The public opinion literature has found that the age of the benefit recipient is an important determinant in the formation of welfare state attitudes. Older people are perceived as more deserving of help and also punished less for not accepting a job. We argue that such a preferential treatment of older people depends on the social and economic context. In this article, we examine public support for demanding active labour market policies in 21 European countries. Relying on a survey experiment varying the age of the unemployed person, our analysis confirms that older unemployed are punished less than younger unemployed for not accepting a job offer. However, this effect varies between countries and our evidence suggests that support for exempting older individuals from demanding active labour market policies disappears as societies age. Moreover, support for stricter sanctions in general is higher in countries with a higher unemployment rate and in countries that already have rather strict active labour market policies. These findings question the public’s role as a veto player in the reform process as it seems unlikely that public opinion will block attempts to further strengthen demanding active labour market policies.

Over the last three decades, Western welfare states have introduced active labour market policies (ALMPs) to accelerate the labour market integration of the unemployed (Halvorsen and Jensen 2004; Weishaupt 2010). Many ALMPs combine enabling-measures like job-search assistance and training with demanding-measures, which put pressure on the unemployed to find a new job. Examples of such demanding-measures include sanctioning for a lack of job search or broadening the definition of acceptable jobs (Eichhorst and Konle-Seidl 2008). However, the strictness of demanding ALMPs differs between target groups and some groups like older people are usually exempted from harsh sanctions and strict conditionality (Buss 2019; Larsen 2008; Petersen et al.2011). At the same time, the growing demand for citizens to be productive members of a society increasingly applies to older individuals as well, as active ageing policies have become a central part of the policy agenda in response to the challenges of ageing populations. One key aim of active ageing policies is to increase the labour market participation of older workers (Walker and Maltby 2012). To that end, countries have limited or abolished early retirement pathways (Hofäcker et al.2015), including the abolition of a period before reaching the retirement age, in which unemployed individuals are excused from job search requirements (Bloemen 2016).

The prospects of policy reforms – in terms of enactment and implementation – crucially depend on their social legitimacy (van Oorschot et al.2017). It is therefore important to understand the determinants of public support for ALMPs and in particular, whether support for sanctions is conditional on characteristics of the unemployed target person (Petersen et al.2011). In this, we follow recent contributions that argue that welfare attitudes are multi-dimensional (Fossati 2018; Gallego and Marx 2017) and that we need to distinguish two facets of the welfare state: first, who pays taxes and social security contributions and, second, who receives benefits (Cavaillé and Trump 2015). In this paper, we provide empirical evidence on this question relying on comparative survey data from 21 European countries. We explore individual attitudes towards sanctioning unemployed individuals who refuse a job offer. We use a survey experiment varying the age of the unemployed person – no age given, 20–25 years old or 50 years old – in order to explore in how far such sanctions depend on the unemployed person’s age.

In line with previous research, we find public support for imposing sanctions on unemployed who do not adhere to job search requirements (Buss 2019; Fossati 2018; Larsen 2008). We show that support for sanctions varies with the institutional context and we find more support for demanding ALMPs in countries with stricter labour market policies. Our main research focus, however, engages with the question whether support for sanctions varies with the age of the target person. We show that unemployed persons in their 50s are punished less than younger unemployed. Our results in general confirm expectations derived from the deservingness argument, which suggests that older workers are perceived as more deserving of help and should be less punished for not accepting a job offer (Buss 2019) since they have usually already contributed to the social system (reciprocity), and have more difficulties in finding a new job (control) (Larsen 2008). Nevertheless, this paper also shows that the country context affects the willingness to punish old and young unemployed differently. As a country’s population becomes older, the preferential treatment of older unemployed is increasingly likely to disappear and the public applies the same sanctions to all unemployed for not accepting a job offer irrespective of their age.

