Welfare state regimes and differences in self-perceived health in Europe: A multilevel analysis
Introduction
The purpose of this paper is to examine and explain between-country differences in self-perceived health in Europe by undertaking a multilevel analysis of the European Social Survey (2002, 2004). It focuses on one main research question: to what extent does welfare state regime classification explain the proportional variation of self-perceived health between European countries, when individual and regional variation is accounted for? This not only implies a need to determine the degree to which self-perceived health actually varies between countries but also to examine whether (and if so, why) the characteristics of certain types of welfare state and welfare state regimes may have a health-protective effect.
It is now widely acknowledged that welfare states are important determinants of health in Europe as they mediate the extent, and impact, of socio-economic position on health (e.g. Bambra, 2006a, Eikemo et al., in press, Navarro et al., 2003). Health status, especially inequalities in health within and between European countries, is largely determined by income inequalities, the distribution of wealth, and other aspects of socio-economic inequalities (Kawachi et al., 1997, Mackenbach et al., 1997, Wilkinson, 1996). Welfare provision in its entirety (social transfers and welfare services) is designed to address these issues of inequality and should, therefore, have a bearing upon health outcomes (Bartley and Blane, 1997, Conley and Springer, 2001, Navarro et al., 2003) Welfare states provide a variety of social transfers (such as housing related benefits, unemployment, pensions, and sickness and disability benefits) as well as key services (most notably health care or social services), which together mediate the relationship between socio-economic position and health. The principles underpinning welfare states, the generosity of social transfers, and entitlements, vary extensively across European countries. For example, in some unemployment benefits are related to previous earnings (e.g. Norway, Germany), whereas in others they are provided at a standard flat-rate (e.g. UK) lowering the relative wage replacement rate (Eikemo & Bambra, 2008). Similarly, entitlement to welfare state benefits and services varies, with some countries providing universal coverage (e.g. Sweden or Norway) whilst others use means-testing (e.g. Ireland or UK).
Welfare state typologies place those welfare states that are the most similar (in terms of principles, provision, etc.) together into clusters of countries with different welfare state regimes, emphasising within regime coherence and between regime differences. Influential within this field is the well known work of Esping-Andersen (1990). In The Three Worlds of Welfare State Capitalism (1990), he classifies welfare states into three regime types (Liberal, Conservative, Social Democratic) on the basis of three principles: decommodification (the extent to which an individual's welfare is reliant upon the market), social stratification (the role of welfare states in maintaining or breaking down social stratification), and the private–public mix (the relative roles of the state, the family and the market in welfare provision). A fourth principle, de-familisation (‘the degree to which individual adults can uphold a socially acceptable standard of living, independently of family relationships, either through paid work or through social security provisions’) (Lister, 1997) was added to the analysis in 1999 (Esping-Andersen, 1999). These principles reflect the relative roles of the state, the family and the market in the provision of welfare. Liberal (UK, Ireland) welfare states are characterised by their basic and minimal levels of provision: social transfers are modest and often attract strict entitlement criteria; recipients are usually means-tested and stigmatised; the dominance of the market is encouraged both passively, by guaranteeing only a minimum, and actively, by subsidising private welfare schemes. The conservative welfare state regime (Germany, France, Austria, Belgium, Italy and, to a lesser extent, the Netherlands) is distinguished by its ‘status differentiating’ welfare programs in which benefits are often earnings related, administered through the employer, and geared towards maintaining existing social patterns. The role of the family is also emphasised and the redistributive impact is minimal. However, the role of the market is marginalised. The Social Democratic regime type (Scandinavian countries), is characterised by universalism, comparatively generous social transfers, a commitment to full employment and income protection, and a strongly interventionist state. The state is used to promote social equality through a redistributive social security system.
