“Generational differences between attitudes towards welfare
provision. Convergence between Eastern and Western Europe?”
Michael Ochsner, FORS - Swiss Centre of Expertise in the Social Sciences, Lausanne, and
Professorship for Social Psychology and Research on Higher Education, ETH Zurich.
Corresponding author, e-mail: [email protected]
Erika Antal, FORS - Swiss Centre of Expertise in the Social Sciences, Lausanne
Ivett Szalma, FORS - Swiss Centre of Expertise in the Social Sciences, Lausanne
Presented in session: “Preferences for redistribution and policy outcomes”
at the 3rd International ESS Conference, 13-15th July 2016, Lausanne, Switzerland
Abstract
The aim of this research is to analyse whether young people – socialized mainly in a globalized
market economy – have different views concerning the role of the state in the welfare system
than the older generations in Eastern and Western Europe. Specifically, we investigate whether
there is a convergence between Eastern and Western Europeans’ attitudes towards the role of
the state in the sense of a less pronounced difference between young Eastern and Western
Europeans than between older Eastern and Western Europeans. This paper presents first results
of a Latent Class Analysis using data from the fourth round of the European Social Survey. The
results suggest that there are no pronounced differences between the younger and older
generations regarding attitudes towards the welfare state. While social support is seen mainly
as government’s responsibility across Europe, perceived social and economic effects of the
welfare state vary among countries.
Keywords
Eastern Europe, Attitudes towards Welfare Provision, Generational Differences
Introduction
Population ageing and shifts in family patterns increase needs for welfare provision, while the
pressures of globalization on national competitiveness restrict the capacity to finance expansive
welfare policies. Europe has been facing this problem of the contradiction between economic
efficiency and social support for some decades now, at least since the transformation of the
socialist countries to a market economy that increased globalization and economic pressures on
the welfare state (see e.g. Bengtsson & Scott, 2011; Kumlin & Stadelmann-Steffen, 2014;
Reeskens & van Oorschot, 2015; Taylor-Gooby, 2011).
More than two decades after this transformation, a new generation emerged that has been
brought up in a globalized world led by economic principles whereas the older generation has
been socialized in (more or less) comprehensive welfare states. This is especially evident in the
post-socialist countries where the older generation was socialised in a regime where the state
ensured a workplace for all its citizens and a younger generation has made completely different
experiences on the labour market and with the state than their parents’ generation (see e.g.
Kluegel, Mason & Wegener, 1999; Mishler & Rose, 2007).
The interaction of a generational shift and the economic pressures of globalization (and
modernization) are often seen as endangering solidarity between generations and leading to a
1
higher spending on old age provision than on child-care and family issues (see e.g., Lee, 1986;
Popenoe, 1988; Preston, 1984). Empirical studies, however, show scarce evidence for a “war
between generations” (Lee, 1986, p. xi), especially for Western Europe (see e.g. EspingAndersen & Sarasa, 2002; Hamil-Luker, 2001; Pampel, 1994; Reeskens & van Oorschot,
2015).
However, welfare policy consists not only of the amount of money that is distributed to different
groups of the society. Welfare policy can take on different forms (see e.g. Bonoli, 2005; EspingAndersen, 1990; Morel, Palier & Palme, 2012) and attitudes to welfare services are
multidimensional (Roosma, Gelissen & van Oorschot, 2013). If we want to understand the role
that people would like the government to take in welfare state matters, we should not only focus
at the opinions on the ideal welfare state (how much money is granted to which groups) but
also look at the perceived effects of the welfare services.
The aim of this research is to analyse whether young people 1 – socialized mainly in a globalized
market economy – have different views concerning the role of the state in the welfare system
than the older generations in Eastern and Western Europe. Specifically, we investigate whether
there is a convergence between Eastern and Western Europeans’ attitudes towards the role of
the state in the sense of a less pronounced difference between young Eastern and Western
Europeans than between older Eastern and Western Europeans. Is there a cohort effect on
attitudes (Inglehart, 1997) or do people adapt to institutional settings (Demartini, 1985)?
Data and methods
We use the fourth wave of the European Social Survey (ESS) for the analysis, which contains
data for 27 European countries (European Social Survey Round 4 Data, 2008). It was
administered between autumn 2008 and spring 2009 and contains a rotating module that focuses
on attitudes towards the welfare state. We use Latent Class Analysis (see e.g. Aragasti, 2002;
Hagenaars & McCutcheon, 2002) to identify clusters of attitudes towards the role of the state
in welfare provision, differentiating between the ideal welfare state (what role the state should
take in resolving some social issues) and the effects of the welfare state (perceived economic
and social consequences of the welfare state in the respondents’ countries). We run separate
models for the younger generation (18 or less years old in 1990, the time of regime change in
Eastern Europe, or 36 or younger at the time of the survey) and the older generation (19 years
or older in 1990).