Our contribution to the literature is twofold. Theoretically, we aim at combining two strands of the literature, the deservingness and the ageism approach, in order to explain how the age of the target person determines the degree of sanctions if someone does not accept a job offer. Moreover, we argue that population ageing has the potential to change which of the two motives to support or oppose sanctions of older unemployed dominates the formation of individual attitudes. Empirically, our article adds to recent contributions in the field of welfare attitudes research specifically examining public attitudes towards demanding ALMPs, in particular the sanctioning of particular groups of unemployed individuals. Buss’ (2019) single-country study convincingly shows that age affects support for sanctions but it remains open whether these findings are also valid in other institutional contexts (see also Carriero and Filandri 2018; Larsen 2008; Petersen et al.2011). Hence, we complement single-country studies with a cross-country survey experiment with novel data from the European Social Survey (ESS) specifically focusing on the effect of the age of the unemployed on support for sanctions across different institutional settings. Therefore, the primary focus of the paper will be on the macro rather than the individual level.

The main aim of the paper is to explore whether support for sanctions differs depending on the age of the targeted person. We are in particular interested in developing explanations for why respondents in some countries would be willing to sanction older unemployed individuals less harshly than younger ones. Subsequently, we briefly present other known drivers of support for sanctions in general, i.e. irrespective of the age of the unemployed, and derive our expectations on how country-level and individual-level characteristics are related to support for sanctioning unemployed individuals.

Sanctioning and age of the unemployed

The deservingness literature revolves around the question to what extent a certain group is considered ‘deserving’ of receiving social benefits. Perceptions of deservingness are based on five aspects, known as the deservingness criteria (van Oorschot 2006): control (whether the welfare recipient is in control of his or her situation of dependency), attitude (whether the recipient is compliant and thankful), reciprocity (whether the recipient has contributed in the past), identity (whether one can identify with the welfare recipient) and need (whether the recipient really needs benefits). One particular finding in this literature is that public support for older individuals is highest, followed by support for the sick and disabled. The unemployed are seen as a markedly less deserving group, as they contribute little to society (van Oorschot 2006). In relation to deservingness of unemployed at different ages, Larsen (2008) shows that Australians are indeed much more inclined to set requirements for younger unemployed individuals (Larsen 2008). They were especially required to make a larger effort in terms of job search behaviour and in terms of their willingness to undergo personal changes such as moving elsewhere to find work (see also Buss 2019). This brings us to our first expectation, which is located at the individual level:

H1: Generally, older unemployed will be punished less when refusing a job offer than younger unemployed individuals.

In order to understand why respondents are more lenient towards older unemployed individuals, we need to link the deservingness concept to ageism (e.g. Koster 2008). Ageism refers to discrimination, perceptions and existing stereotypes of people based on their age (Butler 1969). Ayalon and Tesch-Römer (2018: 6) define ‘[…] ageism as the complex, often negative construction of old age, which takes place at the individual and the societal levels’ and therefore reflects how older individuals are viewed by society. Dordoni and Argentero (2015) add that these stereotypical views regarding older individuals can contain both positive and negative aspects. When looking at older individuals in the labour market, research finds that older workers are typically seen as less productive, less adaptable to change and due to seniority wages as comparably expensive. Positive stereotypes of older workers often relate to their accumulated experience and perceived reliability. These positive as well as negative stereotypes are shared by society as a whole, including employers, trade unions and even older workers themselves (Naegele et al., 2018). These stereotypes could also affect older jobseekers: If an older worker is perceived as less productive, for example, the process of labour market integration might be seen as an unprofitable course of action.

Deservingness perceptions in turn are now largely affected by the public image of a target group, and thus by stereotypes (van Oorschot et al.2017). Negative stereotypes about older workers’ health condition and employability may positively affect support for welfare benefits to older individuals since they would signify a higher need and lower control over the individual’s labour market participation (Dordoni and Argentero 2015). Support can also stem from a more positive attitude if older workers are considered as more honest and committed (Posthuma and Campion 2009). Moreover, older individuals have contributed before through earlier employment, corresponding to the reciprocity criterion (Larsen 2008). From this perspective, the existence of stronger age stereotypes in society should coincide with an inclination to sanction older unemployed individuals less harshly than younger ones. Thus, we expect:

H2: Older unemployed will be sanctioned less than younger unemployed individuals in countries with stronger age stereotypes.