There has been extensive scholarly debate about the theoretical and empirical value of the Three Worlds typology (for a detailed summary see Arts & Gelissen, 2002 or Bambra, 2006b) and as a result of this, modified or alternative typologies have been proposed by others (Bambra, 2004, Bambra, 2005a, Bambra, 2005b, Bonoli, 1997, Castles and Mitchell, 1993, Korpi and Palme, 1998, Navarro et al., 2006), most of which place emphasis on those characteristics of welfare states not extensively examined by Esping-Andersen or which cover more countries. Ferrera's (1996) four-fold typology, which focuses on different dimensions of how social benefits are granted and organised, has been highlighted as one of the most empirically accurate welfare state regime typologies (Bambra, 2007a). Ferrera makes a distinction between the Scandinavian (Social Democratic), Anglo-Saxon (Liberal), Bismarckian (Conservative) and Southern countries (Fig. 1). Although there are clear similarities between Ferrera's and Esping-Andersen's typologies, Ferrera's classification is intended to account for differences in the way welfare is delivered whilst Esping-Andersen's still tends to emphasise the quantity of welfare provided (Bambra, 2007a, Bonoli, 1997). In this way, the additional Southern regime is characterised by a fragmented system of welfare provision which consists of diverse income maintenance schemes that range from the meagre to the generous and a health care system that provides only limited and partial coverage. There is also a strong reliance on the family and charitable sector (Ferrera, 1996). One new challenge to conventional welfare regime typologies concerns the Eastern European countries. These countries have experienced extensive economic upheaval and have undertaken comprehensive social reforms throughout the 1990s (Kovacs, 2002). In comparison with the other member states of the European Union, they have limited health service provision and overall population health is relatively poor. In our analysis, we will use Ferrera's (1996) typology expanded by adding a category for Eastern Europe (Czech Republic, Hungary, Poland, and Slovenia).
Although the focus of this paper lies at country-level, it is important to consider health variations at the individual-level additionally, because health is mainly attributed to individual characteristics. At this level, social inequalities in health have mainly been approached by means of occupational class (Kunst et al., 2005, Kunst and Mackenbach, 1994), educational attainment (Cavelaars et al., 1998, Silventoinen and Lahelma, 2002) and income (Adler et al., 1994, Cavelaars et al., 1998, Fritzell et al., 2004, Marmot, 2002, Subramanian and Kawach, 2006, Subramanian and Kawachi, 2004, Wagstaff and van Doorslaer, 2000), all regarded as indicators of socio-economic status (SES). We have also seen that social network (social contacts with one or several persons) and social support (quality of social interactions) appear to make individuals feel healthier, live longer, feel better and cope with difficulties due to chronic diseases and acute difficulties (Berkman, 1985, House et al., 1988, Pinquart and Sorensen, 2000, Wilkinson, 1999a, Wilkinson, 1999b).
The data used in this study also allow us to investigate the proportional variation of self-perceived health among regions as compared to individual and country-level variation. Even though the extent to which self-perceived health varies among regions within European countries has not been previously investigated, we might expect that self-perceived health in Europe is also related to regional factors. Previous studies of mortality have shown that there is a regional North–South gradient in ischaemic heart disease mortality in both Britain and France, which could be partly explained by people's socio-economic position (Lang et al., 1999, Morris et al., 2001). Another study concluded that ischaemic heart disease mortality is about 50% higher in East compared to West Germany (Muller-Nordhorn, Rossnagel, Mey, & Willich, 2004).
Some European countries are healthier than others (Mackenbach, 2006) and the main aim of this study is to examine and explain disparities of self-perceived health in Europe by means of welfare state regimes, when individual and regional variation are accounted for. Welfare state regime typologies have previously been used to analyse cross-national differences in population health (Chung and Muntaner, 2007, Coburn, 2004, Navarro et al., 2003, Navarro et al., 2006). These studies have invariably all concluded that population health is enhanced by the relatively generous and universal welfare provision of the Scandinavian countries (Chung and Muntaner, 2007, Coburn, 2004, Navarro et al., 2003, Navarro et al., 2006). For example, studies have consistently shown that infant mortality rates (IMR) vary significantly by welfare regime type (Bambra, 2006a, Chung and Muntaner, 2007, Coburn, 2004, Navarro et al., 2006), with rates lowest in the Scandinavian countries and highest in the Southern regimes. These systematic differences in health outcomes may be explained by the relative roles of the state, the family and the market in welfare provision (Esping-Andersen, 1990) with the more highly decommodifying welfare states (Scandinavian) – through income redistribution (Dahl et al., 2006, Subramanian and Kawachi, 2006, Torsheim et al., 2006, Torsheim et al., 2006) and low unemployment (Ferrie et al., 2002, Keefe et al., 2002, Lahelma, 1992, Martikainen and Valkonen, 1996) – providing better protection against the health effects of a low market (socio-economic) position. Furthermore, the welfare state is important to population health in terms of how the state interacts with the family structure (Hatland, 2001), and thereby reduces the welfare burden on families and/or women (the state de-familises the family/women) (Bambra, 2004, Bambra, 2007b, Esping-Andersen, 1999, Korpi, 2000).
