C Bambra, 1 T A Eikemo 2. Research report

Transcription

1 Researh report An appendix is published online only at om/ontent/vol63/issue2 1 Department of Geography, Durham University, UK; 2 SINTEF Health Researh, Norway, and Department of Soiology and Politial Siene, Norwegian University of Siene and Tehnology, Norway Correspondene to: Dr C Bambra, Department of Geography, Wolfson Researh Institute, Durham University, Queens Campus, Stokton on Tees, TS17 6BH; lare. bambra@durham.a.uk Aepted 4 September 2008 Welfare state regimes, unemployment and health: a omparative study of the relationship between unemployment and self-reported health in 23 European ountries C Bambra, 1 T A Eikemo 2 ABSTRACT Bakground: The relationship between unemployment and inreased risk of morbidity and mortality is well established. However, what is less lear is whether this relationship varies between welfare states with differing levels of soial protetion for the unemployed. Methods: The first (2002) and seond (2004) waves of the representative ross-setional European Soial Survey ( respondents, aged years). Employment status was main ativity in the last 7 days. Health variables were self-reported limiting long-standing illness (LI) and fair/poor general health (PH). Data are for 23 European ountries lassified into five welfare state regimes (Sandinavian, Anglo-Saxon, Bismarkian, Southern and Eastern). Results: In all ountries, unemployed people reported higher rates of poor health (LI, PH or both) than those in employment. There were also lear differenes by welfare state regime: relative inequalities were largest in the Anglo-Saxon, Bismarkian and Sandinavian regimes. The negative health effet of unemployment was partiularly strong for women, espeially within the Anglo-Saxon (OR LI 2.73 and OR PH 2.78) and Sandinavian (OR LI 2.28 and OR PH 2.99) welfare state regimes. Disussion: The negative relationship between unemployment and health is onsistent aross Europe but varies by welfare state regime, suggesting that levels of soial protetion may indeed have a moderating influene. The espeially strong negative relationship among women may well be beause unemployed women are likely to reeive lower than average wage replaement rates. Poliy-makers attention therefore needs to be paid to inome maintenane, and espeially the extent to whih the welfare state is able to support the needs of an inreasingly feminised European workfore. The relationship between unemployment and inreased risk of morbidity and mortality is well established. 1 8 However, what is less lear from the existing literature is whether the relationship between unemployment and health varies by welfare state and, if so, the extent to whih this an be explained through referene to the different approahes to soial protetion (partiularly wage replaement rates) taken by different welfare state regimes. In this paper, we examine the extent to whih relative health inequalities between unemployed and employed people vary aross 23 European ountries and by the different approahes to soial protetion taken by the five European welfare state regimes (Sandinavian, Anglo-Saxon, Bismarkian, Southern and Eastern). Unemployment and health At the individual level, studies have partiularly shown that unemployment is assoiated with worse mental health, inluding parasuiide It has also been linked to higher rates of all-ause mortality 6 7 as well as limiting long-term illness, 8 and, in some studies, a higher prevalene of risky health behaviours (among young men), inluding problemati alohol use and smoking. 11 At the area level, rates of unemployment, espeially when used as indiators of deprivation, orrelate with poorer neighbourhood health, 12 and at the ountry level inreases in the unemployment rate have been assoiated with inreased mortality. 13 Researh has also drawn attention to the ontributory role of ill health itself as a fator behind unemployment (diret health seletion), and the importane of ill health-related worklessness in terms of soioeonomi health inequalities. 15 Studies from various ountries have identified poverty as an important intermediary fator in the relationship between unemployment and health. 7 Welfare state regimes Welfare provision, in the form of ash benefits and welfare servies, is aknowledged as an important mediatory fator in terms of the relationship between labour market position and health A ruial aspet of welfare provision, and one whih most differentiates welfare states, is inome maintenane (to prevent poverty), 18 partiularly during adverse events suh as unemployment, old age or long-term sikness absene. Welfare state regimes plae those welfare states that are the most similar (in terms of politial tradition, priniples, levels of provision, et) together, emphasising within-regime oherene and between-regime differenes. 20 There are various ompeting welfare state regime typologies whih emphasise diverse aspets of welfare states suh as soial expenditure levels, deommodifiation or politial traditions (for an overview see Bambra 22 ) Ferrera s fourfold typology, 25 whih fouses on different dimensions of how soial benefits are granted and organised, has been highlighted as one of the most empirially aurate welfare state regime typologies Ferrera 25 makes a distintion between the Sandinavian, Anglo-Saxon, Bismarkian and Southern ountries (box 1). 92 J Epidemiol Community Health 2009;63: doi: /jeh J Epidemiol Community Health: first published as /jeh on 17 Otober Downloaded from on 22 April 2018 by guest. Proteted by opyright.