The relationship between ageism and population ageing has been approached from a modernization perspective (Ayalon 2013; Löckenhoff et al.2009). From this perspective, older individuals lose status in society as society goes through the process of modernization – increased life expectancy and a larger share of older individuals in society being part of this process. This loss of status is the result of a number of reasons, one of them being that older individuals are simply not exceptional anymore. Economic developments undermined their competence, Taylorism initially turning knowledge and experience obsolete and the knowledge economy now making them outdated (Cowgill and Holmes 1972). Socially, individualization hollowed out the privileges of older individuals (Ingelhart and Welzel 2005). Hence, in societies with a larger share of the population in old age, older individuals are a less deserving group. Indeed, Löckenhoff et al. (2009) find that in societies in which a larger share of the population is above age 65, young adults view older people more negatively. Therefore, we expect the following:

H3: As the share of older people in society is larger, older unemployed individuals will be less exempted from sanctions compared to younger ones.

Factors impacting non-age-related support for sanctioning

In the remainder of the theoretical framework, we discuss known determinants of welfare state attitudes in relation to sanctions for unemployed individuals. Since the primary interest of the paper is at the macro-level, we first examine further contextual drivers of sanctioning before discussing individual-level factors.

Previous research has shown that the institutional and the socio-economic context affects aggregate welfare attitudes (Ebbinghaus and Naumann 2018). As to institutional differences, the comparative literature usually assumes some degree of congruence between the existing institutions and public attitudes. Policy responsiveness, i.e. the idea that the democratic public should get what it wants, is a core idea of democracy (Kumlin and Stadelmann-Steffen 2014). Moreover, the policy feedback perspective argues that welfare state institutions shape individual self-interest by providing benefits and services (Pierson 1993). Finally, institutions also convey an idea of a legitimate social order providing orientation for individual values and normative beliefs (Mau 2003). Indeed, more universal welfare schemes, such as pension schemes, are more widely supported (Laenen 2018). Applying this argument that welfare state arrangements reflect the citizens’ values to attitudes towards sanctions for unemployed, one would expect to find more support for this in countries with stronger sanctions.

H4: Support for sanctions will be higher in countries with stricter regulations and more demanding ALMPs.

The financial crisis in 2007/2008 has revived the scientific interest in the consequences of economic downturn on support for the welfare state (Naumann et al.2016; Sachweh 2018). Most studies argue that people are willing to accept greater individual responsibility and want less social protection in good economic times whereas they ask for more governmental protection in bad economic times. Sachweh (2018) shows that stronger welfare state support in hard times is not driven by individual self-interest only, but that also crisis experiences (i.e. job loss) of colleagues and acquaintances lead to more welfare state support. Buss et al. (2017) show that people living in wealthier countries are more likely to be in favour of conditionality, whereas a high unemployment rate reduces such support. However, it is not clear whether these findings apply also to support for sanctions of different age groups of unemployed. On the one hand, a high unemployment rate indicates that the unemployed are less in control of their situation and should thus be sanctioned less. On the other hand, refusing a job offer in times when jobs are scarce might be perceived as a more severe misconduct of the unemployed and hence be punished stronger. The few empirical studies on this question rather lend support to the first argument (Buss et al.2017) and we expect:

H5: Support for sanctions will be lower in countries with higher unemployment rates.

