Deprivation, class and crisis in Europe: a comparative analysis Open Access

Analysing data from an original cross-national survey conducted in 2015 in nine European democracies covering five different types of welfare regime and asking individuals a variety of questions on their deprivation during the crisis, this paper shows that there are important cross-national and cross-class inequalities in deprivation as reported by individuals in different social classes. Cross-nationally, deprivation patterns reflected the welfare regimes of the nine countries as well as the severity of the economic crisis. Working class individuals in countries that were not so deeply affected by the crisis were generally found to be worse off than middle class individuals in countries that were more deeply affected. Semi/unskilled manual classes were found to be the most deprived and class differentials were diminished but not accounted for in multilevel models including a series of controls linked to risk factors and socio-demographic position. At the macro-level, higher inequality as measured by the Gini coefficient was associated with higher levels of reported deprivation. However, cross-level interaction tests did not provide evidence that being in semi/unskilled manual occupations has a further heightened effect on reported deprivation in contexts of higher inequality or lower social spending.

Inequality has been steadily increasing in advanced societies (Piketty, 2014, Nolan and Whelan, 2011, Musterd and Ostendorf, 2013, Dorling, 2014, Atkinson, 2015). Intimately related to the debates on rising inequality are debates on the extent of inequalities linked to social class. In fast-changing societies, multiple sources of disadvantage overlap to marginalise deprived groups. In this paper we examine occupational class in relation to the lived experience of deprivation in the current economic crisis across nine European democracies representing five different welfare regimes. Our specific aim is in analysing in a comparative European perspective the influence of social class in the perception of material deprivation. We look at both cross-national and within-country social class differences in reported deprivation.

Recent scholarship has emphasised the utility of non-monetary indicators of deprivation for identifying the poor as well as to more fully capture the wider aspects of deprivation, disadvantage and social exclusion (Nolan and Whelan 2011). Whereas previous research has tended to employ data from the European Community Household Panel Survey (ECHP, running from 1994 to 2001) and European Union Statistics on Income And Living Conditions (EU-SILC, running from 2003 to 2011), in this study we exploit data from a rich, original survey conducted in 2015 in nine European countries. This allows for analysing the most recent trends in reported deprivation levels as well as cutbacks in consumption and difficulties keeping up payments in terms of class differentials within countries as well as between countries during the latest crisis period. Moreover, given that our survey is cross-national and asks standard questions on reported deprivation across countries this allows for comparing countries from different types of welfare regimes and which experienced different degrees of economic crisis. This type of analysis allows us to make sense of the way in which citizens in different social groups perceived deprivation during the course of the current crisis as well as looking at inequalities between different classes in reported patterns across European countries.

To analyse these questions, we utilise data collected through an original European cross-national survey (N ∼ 18,000) in nine democracies representing five different types of welfare regimes. This survey was designed specifically with our research questions in mind and contains multiple, nuanced indicators of deprivation experiences in times of crisis as well as the relevant individual-level risk factors which we include in multi-level models. Our multilevel models also control for country-level social spending as well as inequality as measured through the Gini coefficient and including cross-level interactions with semi/unskilled manual occupation to test whether inequalities in reported deprivation are exacerbated in contexts marked by lower social spending and higher inequality. This analysis allows us to test for our theoretically-informed hypotheses with respect to the patterns of within-country cross-class and cross-national inequalities in reported deprivation expected based on previous research looking at the European Union in comparative perspective (Nolan and Whelan 2011). In what follows, first we discuss previous literature on deprivation and advances in the study of class. Next we discuss our data and methods. We then present our results and finally conclude with a summary of our key results on cross-national and class-based inequalities during the crisis.