2 More reently, the Eastern European ountries have begun to be onsidered as a distintive regime type. Welfare state regimes and unemployment protetion Soial protetion during unemployment varies by welfare state regime. To a large degree this reflets the historial influene of differing politial traditions, with those ountries experiening more post-war years of Soial Demorati rule providing more generous systems of support. 24 Table 1 breaks down the various harateristis of soial protetion during unemployment in the five different welfare state regimes. In essene, there are three inter-relating priniples underpinning provision: universalism, soial insurane and means-testing. 18 Systems based on universal provision do not make referene to previous ontributions or means-testing and are offered to all itizens as long as speifi demographi, soial or health riteria are fulfilled. Often flatrate benefits are paid. Under soial insurane systems, entitlement to benefits is dependent on previous ontributions and in most ases subsequent benefit levels reflet previous earned inome. Under means-testing, entitlement is restrited on the basis of inome and the (often minimal) finanial support is targeted at those in most need, usually after they have exhausted all other means (eg, personal savings or soial insurane). 31 The welfare provision of different regimes is governed by these three priniples in varying ways. For example, to differing degrees of generosity, universalism is more prominent within the Sandinavian welfare states (high population overage) and the Anglo-Saxon regime (fixed benefit rates for all), whereas soial insurane is the key omponent of provision within the Bismarkian, Southern and Eastern European welfare states. Means-testing is more ommonly a harateristi of the Anglo- Saxon welfare states; however, it is also used for soial assistane payments in other welfare state regimes. For example, in the UK (Anglo-Saxon) unemployment benefit (ontribution-based Job Seekers Allowane) is only payable (for a maximum of 6 months) to those who fulfil the minimum National Insurane ontribution requirement within the 2 years before laiming (table 1). Most laimants do not meet this riterion and are therefore reliant on means-tested soial assistane benefits, partiularly inome-based Job Seeker s Allowane and Inome Support. 32 However, this mixed approah is also evident in Sweden, where there is a soial insurane-based benefit (Unemployment Insurane Benefit) based on past ontributions and whih pays a benefit as a proportion of previous wages, as well as a means-tested soial assistane sheme (Unemployment Assistane Benefit), whih pays a (lower) flat rate. 32 Similarly, a three-tier system is operated in Germany (Bismarkian): those with a full ontribution reord reeive the full unemployment insurane benefit (Arbeitslosengeld), those with a smaller ontribution riteria, reeive a means-tested insurane benefit (Arbeitslosenhilfe) whereas those who do not have a suffiient ontribution reord must rely upon the Sozialhilfe soial assistane sheme. 32 Unemployment protetion in eah welfare state regime therefore represents a omplex mix of these differing priniples. However, there are lear differenes by welfare state regime, due to the influene of differing politial traditions, in terms of how these priniples are operationalised, partiularly in terms of the generosity of benefits paid to the unemployed (replaement rates), the qualifying period and onditions, duration of benefit payments and the waiting period before entitlement is ativated. In eah of these regards, the Sandinavian welfare states are generally more generous than the other welfare state Researh report regimes (table 1), partiularly ompared with the Anglo-Saxon and Eastern European regimes. Welfare state regimes, unemployment protetion and health Differenes in the soial protetion offered to the unemployed ould therefore be an important mediatory fator in the relationship between poverty, unemployment and health. 7 This ould be very important in terms of helping to develop poliy interventions, partiularly in terms of inome maintenane provision, to improve the health of the unemployed, redue inequalities between those in and out of work, and thereby potentially redue the influene of labour market status on health. Indeed, a study omparing means-tested and nonmeans-tested unemployment benefits in three ountries (UK, Germany, USA) found that among the unemployed, those in reeipt of non-means-tested benefits had better health than those in reeipt of means-tested benefits. 3 However, as previous studies of the relationship between unemployment and health have tended to fous either on assoiations between unemployment and health, 7 or hanges to the employment status of people and their subsequent health, within one ountry, 5 8 or in a very limited number of similar ountries, 3 a full examination of the possibly health protetive role of different approahes to soial protetion has not yet been undertaken. Similarly, although there is an emerging omparative soial epidemiology literature that examines differenes in health by welfare state regime, there has to date been little analysis by population sub-group. 22 Therefore, in this study we examine the relationship between unemployment and self-reported health in 23 ountries through referene to the different approahes to soial protetion taken by five different welfare state regimes. Speifially, given the differenes in soial provision by welfare state regime (as desribed in box 1 and table 1), we test the following two inter-related hypotheses: (1) that the self-reported health of the unemployed will be worse than the employed in all welfare state regimes, and (2) that the unemployed in those welfare state regimes with higher levels of soial protetion (the Sandinavian and Bismarkian regimes) will have omparatively better self-reported health than those in the other welfare state regimes (Anglo-Saxon, Southern and Eastern). METHODS Data The data soure is made up of two independent waves of the European Soial Survey (ESS) (merged files from 2002 and 2004), from whih we analysed individuals (aged years) from 21 ountries (table 2). The two health outome variables were self-reported limiting long-standing illness and fair/poor general health. The main objetive of the ESS is to provide high-quality data over time about hanging soial attitudes and values in Europe. The data and extensive doumentation are freely available for downloading at the Norwegian Soial Siene Data Servies (NSD) web site (www. nsd.uib.no). We used two indiators of morbidity available in the ESS: self-reported general health and limiting long-standing illness. Self-reported general health was onstruted from a variable asking How is your (physial and mental) health in general? Eligible responses were very good, good, fair, bad and very bad. We dihotomised the variable into very good or good health versus less than good health ( fair, bad and very bad ). As for limiting long-standing illness, people were J Epidemiol Community Health 2009;63: doi: /jeh J Epidemiol Community Health: first published as /jeh on 17 Otober Downloaded from on 22 April 2018 by guest. Proteted by opyright.