At the individual level, the welfare state attitudes literature distinguishes between two main motives why individuals support the welfare state, self-interest and values (Svallfors 2010). First, the self-interest-interest perspective posits that those who currently benefit or expect to benefit from the welfare state should show stronger support. A very similar logic applies to support for sanctions. The welfare state in general and more specifically unemployment benefits can be seen as a common-pool resource. Punishing free-riders, i.e. those people who benefit from the common good without showing the motivation to contribute to it, is in the interest of everyone as the provision of the public good crucially depends on such a sanctioning mechanism (Ostrom et al.1994). Some contributions even argue that humans have a moral punishment instinct and that people have a universal tendency to punish free-riders even if punishment is costly (van Prooijen 2018). In general, we therefore expect rather high support for sanctioning unemployed who do not adhere to job search requirements. Nevertheless, we expect differences in the degree of support for sanctions and expect lower support among those people who are potential target persons of such sanctions (Fraile and Ferrer 2005). This applies to the current unemployed but also to those people who face a higher risk to become unemployed in the future.

H6: Support for sanctions will be lower among the unemployed, or those who have experienced unemployment before.

Sociological approaches to welfare state attitudes stress the importance of socialization for attitude formation. Accordingly, values, religion and also political ideology shape individual attitudes (Kulin and Meuleman 2015). Political ideology has gained attention in this respect and seems to be the most relevant factor for welfare state attitudes. Left-leaning respondents should be less inclined to support sanctions because they see a higher responsibility of the state to care for those in need and also because they prioritize individual development over economic performance (Fossati 2018).

H7: Support for sanctions will be higher among individuals situating themselves further to the right on the political left-right scale.

We use data from the 8th round of the European Social Survey (ESS). The survey was conducted in 2016 and 2017 and we use data from 21 European countries. In 2016 between 1000–2500 respondents 16 years and older have been surveyed per country. We follow the standard approach in the literature and apply list-wise deletion on cases with missing values on relevant variables. This reduces our sample by 10.5% and we have 26,470 observations in our dataset. One should be aware that list-wise deletion is not the optimal strategy for dealing with missing values and might lead to biased results if values are not missing completely at random (Acock 2005). Yet, we think that 10.5% missing observations are a comparably small share of missing observations, which should limit the possible bias this creates. In Appendix C we provide a more detailed report of missing values which allows evaluating the implications of missingness on the results.

For the dependent variable, we rely on two questions asking how an unemployed person should be sanctioned for turning down a job for different reasons. The reasons for not accepting a job include less pay and a lower level of education required for this job (for exact question wording see Table A1). Response options include ‘Should lose all unemployment benefit’, ‘Should lose about half of unemployment benefit’, ‘Should lose small part of unemployment benefit’ and ‘Should keep all unemployment benefit’. We recode this variable so that the response options have a substantial meaning and show by how much the respondents would cut unemployment benefits (100%, 50%, 25% or 0%).1 We take the average of the two questions as our dependent variable at the individual level. Hence, higher values of the dependent variable mean support for stricter sanctions for refusing a job offer.2

We use a survey experiment in order to explore whether support for sanctions depends on the unemployed person’s age. One group of respondents received the question without mentioning the age of the unemployed person (Control). The experimental treatment consists of information on the age of the unemployed person. Instead of ‘Imagine someone who is unemployed … ’ the question starts with ‘Imagine someone in their 50s who is unemployed … ’ and a third group received a question starting with ‘Imagine someone aged 20–25 who is unemployed … ’. There are slightly more significant differences in the demographic composition of the experimental groups than we would have expected by chance: 7.4% instead of 5% of the possible differences in each country (see Appendix B). This casts some doubts that the randomization has worked perfectly and we include demographic variables as controls in our analysis to account for the fact the experimental groups are not balanced with regard to every demographic variable in all countries.

Hence, we include a set of individual-level variables into our analytical models also because it is of substantial interest, which groups in society support sanctions more than others. We include three age categories (younger than 30 years old, 30–59 years old, and 60 years and older), gender and the highest level of education (ES_ISCED in 5 categories). We distinguish those in paid work from the unemployed, the retired, and those in education or doing housework. Previous unemployment experience also captures the individual labour market status and has been used as an indicator for labour market risks (Fossati 2018). To capture political ideology, we recode the self-placement on an 11-point left-right scale into three groups (0–3: left, 4–6: centre, 7–10: right) and add ‘don’t knows’ as a fourth group.