For a while now the literature on poverty has emphasised that non-monetary measures of deprivation can play an important role for developing our understanding of people's lived experience as well as developing more effective anti-poverty strategies (Nolan and Whelan 2011). Using cross-nationally comparative indicators is crucial when performing comparative analyses (Nolan and Whelan 2011). Using non-monetary indicators provides a clear comparative measure of deprivation cross-nationally. Poverty research uses the definition that people are in poverty when ‘their resources are so seriously below those commanded by the average individual or family that they are, in effect, excluded from ordinary living patterns, customs and activities’ (Townsend 1979: 31). In the United States this is defined as insufficient resources for basic living needs, defined appropriately for the United States today (Citro and Michael 1995). This suggests two core elements of poverty: the inability to participate and the fact that the latter is attributed to inadequate resources (Nolan and Whelan 2011). As Nolan and Whelan (2011) emphasise, in parallel to a large literature e.g. Atkinson et al. (1995) or the Growing Unequal OECD study (2008) which has debated and developed methods to establish income cut-offs to distinguish the poor, non-monetary indicators of deprivation and living standards have also been studied for many years. This focus emerged from Townsend’s (1979) pioneering work on the use of non-monetary indicators of deprivation to show ‘what it meant to be poor in Britain at the time in terms of deprivation of everyday items and activities widely regarded as essential’ and the key point that ‘low income could be used to identify the poor but did not tell us all we need to know about what it was like to be poor and how people arrived in and coped with that situation’ (Nolan and Whelan 2011: 2). A more radical critique of income was put forward by others noting that it failed to identify those unable to participate in society due to lack of resources (Nolan and Whelan 2011: 2). For example, Ringen (1987, 1988) argued that income did not adequately capture poverty as it was both unreliable and indirect a measure; Mack and Lansley (1985) preferred to employ deprivation indicators directly to capture social exclusion in Britain, starting a tradition followed by further British ‘poverty and social exclusion’ studies (Gordon et al. 2000, Pantazis et al. 2006). Other studies identified the ‘consistently poor’ as those both on low income and reporting deprivation in basic items (Callan et al. 1993, Nolan and Whelan 1996) which is also the approach used by the UK combining low income and material deprivation in a range of indicators to monitor child poverty (DWP 2003). Bradshaw and Finch (2003) also looked at ‘core poverty’ – those reporting their own financial situation as very difficult alongside low income and other forms of deprivation. This discussion illustrates the long tradition of using non-monetary indicators as standalone measures as well as in a variety of combinations to gauge deprivation in many European nations and cross-nationally (Nolan and Whelan 2011).

In particular, one of the key advantages of this approach is that it highlights the ways in which poverty and deprivation are ‘not just about money’ and how social exclusion involves poverty which is not just a financial matter of low resources but is also linked to other forms of disadvantage such as in educational opportunities, poor health/access to health services, inadequate housing, as well as exclusion from the labour market (Burchartdt et al. 2002, Nolan and Whelan 2007, 2011). In turn, this recognition has meant that there has been a new focus on measuring and monitoring key dimensions of disadvantage and well-being (Bradshaw and Finch 2003, Boarini and Mira d’Ercole 2006). Indeed, in Europe, the definition of poverty formulated by Townsend (1979) is now widely employed and has also been adopted by the European Union (Nolan and Whelan 2011). The European Council's own definition states that ‘the poor shall be taken to mean persons, families and groups of persons whose resources (material, cultural, and social) are so limited as to exclude them from the minimum acceptable way of life in the Member State in which they live’ (EEC 1985). This definition underlies the EU's Social Inclusion Process which joins member states working to tackle poverty and exclusion through the ‘open method of coordination’ by agreeing common objectives, national plans to promote social inclusion and joint reports by the Commission and Council (Nolan and Whelan 2011).

In this context, an explicitly multidimensional approach to monitoring social inclusion which includes non-monetary indicators has become particularly salient with the EU enlargement since 2004 given that the inclusion of countries with lower living standards has made it much harder to make sense of deprivation cross-nationally (Alber et al. 2007, Kogan et al. 2008). With enlargement, the contrasts between richer and poorer member states by average pro capita income are now much wider and the income poverty thresholds that had been adopted for the richer countries are higher than average income in the poorer ones so that those living in poverty in richer countries have higher standards of living than the better off in the poorest nations (Nolan and Whelan 2011). In this way the ‘at risk of poverty’ and average income pro capita estimates yield widely different pictures so that while the EU strategy has tended to tackle within and between country divergences in living standards as separate issues, there is a deep need for more studies examining cross-nationally comparative non-monetary indicators of deprivation (e.g. Nolan and Whelan 2011). Moreover, this is further an important area of study given also the recent context of economic crisis in Europe (Giugni and Grasso 2018).