3 Researh report Box 1 European welfare state regimes (ranked by levels of soial protetion 1 5, high low). Adapted from Bambra 22 and Eikemo and Bambra Sandinavian Charaterised by universalism, omparatively generous soial transfers, a ommitment to full employment and inome protetion, and a strongly interventionist state. The state is used to promote soial equality through a redistributive soial seurity system. Unlike the other welfare state regimes, the Sandinavian regime type promotes an equality of the highest standards, not an equality of minimal needs and it provides highly deommodifying programmes. 2. Bismarkian Distinguished by its status-differentiating welfare programmes in whih benefits are often earnings related, administered through the employer; and geared towards maintaining existing soial patterns. The role of the family is also emphasised and the redistributive impat is minimal. However, the role of the market is marginalised. 3. Anglo-Saxon State provision of welfare is minimal, soial protetion levels are modest and often attrat strit entitlement riteria, and reipients are usually means-tested and stigmatised. In this model, the dominane of the market is enouraged both passively, by guaranteeing only a minimum, and atively, by subsidising private welfare shemes. The Anglo-Saxon welfare state regime thereby minimises the deommodifiation effets of the welfare state and a stark division exists between those, largely the poor, who rely on state aid and those who are able to afford private provision. 4. Southern The southern welfare states have been desribed as rudimentary beause they are haraterised by their fragmented system of welfare provision that whih onsists of diverse inome maintenane shemes that range from the meagre to the generous and welfare servies, partiularly, the health are system, that provide only limited and partial overage. Reliane on the family and voluntary setor is also a prominent feature. 5. Eastern The formerly Communist ountries of East Europe have experiened the demise of the universalism of the Communist welfare state and a shift towards poliies assoiated more with the Anglo-Saxon welfare state regime, notably marketisation and deentralisation. In omparison with the other member states of the European Union, they have limited welfare servies. asked if they were hampered in daily ativities in any way by any long-standing illness or disability, infirmity or mental health problem. Eligible responses were yes a lot, yes to some extent and no. We dihotomised this variable into yes (regardless of whether to some extent or a lot) and no. Unemployment was measured by omparing unemployed (inluding both those urrently looking for a job and those who are not) with people in paid work. The question asked in the survey was what is your main ativity, the last 7 days? Correlation tests between the reporting of employment status in the ESS largely orrespond with the Organisation for Eonomi Co-operation and Development (OECD) rates from 2003 (table 2, last olumn). People who were urrently under eduation, permanently sik or disabled, retired, doing ommunity or military servie were exluded from the analysis along with those doing housework/looking after hildren. A weight was applied in all analyses to orret for design effets due to sampling designs in ountries where not all individuals in the population have an idential seletion probability. All analyses were done for men and women separately. Analysis Relative health inequalities were alulated applying a series of logisti regression analyses, in whih unemployment was introdued as an independent variable, ontrolled for age, with health outomes as the dependent variables. Prevalene rates and rate differenes were alulated additionally, using diret age standardisation. In addition, to test the robustness of the main findings, three sensitivity analyses were performed. First, the between-regime differenes in the relationship between unemployment and health were tested separately for men and women using the interation employment status*regime within a multi-level design. Seond, one-way analysis of variane was used to examine whether the between-regime differenes in health outomes (overall prevalene, prevalene among unemployed, rate differene and relative inequalities) were greater than the within-regime differenes. Finally, additional adjustments were made for between-regime differenes in the prevalene of unemployment (by sex and ountry) and differenes between regimes in terms of the soioeonomi status (eduation and oupational lass) of the unemployed were also examined. These analyses are detailed further in the online appendix. RESULTS ORs (along with prevalene rates and rate differenes) of ill health are presented in table 3 for men and women within eah welfare regime separately (ountry-speifi data are presented in table 1 in the Web-only appendix). All results in this table indiate that unemployed people feel unhealthier than those who report to be employed. This assoiation is