Our second hypothesis argued that age stereotypes at the country level should be linked to sanctioning older and younger unemployed. As the ESS does not provide such a measure in the 2016 round, we rely on data from the Eurobarometer 83.4 fielded in 2015 instead. Respondents were asked to evaluate the degree of age discrimination towards people being over 55 years old in their country. We use the country averages of this item as an indicator for age stereotypes (see Koster 2018 for a similar strategy). Data is missing for 3 countries which were not part of the Eurobarometer (Iceland, Norway and Switzerland) and we have to drop these countries for the analysis of age stereotypes. Age stereotypes in a country and perceived age discrimination in that country are not the same, but lacking more appropriate survey data on ageism on the country level at the time of the data collection of the ESS wave 8, we use the perceived age discrimination as a proxy. Previous literature has also shown the close relationship of age stereotypes and perceived age discrimination, which both are sub-dimensions of the larger concept of ageism. High rates of perceived age discrimination coincide with negative stereotypes of older people (Voss et al., 2018).

We use the old age ratio as an indicator for population ageing (Eurostat 2015). The old age ratio is defined as the ratio of people older than 65 per 100 people aged from 15 to 65 years. It is a standard indicator for population aging (OECD, 2011).

In order to capture the existing institutional context of active labour market policies in each country, we rely on expert evaluations. Langenbucher (2015) asked national labour market experts to classify policies in terms of demands for occupational mobility and of sanctions for refused job offers by unemployed and then allocated a score for each dimension from 1 (least strict) to 5 (most strict). We prefer such a quantitative indicator for policies to using typologies (e.g. Bonoli 2013) as they are better suited to capture gradual differences between countries, which seem to be particularly important in a policy field where many changes have been taking place over the last decades.

Overall economic success and overall welfare state spending of a country influences individual welfare state attitudes and might also be related to ALMPs (Wlezien 1995). We use social expenditure as share of GDP, the unemployment rate and the gross domestic product per capita as controls in our models.

To account for the cross-national, nested structure of the data – and an intraclass correlation of 0.10 – we estimate two-level linear regression models with random intercepts. We use random slopes for our main independent variable as we assume that the effect of the experiment varies over countries (Heisig and Schaeffer 2019). One main concern about using multilevel models in comparative research is the low number of higher-level units, in our case only 21 countries, which might bias estimates and underestimate confidence levels. Recommendations about the minimum number of countries needed for unbiased results varies between 15 (Heisig et al.2017) and 25 (Bryan and Jenkins 2016, Stegmueller 2013). To address these concerns, we follow recent suggestions and use restricted maximum likelihood estimation (Heisig et al.2017), error correction for small sample sizes and the graphical inspection of our results (Bryan and Jenkins 2016). Moreover, we opt for very simple models and include only one macro indicator at a time.

The main focus of our paper is whether support for sanctions varies with the age of the recipient of unemployment benefits and whether these age-related differences vary over countries (H1-H3). Yet, we also explore how support for sanctions in general are related to the country and to the individual level (H4–H5 and H6–H7). Our empirical strategy is twofold: first, we present the experimental results and explore graphically (Figure 2) whether the country-specific treatment effects of providing the age of the target person varies over countries and whether this difference is related to institutional country characteristics. Such a two-step approach would have been one way to model country effects in a regression, yet we prefer to use a multi-level model with support for sanctions at the individual level as our dependent variable as these models provide more reliable estimates (Heisig et al.2017). Moreover, multi-level models allow us to examine individual-level effects and country-level effects on support for sanctions in one model. Using cross-level interactions between the treatment at the individual-level and country-level characteristics allow us to explore whether age-related differences vary over countries.