Material deprivation indicators are particularly useful when looking at cross-national differences and for examining patterns by class as we do in this study (for a detailed discussion on this see Nolan and Whelan 2011). Given that in this paper we are particularly interested in examining material deprivation during the period of the crisis, we analyse primarily, with original survey data from 2015, whether respondents felt that their household economic condition had deteriorated in the last five years (i.e. since 2010). Moreover, we also analyse an indicator which asks individuals whether they had to reduce the consumption of staple foods in past 5 years for financial/economic reasons. Finally, we analyse an indicator that asks whether they have been struggling with bills. These variables are similar to the material deprivation indicators traditionally used in the literature – particularly those on being able to pay unexpected required expenses, afford consumer durables or whether the household had been in arrears on payments and repayments- based on data analysis of the EU-SILC and the material deprivation indicator included within Laeken indicators adopted by the EU to monitor common progress on poverty and social inclusion since the 2010 outset of the Europe 2020 strategy, with a headline poverty target on reducing by 20 million in 2020 the number of people under poverty and social exclusion.

In particular, in previous research reporting on the patterns of poverty both cross-nationally by welfare regime and by class based on various deprivation measures in the European Community Household Panel Survey and European Union Statistics on Income and Living Conditions, Nolan and Whelan (2011) showed the consequences of different welfare regime arrangements for reported deprivation levels. They also showed that economic vulnerability profiles vary across welfare regimes and therefore different types of welfare regimes – defined by Gallie and Paugam (2000: 3–4) as systems of public regulation that are concerned to assure the protection of individuals and to maintain social cohesion by intervening through both legal measures and the distribution of resources – show different patterns of deprivation. These types of welfare regimes developed by combining Bukodi and Robert's (2007) criteria for the strictness of employment protection legislation (EPL) with those reflected in Esping-Andersen’s (1990) distinction between three ‘worlds of welfare capitalism’ (see further Ferrera 1993, 1996, Bonoli and Palier 2001) are as follows (Nolan and Whelan 2011: 104): (1) The social democratic regime (e.g. Sweden) which assigns the welfare state an important redistributive role; (2) The corporatist regime (e.g. France, Germany and Switzerland) which places less emphasis on redistribution and more on rights to benefits depending on labour market contributions; (3) The liberal regime (e.g. the UK) which emphasises the primacy of the market and sees the state as having a residual welfare role; (4) The southern European regime (e.g. Greece, Italy and Spain) which is characterised by family support systems with poor labour market policies and uneven benefit systems; (5) The post-socialist corporatist regime (e.g. Poland) with transfer-oriented labour market measures and moderate employment protection. (The post-socialist liberal cluster in the Baltic countries have more flexible labour markets and weaker employment protection and are identified as a further group but our study does not include this regime.) Nolan and Whelan (2011) note how the social democratic regime offers a comprehensive coverage and how Maitre et al. (2005) had showed that the proportions of households lifted out of poverty was highest for this regime. They also present a rich discussion of the other types of regimes and their expected deprivation rates relative to each other (Nolan and Whelan 2011: 125–6). This leads to our first hypothesis:

Moreover, socio-economic differentiation patterns will be different from one regime to another (Nolan and Whelan 2011). Thus, we advance the following hypotheses of the type of patterns that we expect cross-nationally for differences between classes, or within-country inequalities which we will capture in the empirical analysis as the ratio of deprivation between the top, professional managerial class and the lowest, semi/unskilled manual class: As noted above, we focus on socio-economic variation based on social class following Nolan and Whelan (2011: 146) who argue that ‘the ongoing dispute relating to its importance can be further clarified by comparative analysis’ (Atkinson 2007, Beck 2007, Goldthorpe 2007, 2010). As such we expect that: It is clear that there is an important overlap here between welfare regimes and typologies of social spending or levels of inequality and as such we control for these as level 2 variables in our multi-level models to capture whether perceived deprivation is higher in contexts marked by greater inequality or lower levels of social spending. At an aggregate level, these cross-national differences can be captured by the use of measures of social spending i.e. the extent to which nations spend on social services and the Gini coefficient of inequality. Based on this we expect that: Moreover, we expect that more unequal national contexts and those characterised by lower levels of social spending will exacerbate class differentials and in other words the likelihood that those in semi/unskilled manual occupations will report deprivation: Furthermore, scholarship has shown that apart from occupational class also other individual-level factors make people more at risk of deprivation. The seminal studies on class focused on the extent of stratification in society and on issues of social mobility (Erikson et al., 1979). Traditionally, class has been understood through occupational status and a worker's position relative to the means of production; of key interest was the understanding the extent of social mobility and its impact on the working class (Heath 1981). Goldthorpe et al. (1967) and others set out specifically to consider the embourgeoisement thesis (Ryan and Maxwell 2016). Goldthorpe et al. (1967) argued that despite increased affluence over time, manual workers in their study (i.e. the working class) still experienced lower mobility than non-manual/middle class individuals. This was found with respect to income earned and also relative to the nature of work: repetitive forms of work with little hope for promotion or supervisory roles. Moreover, in terms of sociability, manual workers’ networks had remained narrow and limited to family members and a few other working class contacts. Most importantly, with respect to the thesis of embourgeoisement, these workers did not express views showing that they now saw the Conservative party as representing their needs. The key conclusions of this study were that positions in a stratified hierarchy were not solely based on income or possessions but more widely in terms of life-chances, experiences and the nature of relationships with other groups (Goldthorpe et al., 1967: 27): the crucial distinction remaining that manual workers must sell their labour for income.

Goldthorpe et al. (1967) largely adopted a neo-Weberian approach to class and stratification. Others such as Crompton (1987) argued in favour of a Marxist analysis of class for understanding white collar workers or the ‘propertyless middle class’ that were neither proletariat nor bourgeoisie and the ways in which the expansion of the middle class had challenged the traditional distinction between manual and non-manual workers in Western societies. A similar argument has been echoed more recently in work on ‘the precariat’ and the argument that new sources of inequalities are not captured through traditional distinctions (Standing 2011).

Above all, an understanding of class is linked to questions of inequality since classes are understood in relation to one another in a system of hierarchy and stratification. Classes are distinguished by the nature of people's employment relationships (e.g. employers and employees), the nature of the wage contract and life chances (Goldthorpe 2000). It remains clear that questions of class differences with respect to the extent to which classes have to deal with the negative effects of economic crisis for example have critical implications in terms of their relative well-being and life chances. While more cultural approaches to class have also been proposed to study deprivation, it remains critical, as argued by many (e.g. Devine and Savage 2000, Savage and Williams 2008), to examine how class inequalities drive material deprivation in contemporary European societies.