signifiant for all outomes, with the single exeption of men with limiting longstanding illness (OR 1.67) in the Anglo-Saxon welfare regime. There are also lear differenes by welfare state regime. Relative inequalities between employed and unemployed were largest in the Anglo-Saxon (men OR PH 2.97, 1.92 to 4.60; women OR LI 2.73, 1.50 to 4.95 and OR PH 2.78, 1.63 to 4.73) Bismarkian (men only OR LI 2.21, 1.74 to 2.79 and OR PH 2.72, 2.21 to 3.35) and Sandinavian (women only OR LI 2.28, 1.71 to 3.03 and OR PH 2.99, 2.34 to 4.00) regimes, and smallest in the Southern (men OR PH 1.82, 1.35 to 2.46; women OR LI 1.52, 1.03 to 2.25 and OR PH 1.66, 1.31 to 2.11) and Eastern (women only OR LI 1.65, 1.24 to 2.10 and OR PH 1.76, 1.38 to 2.25) welfare state regimes. Aording to the size of rate differenes and ORs, it appears that the negative health experienes of being unemployed are partiularly strong for women within the Anglo-Saxon (OR LI 2.73 and OR PH 2.78) and Sandinavian (OR LI 2.28 and OR PH 2.99) welfare regime. Although the ORs of men s reporting of limiting long-standing illness do not show a distint pattern (exept from the non-signifiant results in the Anglo-Saxon regime), the reporting of poor general health within the Anglo- Saxon regime again demonstrates the largest ORs. The sensitivity analyses (presented in the online appendix) show that welfare state regimes are strongly related to the assoiation of unemployment and women s health and, in terms of health outomes, that within-welfare state regime variane is signifiantly smaller than between-welfare state regime variane for measures of prevalene (but not with regard to rate differenes and relative inequalities). The assoiation between 94 J Epidemiol Community Health 2009;63: doi: /jeh J Epidemiol Community Health: first published as /jeh on 17 Otober Downloaded from on 22 April 2018 by guest. Proteted by opyright.

4 33 34 Table 1 Charateristis of unemployment protetion in 23 European ountries, ranked by welfare state regime (2004) Welfare regime (1 5, high low) Country Funding system Qualifying period* rate differenes and ORs was more evident for women than for men. The additional adjustments made for the prevalene of unemployment onfirmed the main findings, that the assoiation between unemployment and health varies by welfare state regime, and in addition, a high orrelation as found between the original ORs and the ORs adjusted for the prevalene of unemployment (r = 0.85 or higher). This suggests that it is not the higher prevalene of unemployment in some welfare states whih has driven the observed differenes in the health of the unemployed by welfare state regime. The sensitivity analyses also found that unemployed men and women were more likely to be from the lower soioeonomi groups than employed people in all welfare state regimes. DISCUSSION Our study has found that the relationship between unemployment and health is onsistent aross all 23 European ountries Initial net replaement rate (% of net average wages){ Researh report Unemployment insurane benefit duration (months){ 1. Sandinavian Denmark Subsidised voluntary 12 months in last 3 years insurane Finland Voluntary subsidised 43 weeks in last 2 years insurane and soial assistane system Norway Soial insurane Annual earnings in last year equal to % of base amount Sweden Subsidised programme 6 months in last 12 months of basi insurane and voluntary inome-related insurane 2. Bismarkian Austria Soial insurane 28 weeks in last 12 months Belgium Soial insurane 468 days in last 27 months 61 No limit 0 Frane Soial insurane and 6 months in last 22 months soial assistane Germany Soial insurane and 12 months in last 2 years soial assistane Luxembourg Soial insurane 26 weeks in last 12 months Netherlands Soial insurane and 26 weeks in last 39 weeks soial assistane Switzerland Soial insurane 12 months in last 2 years Anglo-Saxon Ireland Soial insurane and 39 weeks in last 12 months soial assistane United Kingdom Soial insurane and Contributions equivalent to 25 and soial assistane 50 times the lower earnings limit must have been paid in the last 2 years 4. Southern Europe Greee Soial insurane 125 days in last 14 months Italy Soial insurane 2 years of insurane ontributions with weeks ontributions in last 2 years Portugal Soial insurane and 540 days in last 24 months soial assistane Spain Soial insurane 12 months in last 6 years Eastern Europe Czeh Republi Soial insurane 12 months in last 3 years 56 5 Hungary Soial insurane 12 months in last 4 years Poland Soial insurane Earnings in 18 months prior to laim must be at least equivalent to the minimum wage Slovenia Soial insurane 12 months in last 18 months 56 8 Waiting period (days)1 *For unemployment insurane benefits. {Net replaement rate = (benefit inome when unemployed tax on benefit inome)/(earned inome + benefit inome when employed 2 tax on earnings and benefits) 6100; it is assumed that the unemployed worker is 40 years old and has an uninterrupted employment reord of 22 years. Benefits inluded in alulation: unemployment insurane, unemployment assistane, soial assistane, family benefits, housing benefits. {Months at equivalent to the initial rate for the Czeh Republi, the Slovak Republi and Spain, where the benefit level delines overtime (eg, for Spain, where the nominal replaement rate delines from 70% to 60% after 6 months, the month s equivalent initial rate is alulated as 6 months plus six-sevenths of 18 months). In most ountries after the insurane period ends the unemployed person is entitled to laim soial assistane (whih may be means-tested). 