We start our analysis with a comparison of the average sanction for not accepting a job offer because the job would pay less or require a lower level of education. In our control group in which we do not provide the unemployed person’s age, respondents would cut benefits by 40.9% on average. Respondents support stricter sanctions against younger unemployed whose unemployment benefits would be cut by 42.2% whereas unemployed persons in their 50s would face less strict sanctions (39.4%). Although these differences are small, a series of t-test shows that all differences between the experimental groups are significant different from zero (control vs. 20–25 years old: diff: −1.31 [−2.28;−0.36], control vs. in their 50s: diff: 1.48 [0.01;2.44]).3 Hence, our analysis supports the expectation that the public in general supports sanctioning unemployed who do not accept a job offer. Though differences are modest, we also show that the strictness of these sanctions varies with the age of the unemployed person and that younger unemployed face stricter sanctions whereas older unemployed are punished less for not accepting a job offer.

As the main interest of this article is to understand how far this pattern varies with the institutional and socio-economic context of a country, we continue our analysis with a comparative analysis of differences between countries. Figure 1 shows differences in average sanctions over the three experimental groups for each of our 21 countries separately. The bars show by how much respondents would, on average, cut unemployment benefits if an unemployed person refuses a job. We observe some variation in support for punishing the unemployed for not accepting a job between countries (ICC: 10%). Respondents in Italy, Slovenia, Poland, Norway and Spain would support the harshest sanctions against unemployed who reject a job offer cutting their unemployment benefits by around 50%, whereas respondents in Germany, Lithuania, Estonia and Austria would also reduce unemployment benefits but to a much lesser extent, around 30% reduction for not accepting a job offer. This is some preliminary evidence for our expectation that the existing institutional regulations but also the current labour market situation shapes support for sanctions. This finding is supported by Italy, Slovenia and Poland being among the countries with the strictest sanctions for not accepting a job offer, whereas high unemployment rates might be an important explanatory factor for the high support for sanctions in Italy and Spain.
Figure 1.

Support for sanctioning unemployed who do not accept a job offer in 21 European countries.

Figure 1.

Support for sanctioning unemployed who do not accept a job offer in 21 European countries.

Close modal

In line with Hypothesis 1, our results suggest that older individuals are perceived as more deserving and older unemployed individuals are usually sanctioned less for not accepting a job offer – but this general pattern is not consistently found in all countries. There is a general pattern that young unemployed (crosses) are punished more than the control group (with the exception of Switzerland, UK, Iceland and Poland). In contrast, older unemployed in their 50s (diamonds) face lower sanctions in most countries, with notable exceptions in particular in Portugal and Spain.

The main focus of this paper is on finding explanations for this cross-country variation and in understanding why respondents in some countries apply stricter sanctions towards older unemployed than in others. We argued that fewer age stereotypes within a country but also a higher share of older individuals should be correlated with smaller age-related differences in attitudes towards sanctioning. To investigate our expectations, we plot the treatment effect – i.e. the difference in sanctions for people in their 50s compared with the control group – against country-level characteristics. In Figure 2 we show the difference between the average sanction for an older unemployed compared with the average sanction of the control group. A negative value means that older people are punished less for not accepting a job whereas a positive value means that older unemployed are punished more for not accepting a job offer.
Figure 2.

Preferential treatment of older unemployed, age stereotypes and population ageing.

Figure 2.

Preferential treatment of older unemployed, age stereotypes and population ageing.

Close modal

The left panel in Figure 2 shows how age stereotypes and the preferential treatment of older individuals are correlated. We argued that strong age stereotypes, i.e. a high perceived discrimination of people 55 years and older, would be related to stronger treatment effects meaning that older unemployed would be sanctioned much less than younger unemployed in these countries. Our results do not support this expectation and we only find a very weak relationship between age stereotypes and the strength of our treatment effect (r = 0.23). If at all, the preferential treatment of older people is weaker in countries where their perceived discrimination is higher. Further, this very small positive correlation is mainly driven by Portugal, where perceived discrimination of older people is high and unemployed individuals in their 50s are even sanctioned more than the control group for not accepting a job. Hence, little support is found for Hypothesis 2.