A further contribution of our study is to control by social groupings other than class and analyse the extent to which risk factors which make various groups more vulnerable to having experienced a deterioration in financial conditions as a result of the current crisis are associated with or account for the effect of class when we control for them in subsequent models. Moreover, as noted above we also test whether country-level inequality and levels of social spending exacerbate class inequalities in this respect. We include important controls pertaining to socio-demographic dimensions discussed in the literature such as gender (Skeggs 2004), generation (Chauvel 2006) and education (Vincent et al. 2013). The literature tends to argue that austerity spending cuts linked to economic crisis tend to be most damaging for women since they are generally more likely to be in caring roles and to use social services (Stacey 1981, Women’s Budget Group 2015). Moreover, the literature has emphasised the economic difficulties that young generations are experiencing in relation to their parents (Chauvel 2006) and higher levels of education are seen as a means to attenuate class differentials in material outcomes (Vincent et al. 2013). Additionally, as is well known, issues of class inequality are intermingled with other sources of poverty and multiple deprivation relating to type of occupation and health. Indeed, poverty and deprivation have been shown to be associated with higher mortality and morbidity rates and lower life expectancy as well as with work in unsafe occupations and the more likely exposure to toxic sites (Seccombe 2002). The literature on health inequalities clearly shows that both subjective and objective measures of deprivation are linked to health outcomes (Weitz 2001). Furthermore, deprivation is also associated with a greater likelihood that one will be living alone and not be married or have children and have lower levels of social contact since it diminishes the chances that one has to marry given economic insecurity makes marriage less attractive (Wilson 1996). Furthermore, deprivation and other types of hardship such as unemployment and precarious work conditions have been shown to undermine marriages (Conger et al. 1999). Conger et al. (1994) suggested that hardship leads to depression which in turn contributes to more challenging marital relationships and dissatisfaction. More generally, scholarship has highlighted different types of individual-level factors which might mitigate the risk of deprivation: (1) individual level factors such as personality and dispositions e.g. good communication/problem-solving skills and self-efficacy such as those provided by a good education, good mental and physical health (Garmezy 1991); (2) family factors that might allow shielding from the more negative effects of deprivation e.g. companionship, social contact and support which can shape a family's ability to endure in the face of risk factors (Seccombe 2002); (3) community factors e.g. wider webs of social contacts (Bowen et al. 2000). In situations of deprivation, social ties can serve as almost a form of informal insurance, providing financial help, and physical assistance (Aldrich 2010). Money-lending, a place to stay, help with looking after the children and information are all resources that individuals can rely on their friends to provide even when it may not be accessible from organisations such as the local government, professional childcare services, and other institutions (Aldrich 2010).

We use an original and rich new source of data from 2015 which allows us to capture cross-national and cross-class reported deprivation during the economic crisis in Europe. More specifically, in order to test our hypotheses we rely on data from an original cross-national survey conducted in 2015 in the context of the Living with Hard Times: How European Citizens Deal with Economic Crises and Their Social and Political Consequences (LIVEWHAT) project funded by the European Commission under the auspices of their 7th Framework Programme (grant agreement number 613237). The survey was conducted in nine European countries (for a total N of approximately 18,000 respondents with approximately 2,000 N per country): France, Germany, Greece, Italy, Poland, Spain, Sweden, Switzerland and the UK by a specialised polling agency (YouGov) using online panels with the methodologies available in each country and quota balanced to match national population statistics in terms of region, sex, age, and education level. Given the strong association between education and social class this would support the adequate observation of social class. Moreover, the country cases conveniently cover all welfare regime typologies discussed in the theory section with the exception of the liberal variant of the post-socialist model of the Baltic. The total final sample consisted of 17,629 individuals once missing cases were deleted.

As detailed in the discussion section, most studies of deprivation have tended to use the ECHP and EU-SILC datasets. These do not include indicators relating specifically to deterioration in household living standards or the period of the crisis. Moreover, given the data is at the household level in these studies our individual level survey allows to control for further individual level risk factors associated with deprivation to test whether class differentials can be explained by these.

Our main dependent variable is reported household deprivation in the last five years. This variable asks individuals whether their household economic situation had deteriorated in the last five years (i.e. between 2010 and 2015). We also examine two further measures of reported deprivation: whether individuals had to reduce the consumption of staple foods for economic reasons in the past 5 years, and whether they are struggling to keep up with bills.

As noted in our theoretical section, given the continued importance for socio-economic differentiation, our main independent variable is the social class of the chief wage earner. The eight occupational classes investigated are as follows: 1. Professional or higher technical work – work that requires at least degree-level qualifications (e.g. doctor, accountant, schoolteacher, university lecturer, social worker, systems analyst); 2. Manager or senior administrator (e.g. company director, finance manager, personnel manager, senior sales manager, senior local government officer); 3. Clerical (e.g. clerk, secretary); 4. Sales or services (e.g. commercial traveller, shop assistant, nursery nurse, care assistant, paramedic); 5. Foreman or supervisor of other workers (e.g building site foreman, supervisor of cleaning workers); 6. Skilled manual work (e.g. plumber, electrician, fitter); 7. Semi-skilled or unskilled manual work (e.g. machine operator, assembler, postman, waitress, cleaner, labourer, driver, bar-worker, call centre worker); 8. Other (e.g. farming, military).