The attitudes towards the ideal welfare state, or what role the state should take in resolving
some social issues, are measured using the following five variables: How much responsibility
do you think governments should have to…: a) ensure a job for everyone who wants one, b)
ensure a reasonable standard of living for the unemployed, c) ensure adequate health care for
the sick, d) ensure a reasonable standard of living for the old, e) ensure sufficient child care
services for a working parents, and f) provide paid leave from work for people who temporarily
have to care for sick family members 2.
The attitudes towards the effects of the welfare state, i.e. the perceived economic, moral and
social consequences of it, are measured using eight variables: To what extent do you agree or
disagree that social benefits and services in [your country]…: a) place too great a strain on the
economy, b) cost businesses too much in taxes and charges, c) make people lazy, d) make
people less willing to care for one another, and e) make people less willing to look after
In this research, we cannot differentiate between age and cohort. We therefore refer to the commonly
used term “generation” that can have both connotations.
2 We changed the order of the variables to group them thematically. The variables have the following
names/order in the questionnaire: d15, d18, d16, d17, d19, d20.
1
2
themselves and their family, f) prevent widespread poverty, g) lead to a more equal society, h)
make it easier for people to combine work and family life 3.
In order to arrive at comparable and easy to interpret measures for the two groups of attitudes 4,
we recoded the 11-scale items for the ideal welfare state and the 5-scale items of the economic
and social consequences into a categorical variable with 3 values, pro-government, neutral, promarket (i.e., not governments responsibility) on the one hand and negative, neutral and positive
on the other hand.
To calculate the latent classes we used gender, whether the respondent is on the labour market,
and whether the respondent has a tertiary education as active covariates (influencing the class
membership). We used the poLCA package in R (Linzer & Lewis, 2014) to run our analyses.
Results
Regarding the ideal welfare state, we extracted four classes. The first class is characterised by
pro-government attitudes. The members of this class think it is the government’s responsibility
to provide services for the disadvantaged. The second class thinks that providing health care
and a reasonable standard of living for the old is the government’s responsibility while the other
items score in the middle or slightly pro-government. The third class scores in the middle for
all items. The fourth class is more economy-oriented but score in the middle for the health care
and old-age provisions.
If we look at how many people belong to which class, we can see that there is not much
difference between the two age groups. Class 1 (government-oriented) makes up 40% of the
sample for the older group whereas it sums up to 44% for the younger group. Class 2 (neutralgovernment) sums up to 40% and 38%, Class 3 (neutral) to 18% and 15% and Class 4 (marketoriented) to 2% and 3%, respectively.
The class-memberships, i.e. the preferences for the role of government in welfare provision,
can vary across countries. This is indeed the case. While in Greece, Lithuania and Ukraine more
than 60% of the respondents belong to Class 1 (government-oriented), less than 20% belong to
this class in Switzerland and The Netherlands for both age groups. Figure 1 schematically
shows the preferences for the organisation of welfare provision in countries for both age groups
(the size of the bubble represents the size of the country’s population). There are three groups
of countries: In the countries of the first cluster, both the younger and the older generation prefer
a welfare model in the responsibility of the government. It comprises Eastern Southern
European countries. In the second group, both old and young are neutral but see the old age
provision and health care in the responsibility of the government. The Western European
countries and Slovakia and Czech Republic are in this group. The third group consists of the
Northern European countries plus Poland, Slovenia, and Estonia, where the older generation is
predominantly in the government-neutral class and the younger generation sees welfare
provision as a responsibility of the government.
3 Also here, we changed the order of the variables to group them thematically. The variables have the
following names/order in the questionnaire: d21, d25, d27, d28, d29, d22, d23, d26. Also, we harmonised
the scale of the items so that higher ratings indicate positive effects of the welfare state for all the items.
4 Roosma, Gelissen & van Oorschot (2013) examine the dimensionality of welfare attitudes in detail. They
find measurement invariance across a range of 22 countries using the same ESS data as in this paper (they
name the dimensions differently: “Range” for ideal, “Outcomes goals” and “Outcomes unintended” for
effects). Roosma, van Oorschot & Gelissen (2014) use the first dimension in a Latent Class Analysis and call
it “should”-dimension (as opposed to the “is”-dimension, not used in this paper). Van Oorschot, Reeskens
& Meulemann (2012) test the measurement invariance of the three components of the effects dimension
successfully in 25 countries using the ESS data and call them social, moral and economic consequences of
the welfare state.