Previous studies of health differences between welfare state regimes (e.g. Bambra, 2006a, Chung and Muntaner, 2007, Coburn, 2004) have used mortality (especially IMR) or life expectancy data as their health outcomes (Navarro et al., 2006); they have tended to rely on Esping-Andersen's three-fold classification of welfare states (often excluding Southern and Eastern Europe), and they have seldom utilised a multilevel design (except for the studies of Olsen & Dahl, 2007 and Chung & Muntaner, 2007, in which two levels were applied). Therefore, this paper is the first to focus particularly on morbidity (self-perceived health) differences between welfare state regimes in Europe. This study is also distinguished from others in this field because we use a five-fold typology of welfare states, and we also use multilevel analysis with three levels.
Section snippets
Data and methods
This study is based on the cumulative data file (edition 2.0) for the first two rounds (fielded in 2002 and 2004) of the European Social Survey (ESS), which was released on January 29, 2007. The main objective of the ESS is to provide high quality data over time about changing social attitudes and values in Europe. Extensive descriptions of the ESS are at the ESS web site (www.europeansocialsurvey.org). The data and extensive documentation are freely available for downloading at the Norwegian
Results
The interpretation of the results from Table 2 is very straightforward, as it simply shows how much of the total variance of self-perceived health that is attributed to each of the three levels. This is calculated as the ratio of the random country variance (i.e. the intercept) to the total variance. For example, the country-level variance of poor general health using the 2nd order PQL method is 0.318, which gives a proportional variance (in percent) of 8.68. The calculation is [0.318/(3.29 +
Discussion
Summarising the results, this study has shown that nearly 90% of the variation of self-perceived general health outcomes was at the individual-level. Country-level characteristics accounted for around 10% of disparities in self-perceived health. Intra-country regional variation, however, was almost non-existent after controlling for individual and country-level variation. The key finding of this study is that the Scandinavian and Anglo-Saxon welfare regimes seem to have arrangements that give
Conclusion
This study confirms what previous studies on the social determinants of health have shown; that socio-economic position is important in explaining disparities in health at the individual-level. However, going further into the mechanisms of different welfare state regimes is an important path to follow in the process of identifying interventions to improve public health, as welfare regime appears to account for approximately half of the national-level variation of health inequalities among
Acknowledgement
We are grateful to Harvey Goldstein (University of Bristol) for providing us with useful comments related to the methods section of this article.
References (86)
- et al.
Socioeconomic factors, material inequalities, and perceived control in self-rated health: cross-sectional data from seven post-communist countries
Social Science & Medicine
(2000) Self-perceived health in East and West Europe: another European health divide
Social Science & Medicine
(1998)- et al.
Welfare state matters: a typological multilevel analysis of wealthy countries
Health Policy
(2007) Beyond the income inequality hypothesis: class, neo-liberalism, and health inequalities
Social Science & Medicine
(2004)- et al.
For whom is income inequality most harmful? A multi-level analysis of income inequality and mortality in Norway
Social Science & Medicine
(2006) - et al.
Socioeconomic inequalities in morbidity and mortality in western Europe
Lancet
(1997) - et al.
Excess mortality of unemployed men and women during a period of rapidly increasing unemployment
Lancet
(1996) - et al.
Politics and health outcomes
Lancet
(2006) - et al.
Health differences between European countries
Social Science & Medicine
(2007) - et al.
Whose health is affected by income inequality? A multilevel interaction analysis of contemporaneous and lagged effects of state income inequality on individual self-rated health in the United States
Health & Place
(2006)