As justified in the theoretical section we also include controls for gender, generation, education level, employment status, health, whether the respondent lived alone or had children at home as well as frequency of social contact with friends and participation in associations. To account for structural effects on reported deprivation, we include measures of social spending and inequality (Gini coefficient) at the aggregate level and furthermore to examine whether this has implications for class-based inequalities by conducting cross-level interactions tests. Variable descriptive statistics are presented in Table 1.

Our dependent variable is measured at the individual level. However, our respondents are nested in their respective countries, so to capture the hierarchical structure of the data, we specify multilevel models to take into account the two-level nature of the data (country and individual). This type of model is useful to correct for the within-country dependence of observations (intraclass correlation) and the clustered nature of the data. Since our dependent variable is dichotomous, we estimate logistic multilevel models with a Gaussian link function. In the results section below, After presenting the descriptive results by class and country to test whether patterns reflect H1-2 on cross-national differences and social-differentiation patterns, we then apply a more analytical strategy and specify nine nested multilevel models including subsequently in the five first models a greater number of controls to test the resilience of class differentials to various factors that tend to be associated with deprivation and social exclusion as discussed in the theory section, to test for H3-4. In the last four models we include the level 2 controls to test for H5-6 and their respective cross-level interactions with semi/unskilled manual occupational class to test for H7-8.

Deprivation can be understood in absolute terms, as a proportion of individuals in a given class that reported deprivation. However, deprivation can also be understood in relative terms, or as inequality, in terms of the proportion of individuals in one class that reported deprivation relative to individuals in other classes. As detailed in H1 and H2 we expect different patterns based on welfare regimes cross-nationally. As such in what follows we comment on both overall and relative results. The first concern of our analysis is to look at the implications of class inequalities for deprivation cross-nationally. Table 2 shows the proportion of individuals in each social class that reported household level deprivation in terms of household economic conditions having deteriorated in the last five years (i.e. between 2010 and 2015). Examining the data in relation to H1 the lowest levels of reported deprivation are to be found, as expected, in the social democratic regime (Sweden), while the next lowest levels are found, also as hypothesised, in two corporatist regimes (i.e. Germany and Switzerland). However, against H1, the third corporatist regime, France, exhibits higher levels of reported deprivation than the liberal regime (UK) as well as the post-socialist corporatist regime (Poland), and more akin to the higher levels reported in the southern European regimes of Italy and Spain but not as high as Greece. As such we find mixed evidence with respect to H1 for reported deprivation: countries that experienced a deeper economic crisis relative to the others in their welfare regime group stand out with higher levels of reported deprivation and the southern European regime countries report higher deprivation than the corporatist post-socialist regime despite the predictions of H1-presumably also linked to the fact that in this bloc the crisis was deeper than in Poland. Thus, the reported deterioration indicator shows that while patterns broadly fit those expected in H1 there is some movement in the expected ranking relative to the depth of the latest economic crisis.

If we examine the evidence for H1 with respect to the indicators on the reduced consumption of staple foods in the past 5 years for economic reasons reported in Table 3, we can see that here the patterns largely reflect those found above though overall absolute levels are slightly lower. The social democratic regime and the two corporatist regimes (Germany and Switzerland), as well as the liberal regime exhibit the lowest levels of deprivation, but France exhibits higher levels, closer to those reported in some of the southern European regimes (Italy) and the post-socialist corporatist regime (Poland) which according to theory should have shown the highest levels of deprivation. Rather, levels of deprivation in Italy and Greece as well as France (two southern European and one corporatist regime) are higher here suggesting that at the deeper economic crisis may have contributed to this slightly different ranking relative to the hypothesised expectations.

Finally, examining the evidence for H1 with respect to the third indicator that reports those saying their household is struggling with bills more generally as presented in Table 4, we can see that H1 is supported to some extent: the social democratic regime (Sweden) exhibits the lowest levels, this is followed by one corporatist regime (Germany) and then the liberal regime (UK). However, the other two corporatist regimes (Switzerland and France) display higher levels of struggling with bills even relative to southern European regimes (Spain) and the post-socialist corporatist variant (Poland). The highest levels of reported financial difficulties are once more found in southern European regimes which were also more deeply affected by the economic crisis (Italy and Greece).