3
Figure 1. Groups of countries regarding the ideal welfare state by generation
Regarding the perceived consequences of the welfare state, the Latent Class Analysis suggests
six classes. Unsurprisingly, the classes differentiate attitudes towards effects on the economy
and moral behaviour (items a to e), representing a market-oriented ideology of welfare
distribution, from attitudes towards effects on social policy goals (items f to h), representing a
government-oriented ideology of welfare (see also van Oorschot et al., 2012; Roosma et al.,
2013). Class 1 is characterised by market-oriented attitudes (they perceive negative effects for
all items). Class 2 members see negative effects concerning the economy and moral behaviour
but positive effects concerning social policy goals (e.g. welfare services are a strain to economy
and make people lazy but reduce poverty). Class 3 includes the respondents who see no positive
consequences regarding social policy goals but do not see negative consequences for the
economy and moral behaviour. Class 4 is neutral regarding all items, Class 5 is neutral
regarding effects on the economy and moral behaviour but sees positive effects regarding social
policy goals and Class 6 is characterised by being in favour of welfare provision thus seeing no
negative consequences on the economy and moral behaviour but positive effects on social
policy goals. Also regarding the perceived consequences of the welfare state, the differences
between the age groups across all countries are small. Class 1 sums up 12% of the older
generation and 13% of the younger generation. Class 2 combines 17% and 17%, Class 3 16%
and 15%, Class 4 14% and 12%, Class 5 13% and 12% and Class 6 28% and 26%, respectively.
4
Figure 2. Groups of countries regarding the perceived consequences of the welfare state by generation:
Economy and moral behaviour
Figure 2. Groups of countries regarding the perceived consequences of the welfare state by generation:
Social policy goals
5
There is again considerable variance across countries. Figure 2 and 3 provide the schematic
illustration of countries regarding attitudes towards the effects of the welfare state by
generation. Combining the classes for the old and young generation, we find eight groups of
countries. The first group is characterised by predominantly seeing no negative effects
regarding social policy issues (label “Social” in the legend) but negative economic and moral
effects (label “Eco” in the legend), the second group mainly sees negative effects concerning
social policy but no negative economic and moral effects, the third mostly sees no negative
effects, the fourth predominantly sees negative effects on the economy and moral behaviour as
well as social policy issues, the fifth is undecided for both economic and moral as well as social
policy issues. In groups one to five, there is no difference between the older and the younger
generation. Group 6 predominantly sees no negative effects on social policy issues and the older
generation mainly sees no negative economic and moral effects, while the younger generation
sees some negative economic and moral effects. In group 7, the older generation predominantly
sees negative effects on social policy issues while the younger generation sees no such effects,
both generations mostly see negative economic and moral effects. Finally, in group 8, the older
generation is undecided whether there are negative effects on social policy or economic and
moral issues while the younger generation mainly sees no negative effects for neither economic,
moral nor social policy issues. For ease of interpretation, we separate the figures for the effects
on the economy and individual behaviour and the social policy goals, interlinking the two with
colouring according to the country groups, which are constructed along the two issues. Figure
2 shows the results regarding the economy and moral behaviour and figure 3 shows the results
regarding the social policy goals.
Conclusion/Discussion
The results suggest that the European population generally expects a socially supportive welfare
state (see also van Oorschot et al., 2012; Roosma et al., 2013; Roosma et al., 2014). It is seen
as the government’s responsibility to provide some basic support for the needy. Especially
strong are the expectations for the provision of a reasonable standard of living for the old and
an adequate health care for the sick. Eastern and Northern European countries are slightly
more welfare oriented (i.e. expect the government to provide welfare services), but the Czech
Republic and Slovakia are similar to the Western European countries. Interestingly, there are
no considerable differences between the older and the younger generation regarding the ideal
welfare state. Both generations prefer a governmental solution of welfare provision.
More differences between countries are revealed when it comes to the societal as well as
economic and moral consequences of welfare provision. While in Germany, France, Great
Britain, Poland and Hungary the respondents see some negative economic and moral
consequences, most countries do not see such issues. In Belgium, Switzerland, Portugal, Ireland
and Slovenia, however, the older generation does not see negative economic and moral
consequences, but the younger generation does. Looking at the possible consequences of
welfare provision regarding social policy goals, Russia, Bulgaria, Latvia, Ukraine, Hungary
and Bulgaria do not see positive effects, while in most countries, people see positive effects
regarding social policy goals. In Poland, however, the older generation observes no positive
effects on social policy while the younger generation does.
We can conclude that differences in the attitudes towards the welfare state between the older
and younger generation are rare. Only Poland, Belgium, Switzerland, Portugal, Ireland and
Slovenia show big differences between the generations, yet only regarding one of the three
dimensions examined here. Thus, we cannot support Inglehart’s (1997) thesis of a cohort effect
in this subject. It is more likely that the attitudes towards the welfare state are guided by
institutional experience. This is especially true for the evaluation of perceived consequences of
6
the welfare state. The question of whether there are such consequences is linked to the actual
policy in the countries. Therefore, we find more country differences with regard to these issues.