1No data available. with the unemployed in eah ountry reporting worse selfreported health than the employed (either LI, PH or both). This is in keeping with our first hypothesis and in line with the majority of the existing researh literature For example, a longitudinal Swedish study found that self-reported physial health dereased with the advent of unemployment and that poorer self-reported physial health inreased the likelihood of future unemployment. 5 Similarly, a longitudinal study of UK men found an inreased risk of limiting long-standing illness among the unemployed. 8 It seems, therefore, that even though the levels of soial protetion offered to the unemployed vary by welfare state (and welfare state regime), in all ountries, a relationship exists between unemployment and poorer selfrated health. This suggests that urrent wage replaement rates, even in the more generous welfare states, are not suffiient to overome the finanial effets of unemployment on health. On the other hand, it may indiate the importane for health of the non-finanial losses assoiated with unemployment (eg, soial J Epidemiol Community Health 2009;63: doi: /jeh J Epidemiol Community Health: first published as /jeh on 17 Otober Downloaded from on 22 April 2018 by guest. Proteted by opyright.

5 Researh report Table 2 Welfare regime Country statistis Country isolation), as demonstrated in Rudas et al s 40 study of unemployed Italian workers who despite reeiving a 100% replaement rate still reported elevated levels of physial and mental morbidity. 7 Although we have found a onsistent ross-european relationship between unemployment and poorer self-reported health, we have also identified differenes in the magnitude of the relationship by welfare state regime. Speifially, we have found that relative inequalities are largest in the Anglo-Saxon, Bismarkian (men only), and Sandinavian (women only) regimes, and smallest in the Southern and Eastern (women only). The findings for the Anglo-Saxon welfare state regime are perhaps unsurprising given that wage replaement rates for the unemployed are the lowest in these welfare states, and that benefits are means-tested and subjet to strit entitlement rules. The unemployed in the Anglo-Saxon welfare states are therefore at a great finanial disadvantage in omparison to those in and 2004 ombined Response rate (%) Inluded in analysis Response rate Inluded in analysis Unemployed in ESS (OECD rates 2003) Men Women (%) Men Women Total Men* Women* Denmark (5.1) 7.2 (5.7) Sandinavian Finland (9.2) 8.4 (8.9) Norway (4.8) 3.7 (3.9) Sweden (6.3) 6.7 (5.2) Austria (4.3) 3.7 (4.1) Bismarkian Belgium (7.4) 11.3 (8.0) Frane (8.8) 11.2 (11.0) Germany (9.6) 11.7 (8.8) Luxembourg (3.0) 3.7 (4.7) Netherlands (4.1) 3.9 (4.5) Switzerland (3.8) 2.7 (4.5) Ireland (4.8) 4.8 (3.9) Anglo-Saxon UK (5.5) 4.5 (4.1) Greee (6.0) 14.2 (14.3) Southern Europe Italy (6.7) 15.9 (11.6) Portugal (5.6) 11.8 (7.3) Spain (8.2) 12.9 (15.9) Czeh Republi (6.1) 10.8 (9.9) Eastern Europe Hungary (6.1) 5.7 (5.6) Poland (19.0) 16.2 (20.4) Slovenia (NA) 11.1 (NA) *Correlation between unstandardised European Soial Survey (ESS) rates and Organisation for Eonomi Co-operation and Development (OECD) rates is 0.82 for men and 0.88 for women. NA, not available. employment and this may well explain the magnitude of inequality as finanial strain has been found to be an important fator in the relationship between unemployment and ill health. Furthermore, means-tested benefits are assoiated with stigma 18 and so the non-finanial problems of unemployment may be greater in the Anglo-Saxon welfare states. Our findings are in keeping with broader based studies of welfare state regimes and health indiators whih have found that overall population health tends to be worse in the welfare states of the Anglo-Saxon regime. It is harder to explain the findings for the Bismarkian (men only), and Sandinavian (women only) regimes and ertainly these are in ontradition to the expetations outlined in our seond researh hypothesis. Interestingly, unlike the Anglo- Saxon welfare states they apply only to one or other gender. It is possible that the status differentiating Bismarkian welfare states may tend to exasperate the relationship between Table 3 Prevalene rates, rate differenes and ORs (95% CI) for eah welfare regime separately (N = ) Limiting long-standing illness Poor/fair general health Sex Welfare regime Prev (%) Unemp (%) (RD) OR (95% CI) Prev (%) Unemp (%) (RD) OR (95% CI) Men Sandinavian (13.5) 1.96 (1.47 to 2.61) (17.0) 2.27 (1.72 to 3.01) Bismarkian (12.0) 2.21 (1.74 to 2.79) (19.8) 2.72 (2.21 to 3.35) Anglo-Saxon (5.7) 1.67 (0.99 to 2.81) (16.9) 2.97 (1.92 to 4.60) Southern (6.2) 2.07 (1.34 to 3.18) (12.6) 1.82 (1.35 to 2.46) Eastern (10.8) 1.89 (1.43 to 2.52) (17.8) 2.15 (1.67 to 2.76) Women Sandinavian (17.0) 2.28 (1.71 to 3.03) (18.7) 2.99 (2.34 to 4.00) Bismarkian (9.4) 1.87 (1.48 to 2.37) (13.8) 2.06 (1.67 to 2.55) Anglo-Saxon (13.7) 2.73 (1.50 to 4.95) (14.8) 2.78 (1.63 to 4.73) Southern (4.5) 1.52 (1.03 to 2.25) (10.1) 1.66 (1.31 to 2.11) Eastern (7.0) 1.65 (1.24 to 2.19) (12.0) 1.76 (1.38 to 2.25) Prev, total prevalene; Unemp, prevalene among unemployed; RD, rate differene between employed and unemployed. All measures were age-standardised. 96 J Epidemiol Community Health 2009;63: doi: /jeh J Epidemiol Community Health: first published as /jeh on 17 Otober Downloaded from on 22 April 2018 by guest. Proteted by opyright.

6 unemployment and poor health by restriting aess to the higher level soial insurane benefits. The length of entitlement to soial insurane is also omparatively low in the Bismarkian welfare states (table 1). That relative inequalities are greater between men than women, may also be in part due to stigma as the familial approah of the Bismarkian welfare states emphasises the male breadwinner role. In terms of the Sandinavian welfare state regime, the relatively large inequalities between employed and unemployed women may well reflet the fat that women are less likely to meet the qualifiation riteria for soial insurane payments (for example due to higher rates of part-time working) 47 and are therefore dependent on soial assistane benefits whih have a lower overall replaement rate. 48 It is of interest that the smallest relative inequalities between employed and unemployed were found in the Southern and Eastern welfare states. For example, the health differenes between unemployed and employed people in the East European welfare regime were never larger than OR = 2.15 throughout the study (table 3). This is somewhat ounter to the wider inequalities in health literature, whih suggests that relative inequalities in health by soioeonomi status should be larger in these ountries. This finding is therefore very surprising and learly requires further analysis (perhaps looking at individual ountries in these regimes in more depth), not least as the replaement rates and eligibility riteria for the Southern and Eastern welfare state regimes are not partiularly generous, holding a fairly moderate position in relation to other regimes (table 1). One possible explanation for the finding is that the more traditional family model in these ountries means that additional material, and non-material, support is provided by the family to unemployed members thus buffering the impat of unemployment on health. Our main results and the sensitivity analyses also suggest that there is an important gender dimension to the relationship between unemployment and poorer self-reported health. Health inequalities between the unemployed and employed were larger among women, most strikingly in the Anglo-Saxon and Sandinavian welfare state regimes. First, this is in ontrast to most single ountry, longitudinal studies, in whih the relationship between unemployment and poor health has generally found to be more notieable among men. Caution should therefore be applied to our findings until they are repliated. However, from a soial protetion perspetive it is less surprising that women experiene a more adverse impat on health of unemployment. Women are often not entitled to the higher value soial insurane benefits, due to a less oherent employment history, for example, part-time work, periods out of work due to aring et, and therefore have to rely on lower level soial assistane, whih provides muh lower replaement rates, even in the more generous Sandinavian welfare states. It is also possible that the seletion effet is stronger for women than men, that is that unhealthy women are more likely to beome unemployed than unhealthy men. Future researh learly needs to explore further the relationship between women, unemployment and health, and the role that the welfare state an play in supporting the needs of an inreasingly feminised European workfore. Limitations Although the ESS presents an outstanding opportunity to investigate ross-national patterns of health inequality, as the survey asks the same questions at the same time in all ountries, we aknowledge that there are many issues whih may affet What is already known on this subjet Single-ountry studies have shown that unemployment is assoiated with worse morbidity and mortality. Poverty may be an important mediatory fator in this relationship. Different types of European welfare state (welfare state regimes) offer different levels of soial protetion to the unemployed. Unemployed people in reeipt of means-tested benefits have worse health than those in reeipt of