With respect to the evidence from these three indicators for H2 on social differentiation patterns within countries – where the patterns are expected to be the same as for H1 with the variation that here liberal regimes should have the highest levels of inequality – we find that while there is some evidence for this with respect to the indicator for the reduced consumption of stable foods (Table 3), by and large patterns do not confirm H2. The highest levels of inequality as captured by the ratio between those in the semi/unskilled manual class and those in the professional class for the reported household deprivation measure are found in the social democratic regime. However, it should be noted here that levels of reported deprivation are much lower than in the other countries. Even amongst the semi/unskilled manual class only 32 percent report deprivation (relative to 16 percent in the professional class) whereas in the southern European regime of Greece which was badly hit by the crisis, on top of the much weaker transfer systems and poor population coverage, there is virtually no inequality between classes in reported levels but even amongst the professional class 84 percent report deprivation (relative to 87 percent in the unskilled manual class). As such, these results emphasise the gross cross-national differences in deprivation while also noting that higher levels of inequality and differentiation within countries should be considered with respect to overall reported levels in the country as a whole.

Italy, on the other hand was one of the countries where the proportion of deprivation in the semi/unskilled manual class was quite high and as such one could argue that the poorest individuals here are particularly worse off, both in absolute terms and also in terms of their relative experience to those in more fortunate positions. Other than Italy, countries with the highest levels of deprivation in the semi/unskilled working class (Greece, Spain and France) tended to have relatively lower levels of inequality (with ratios of 1.03, 1.39 and 1.37, respectively). The countries with lower proportions experiencing deprivation tended also to be more unequal – including the UK and Switzerland (ratios of 1.52 and 1.45, respectively). Poland had relatively lower levels of absolute deprivation accompanied by relatively more equality as well (ratio, 1.31). As such, on balance here, evidence for H2 is weak.

Next, in order to test for H3-8 we ran a series of multilevel models with class as the key independent variable and examining the extent to which class and other risk factors account for reported deprivation during the economic crisis with results reported in Table 5. Firstly, testing and confirming H3 we can see that there is a strong class effect on reported deprivation with a clear gradient from the professional to the less skilled manual classes (the ‘other’ category is more mixed). Testing for H4 by looking at the results from subsequent models, we can see that while the effect of class is gradually diminished with the addition of more risk factors and controls in subsequent models, it remains strong throughout (we ignore results in Models 8 and 9 here as these contain cross-level interactions). Testing for H5 in Model 6 specifically, we can see that, against expectations, there is no direct effect of social spending on reported deprivation. As such, other features of welfare regimes are likely to be more relevant at the macro-level for reported individual level deprivation, including inequality levels as tested for in Model 7 and confirming H6 with the significant and positive effect for the Gini coefficient on reported deprivation. However, with respect to H7 and H8 tested for through the cross-level interactions included in Models 8 and 9 we find no evidence to support the argument that being in semi/unskilled manual occupations has a further heightened effect on reported deprivation in contexts of higher inequality or lower social spending.

Finally, the effects of the controls generally reflected those suggested in the theory section based on extant literature. The generational divide prominent in the press with the baby-boomers, or the 60–70s ‘lucky’ generation, appears to stand up to scrutiny in that they are less likely to be deprived than the 1980s generation. However, the youngest two generations are found to be about as well off as the oldest, Post-WWII generation. The models also show that education level and associational participation are not linked to reported deprivation.

Social class is perhaps the most contested and scrutinised concept in sociology. Intimately related to the debates on the meaning of social class are debates on the extent of inequalities linked to class. In fast-changing societies, multiple sources of disadvantage overlap to marginalise deprived groups. In this paper, we examined cross-national and within-country inequalities by social class in reported deprivation during the crisis. We know that inequality has been steadily increasing in advanced societies. Despite being in employment, many individuals in advanced democracies remain financially vulnerable. By analysing data from a new cross-national survey conducted in 2015 in nine European democracies representing five different types of welfare regime and asking individuals a variety of questions on their deprivation, this paper shows that there are important inequalities as reported by individuals in different social classes and cross-nationally. In general, we found that individuals in manual occupations in countries that were not so deeply affected by the crisis were still worse off than middle class individuals in countries that were more deeply affected. Semi/unskilled manual classes were found to be the most deprived.