Here, Demartini’s (1985) thesis of the influence of institutional settings might be an important
driver.
This paper presents preliminary results of our study. Therefore, there are some limitations. First,
we will delve deeper into the influence of the covariates we used for the classification procedure
on the classification results (gender, education, labour market status). Second, we have used
complete case analysis. Since there are more than 5% missing values in the data, we will take
the information we have from all individuals into account using Full Information Maximum
Likelihood. Third, we will take the multilevel structure of the data into account and check
whether a Multilevel Latent Class Analysis will yield different results.
References
Agresti, A. (2002). Categorical data analysis, second edition. Hoboken, NJ: John Wiley &
Sons.
Bengtsson, T., & Scott, K. (2011). Population aging and the future of the welfare state: the
example of Sweden. Population and Development Review, 37(1): 158-170.
Bonoli, G. (2005). The politics of new social policies: providing coverage against new social
risks in mature welfare states. Policy and Politics, 33(3): 431–449
European Social Survey Round 4 Data (2008). Data file edition 4.3. NSD - Norwegian Centre
for Research Data, Norway – Data Archive and distributor of ESS data for ESS ERIC.
Demartini, J. R. (1985). Change agents and generational relationships: a re-evaluation of
Mannheim’s problems of generations. Social Forces, 64(1): 1-16.
Esping-Andersen, G. (1990). Three worlds of welfare capitalism. Cambridge, UK: Polity
Press.
Hagenaars, J., & McCutcheon, A. (eds.) (2002). Applied Latent Class Analysis. Cambridge:
Cambridge University Press.
Hamil-Luker, J. (2001). The prospects of age war: inequality between (and within) age
groups. Social Science Research, 30(3): 346–400. doi:10.1006/ssre.2001.0703
Inglehart, R. (1997). Modernization and postmodernization - cultural and political change in
43 societies. Princeton, NJ: Princeton University Press.
Kluegel, J. R., Mason, D. S., & Wegener, B. (1999). The legitimation of capitalism in the
postcommunist transition: public opinion about market justice, 1991-1996. European
Sociological Review, 15(3): 251–283.
Kumlin, S., & Stadelmann-Steffen, I. (2014). Citizens, policy feedback, and European welfare
states. In S. Kumlin & I. Stadelmann-Steffen (eds.), How welfare states shape the democratic
public. Policy feedback, participation, voting, and attitudes (pp. 3–16). Cheltenham, UK:
Edward Edgar.
7
Lee, D. E. (1996). Generations and the challenge of justice. New York, NY: University Press
of America.
Linzer, D., & Lewis, J. (2014). Package poLCA. Version 1.4.1. Retrieved from:
https://cran.r-project.org/web/packages/poLCA/poLCA.pdf
Mishler, W., & Rose, R. (2007). Generation, age, and time: the dynamics of political learning
during Russia’s transformation. Studies in Public Policy, 412: 1–44.
Morel, N., Palier, B., & Palme, J. (eds.). (2012). Towards a social investment welfare state?
Bristol, UK: Policy Press.
Pampel, F. C. (1994). Population aging, class context, and age inequality in public spending.
American Journal of Sociology, 100(1): 153–195.
Preston, S. (1984). Children and the aged in the US. Scientific American, 251(6): 44–49.
Popenoe, D. (1988). Family change and decline in modern societies. New Brunswick, NJ:
Transaction Publishers.
Reeskens, T., & van Oorschot, W. (2015). Is intergenerational solidarity under pressure?
Comparative analyes of age cleavages in opinions about government support for the young
and the old. In: I. Salagean, C. Lomos & A. Hartung (eds). The young and the elderly at risk:
Individual outcomes and contemporary challenges in European societies (pp. 229–247).
Antwerp, BE: Intersentia Publishers.
Roosma, F., Gelissen, J., & van Oorschot, W. (2013). The multidimensionality of welfare
state attitudes: a European cross-national study. Social Indicators Research 113: 235–255.
doi:10.1007/s11205-012-0099-4
Roosma, F., van Oorschot, W., & Gelissen, J. (2014). The preferred role and perceived
performance of the welfare state: European welfare attitudes from a multidimensional
perspective. Social Science Research, 44: 200–210. doi:10.1016/j.ssresearch.2013.12.005
Taylor-Gooby, P. (2011). Opportunity and solidarity. Journal of Social Policy, 40(3): 453–
470. doi:10. 1017/S0047279410000681
Van Oorschot, W., Reeskens, T., & Meulemann, B. (2012). Popular perceptions of welfare
state consequences: a multilevel cross-national analysis of 25 European countries. Journal of
European Social Policy, 22(2): 181–197. doi:10.1177/0958928711433653
8