entitlement benefits. What this study adds This study examines whether the relationship between unemployment and health varies by European welfare state regime and, if so, the extent to whih this an this be explained through referene to the different types of soial protetion. The negative relationship between unemployment and health is onsistent aross Europe but varies by welfare state regime, suggesting that levels of soial protetion may indeed have a moderating influene. The negative relationship is partiularly strong among women and in those ountries with low replaement rates and whih utilise means-tested benefits. Poliy impliations Researh report Unemployment has a negative relationship with health; this may in part be due to the loss of inome assoiated with unemployment. Inome levels for the unemployed therefore need to be adequate to prevent health damage. Relative health inequalities between the employed and unemployed were greatest in those welfare states that utilised means-tested benefits. Welfare state arrangements need to be more sensitive to moderating the effets of unemployment on the health of women, partiularly as the European workfore is beoming inreasingly feminised. the omparability of multi-ountry studies, suh as variations in response rate (table 2), modes of data olletion, translations, ultural interpretation and ondut. Our study is further limited by utilising only self-reported health measures whih may vary by ountry, soioeonomi or employment status and/or ulture. For example, an unemployed immigrant living in Spain may use different riteria to define his or her health than an unemployed Finn living in Finland. However, studies have found a strong relationship, whih does not vary by soioeonomi status, 51 between self-reported health and mortality. 52 Similarly, the measure of unemployment (unemployed in the last 7 days) may obsure important between ountry differenes in the omposition of the unemployed population (online appendix). Further, the omplex nature of the relationship between unemployment and health means that, despite onduting a number of sensitivity analyses, we J Epidemiol Community Health 2009;63: doi: /jeh J Epidemiol Community Health: first published as /jeh on 17 Otober Downloaded from on 22 April 2018 by guest. Proteted by opyright.

7 Researh report have not overed all the possible fators influening between ountry differenes. Another possible limitation is our hoie of welfare state regime typology. As noted in the introdution, there is a multitude of ompeting welfare state regime typologies and no ategorisation has yet been generally aepted as the standard typology (although Ferrera s is one of the most aurate in terms of how soial benefits are granted and organised). We also arried out a number of sensitivity analyses. However, it must be aknowledged that if the typologies of other authors were used it may have resulted in different results. Finally, as the ESS data is ross-setional, we annot rule out seletion effets. Competing interests: None delared. REFERENCES 1. Martikainen P, Valkonen T. Exess mortality of unemployed men and women during a period of rapidly inreasing unemployment. Lanet 1996;348: Martikainen P, Valkonen T. The effets of differential unemployment rate inreases of oupation groups on hanges in mortality. Am J Publi Health 1998;88: Rodriguez E. Keeping the unemployed healthy: the effet of means-tested and entitlement benefits in Britain, Germany and the United States. Am J Publi Health 2001;91: Bartley M. Unemployment and ill health: understanding the relationship. J Epidemiol Community Health 1994;48: Korpi T. Aumulating disadvantage: longitudinal analyses of unemployment and physial health in representative samples of the Swedish population. Eur Soiologial Rev 2001;17: Morris JK, Cook DG, Shaper AG. Loss of employment and mortality. BMJ 1994;308: Bartley M, Ferrie J, Montgomery SM. Health and labour market disadvantage: unemployment, non-employment, and job inseurity. In Marmot M, Wilkinson RG, eds. Soial determinants of health. Oxford: Oxford University Press, 2006: Bartley M, Plewis I. Aumulated labour market disadvantage and limiting long-term illness: data from the ONS longitudinal study. Int J Epidemiol 2002;31: Montgomery SM, Cook DG, Bartley M, et al. Unemployment pre-dates symptoms of depression and anxiety resulting in medial onsultation in young men. Int J Epidemiol 1999;28: Platt S. Parasuiide and unemployment. Br J Psyhiatry 1986;149: Montgomery SM, Cook DG, Bartley M, et al. Unemployment, igarette smoking, alohol onsumption and body weight in young British men. Eur J Publi Health 1999;8: Stafford M, MCarthy M. Neighbourhoods, housing and health. In Marmot M, Wilkinson RG, eds. Soial determinants of health. Oxford: Oxford University Press, Brenner H. Politial eonomy and health. In Amik B, ed. Soiety and health. Oxford: Oxford University Press, 1995: Jusot F, Khlat M, Rohereau T. Job loss from poor health, smoking and obesity: a national prospetive survey in Frane. J Epidemiol