With this investigation we hope to have made a valuable contribution to the study of cross-national and cross-class differences in deprivation in Europe building on the insights provided in recent scholarship on poverty and deprivation, in particular the work by Nolan and Whelan (2011). To this literature we hope to have added some insights on the dimension of analysing countries during the economic crisis by using a rich and original comparative individual-level survey dataset comprising nine European countries covering five different types of welfare regimes collected in 2015 which also allowed us to control for various individual-level risk factors. Moreover, in our multilevel models we also tested for whether individual level characteristics interacted with aggregate level factors for exacerbating class differentials in deprivation in more unequal or welfare-poor contexts.

We showed that, while countries normally fulfilled the expected welfare regime patterns as hypothesised, those where the crisis was deeper exhibited reported higher reported deprivation levels than would be expected from their welfare regime alone. Moreover, we found the highest levels of cross-class inequality in those countries where overall reported deprivation levels were lower, so that the middle class situation in worse off countries was comparable to that of the manual classes in the richer nations. We also found evidence that class effects on deprivation diminished but persisted with the inclusion of various controls across models as well as that more unequal macro-level contexts are linked to higher reported deprivation. In this way, we hope to have shown the value of investigating the relationship between class and deprivation in the context of the economic crisis. In a context of growing inequality across the globe our study examined how the crisis was experienced by European citizens and how stratification impacted on these experiences.

Overall, our results show the importance of examining both within and between country differences in reported deprivation in Europe. Future studies should seek to develop these analyses and further disentangle the underlying mechanisms for class inequalities and deprivation and provide further nuanced evidence-based advice to national and supranational bodies such as the EU (see for e.g. Nolan and Whelan 2011 for an excellent example of this) for developing the most suited targets for effective initiatives of poverty alleviation within and across European countries.

This work was supported by the European Commission 7^th Framework Programme: project name Living with Hard Times: How European Citizens Deal with Economic Crises and Their Social and Political Consequences (LIVEWHAT) [grant agreement number 613237] coordinated by the University of Geneva (Marco Giugni). The authors are extremely grateful to all the participants at the LIVEWHAT paper workshop in Warsaw on 19-20 May, 2016 for their feedback and in particular to Marco Giugni and Maria Mexi. We would also like to thank the Editor and the reviewers at European Societies for their useful comments.

No potential conflict of interest was reported by the authors.

Maria Grasso is Professor at the Department of Politics at the University of Sheffield. She is the author of Generations, Political Participation and Social Change in Western Europe (2016) and co-editor of Austerity and Protest: Popular Contention in Times of Economic Crisis (with M. Giugni, 2015). Her research focuses on political sociology and political engagement. She is European Editor of Mobilization: An International Quarterly.

Sotirios Karampampas holds a PhD in Politics from the University of Sheffield. He received his BA in Political Studies and Public Administration from the University of Athens (2008), and his MA in International Politics and Security Studies with Distinction from the University of Bradford (2009). His research interests are in contentious politics, social movements, political violence and radicalisation. His work has recently been published in Acta Politica and Politics and Policy.

Luke Temple is a Teaching Associate in Political Geography at the University of Sheffield. His research looks at political and electoral geography, democratic theory, and patterns of digital engagement.

Barbara Yoxon is an Associate Lecturer in Quantitative Methods and Political Science at the University of York. She received her doctorate in political science at the Department of Politics, at the University of Sheffield and has been awarded the BISA Michael Nicholson Prize for the best doctoral thesis in International Studies in 2018. Her main interests include democratisation, political prejudice, nationalism, international conflict, and regime change.

Maria Grassohttp://orcid.org/0000-0002-6911-2241

Sotirios Karampampashttp://orcid.org/0000-0002-0908-8315

Luke Templehttp://orcid.org/0000-0002-2605-2285

Barbara Yoxonhttp://orcid.org/0000-0002-0810-1282