who were traditionally unequal with their partner and lived in an equal municipality had higher levels of sick-leave. Mothers who earned less and/or had a lower occupational position than their partner had lower levels of sick-leave, especially when they lived in a traditional municipality.

Study III aimed to classify 22 OECD countries into homogenous clusters, using hierarchical cluster analysis for 1979, 1989, 1999 and 2004. Included were indicators of taxes, parental leave, pensions, social insurances and social services in kind to reflect Sainsbury’s three gender policy regimes. The empirical classification did not completely correspond to Sainsbury’s theoretical typology. Rather than an emerging separate gender roles regime, there was a compensatory breadwinner cluster, distinguished by compensatory measures in the pension system directed at carers and extended childcare leave following maternity leave. A male breadwinner cluster was found, with the core feature of having a tax system that supports single (as opposed to dual) earner families. From 1989 there was an earner-carer cluster, characterised by generous parental leave, high social services expenditure, and universal basic pensions. In Study IV, the cluster solution for 2004 and specific policy indicators for 1973-2008 were used as predictors, and the outcomes were mortality from external causes and circulatory disease. For this repeated measures model, hierarchical linear regression was used. Both the earner-carer cluster and specific policies, i.e. generous parental leave, high social services expenditure and universal basic pensions, were associated with a smaller gender gap in external cause mortality, primarily due to increased female mortality. For circulatory disease mortality, both the earner-carer and the compensatory breadwinner cluster experienced a larger decrease in male mortality over time.

In conclusion, the relationship between gender equality and health differed according to the unit of analysis and the outcome studied, and also varied between men and women. Based on these results, hypotheses for future studies are formulated.