Community Health 2008;62: Bambra C, Pope D. What are the effets of anti-disriminatory legislation on soioeonomi inequalities in the employment onsequenes of ill health and disability? J Epidemiol Community Health 2007;61: Bartley M, Blane D. Soioeonomi determinants of health: health and the life ourse: why safety nets matter. BMJ 1997;314: Dahl E, Fritzell J, Lahelma E, et al. Welfare state regimes and health inequalities. In Siegrist J, Marmot M, eds. Soial inequalities in health. Oxford: Oxford University Press, 2006: Diderihsen F. Impat of inome maintenane poliies. In Makenbah J, Bakker M, eds. Reduing inequalities in health: a European perspetive. London: Routledge, 2002: Lahelma E, Arber S. Health inequalities among men and women in ontrasting welfare states: Britain and three Nordi ountries ompared. Eur J Publi Health 1994;4: Eikemo TA, Bambra C. The welfare state: a glossary for publi health. J Epidemiol Community Health 2008;62: Bonoli G. Classifying welfare states: A two-dimension approah. J So Poliy 1997;26: Bambra C. Going beyond the three worlds of welfare apitalism: regime theory and publi health researh. J Epidemiol Community Health 2007;61: Esping-Andersen G. The three worlds of welfare apitalism. London: Polity, Navarro V, Shi L. The politial ontext of soial inequalities and health. Int J Health Serv 2001;31: Ferrera M. The southern model of welfare in soial Europe. J Eur So Poliy 1996;6: Eikemo T, Bambra C, Judge K, et al. Welfare state regimes and differenes in selfpereived health in Europe: a multi-level analysis. So Si Med 2008;66: Bambra C. Sifting the wheat from the haff: A two-dimensional disriminant analysis of welfare state regime theory. So Poliy Adm 2007;41: Bambra C, Pope D, Stanistreet D, et al. Gender, health inequality and welfare state regimes: a ross-national study of twelve European ountries. J Epidemiol Community Health doi: /jeh Eikemo TA, Huisman M, Bambra C, et al. Health inequalities aording to eduational level under different welfare regimes: a omparison of 23 European ountries. Soiol Health Illn 2008;30: Esping-Andersen G. Soial foundations of post-industrial eonomies. Oxford: Oxford University Press, Rhodes M. The Welfare State: internal hallenges, external onstraints. In Rhodes M, Vinent A, eds. Developments in Western European politis. London: Mamillan, Eurostat. Soial protetion in the EU member states. Luxembourg: Eurostat, Organisation for Eonomi Co-operation and Development. Benefits and wages: OECD indiators. Paris: OECD, International Soial Seurity Assoiation. Soial seurity programmes throughout the world. Geneva: ISSA, Bambra C. Health status and the worlds of welfare. So Poliy Soiety 2006;5: Chung H, Muntaner C. Welfare state matters: A typologial multilevel analysis of wealthy ountries. Health Poliy 2007;80: Martikainen P, Lahelma E, Marmot M, et al. A omparison of soioeonomi differenes in physial funtioning and pereived health among male and female employees in Britain, Finland and Japan. So Si Med 2004;59: Navarro V, Muntaner C, Borrell C, et al. Politis and health outomes. Lanet 2006;368: Raphael D, Bryant T. The welfare state as a determinant of women s health: support for women s quality of life in Canada and four omparison nations. Health Poliy 2004;68: Rudas N, Tondo L, Musio A, et al. Unemployment and depression: results of a psyhometri evaluation. Minerva Psihiatr 1991;32: Kessler RC, Turner JB, House JS. Intervening proesses in the relationship between unemployment and health. Psyhol Med 1987;17: Jakson P, Warr P. Unemployment and psyhologial ill-health: the moderating role of duration and age. Psyhol Med 1984;14: Coburn D. Beyond the inome inequality hypothesis: lass, neo-liberalism, and health inequalities. So Si Med 2004;58: Conley D, Springer K. Welfare states and infant mortality. AJS 2001;107: Bambra C. The worlds of welfare: Illusory and gender blind? So Poliy Soiety 2004;3: Bambra C. Defamilisation and welfare state regimes: A luster analysis. Int J So Welf 2007: European Foundation for the Improvement of Living and Working Conditions. Part-time work in Europe 2004 [ited ]; eu/ewo/reports/tn0403tr01/tn0403tr01_2.htm (aessed 21 Ot 2008). 48. Saarela J. Replaement rates and labour market behaviour. Soioeonom Plann Si 2006;40: Leinsalu M, Vagero D, Kunst A. Estonia : enormous inrease in mortality differenes by eduation. Int J Epidemiol 2003;32: Kunst A, Groenhof F, Anderson O. Oupational lass and ishemi heart disease mortality in the United States and 11 European ountries. Am J Publi Health 1999;89: Burstrom B, Fredlund P. Self rated health: Is it as good a preditor of subsequent mortality among adults in lower as well as in higher soial lasses? J Epidemiol Community Health 2001;55: Idler EL, Benyamini Y. Self-rated health and mortality: A review of twenty-seven ommunity studies. J Health So Behav 1997;38: J Epidemiol Community Health 2009;63: doi: /jeh J Epidemiol Community Health: first published as /jeh on 17 Otober Downloaded from on 22 April 2018 by guest. Proteted by opyright.