The right panel in Figure 2 shows that population ageing is related to smaller treatment effects. The strongest negative effect of the 50s-treatment is found in those countries with the youngest population like Poland, Iceland and Norway. In most countries with a high share of older individuals, the treatment has a very small or no significant effect on sanctions and older unemployed are treated more or less as the younger unemployed. Moreover, there are two countries in which older people are punished much more for not accepting a job offer, Portugal and Spain. In Portugal, the average punishment for someone in the control group is 10 percentage points lower than for an unemployed in his or her 50s. Hence, we find a positive correlation between the old age dependency ratio and the treatment effect (r = 0.54). This suggests that, as populations become older, the preferential treatment of older unemployed compared to the general and the younger unemployed gets weaker – and might even be reversed. Hence, if a larger share of the working population is old, insistence to actively integrate also older unemployed into the labour market via demanding ALMPs increase.

We further test the relationship between the country context and attitudes towards demanding ALMPs with a multilevel linear regression. Thus, we control for compositional effects and provide a more formal test of our hypotheses (Table 1). It first confirms that the experimental manipulation of the age of the unemployed target person has a significant effect on support for sanctions of unemployed. Compared to the control group in which no age is provided to the respondents, younger unemployed are punished more and average cuts in unemployment benefits are 1.5 percentage points higher. In contrast, average cuts of unemployment benefits for an older unemployed individual are 1.6 percentage points lower compared to the control group.

Table 1.
Support for sanctions for the unemployed – country-level and cross-level interaction effects of a multilevel linear regression with random intercepts and random slopes for the experimental groups.
(1)(2)(3)(4)(5)(6)(7)
        
Country level        
Age stereotypes −0.113       
 [−0.33,0.10]       
Old age ratio  0.002     −0.001 
  [−0.01,0.02]     [−0.01,0.01] 
Unemployment rate   0.017**     
   [0.01,0.03]     
Social expenditure    0.004    
    [−0.01,0.01]    
Strictness of unemployment     0.050   
benefit eligibility     [−0.01,0.11]   
GDP per capita      −0.000  
      [−0.00,0.00]  
Experimental groups        
Control (no age) Ref. cat. Ref. cat. Ref. cat. Ref. cat. Ref. cat. Ref. cat. Ref. cat. 
20–25 years old 0.015** 0.014** 0.014** 0.014** 0.014** 0.014** 0.014** 
 [0.01,0.03] [0.00,0.02] [0.00,0.02] [0.00,0.02] [0.00,0.02] [0.00,0.02] [0.00,0.02] 
        
in their 50s −0.013* −0.016*** −0.016*** −0.016*** −0.016*** −0.016*** −0.016*** 
 [−0.02,−0.00] [−0.03,−0.01] [−0.03,−0.01] [−0.03,−0.01] [−0.03,−0.01] [−0.03,−0.01] [−0.03,−0.01] 
Cross-level interactions        
Control (no age) # Old age ratio       Ref. cat. 
20–25 years old # Old age       0.001 
ratio       [−0.00,0.00] 
in their 50s # Old age ratio       0.003* 
       [0.00,0.01] 
Individual-level controls Yes (see Table 2
N 18 21 21 21 21 21 21 
n 23722 26470 26470 26470 26470 26470 26470 
(1)(2)(3)(4)(5)(6)(7)
        
Country level        
Age stereotypes −0.113       
 [−0.33,0.10]       
Old age ratio  0.002     −0.001 
  [−0.01,0.02]     [−0.01,0.01] 
Unemployment rate   0.017**     
   [0.01,0.03]     
Social expenditure    0.004    
    [−0.01,0.01]    
Strictness of unemployment     0.050   
benefit eligibility     [−0.01,0.11]   
GDP per capita      −0.000  
      [−0.00,0.00]  
Experimental groups        
Control (no age) Ref. cat. Ref. cat. Ref. cat. Ref. cat. Ref. cat. Ref. cat. Ref. cat. 
20–25 years old 0.015** 0.014** 0.014** 0.014** 0.014** 0.014** 0.014** 
 [0.01,0.03] [0.00,0.02] [0.00,0.02] [0.00,0.02] [0.00,0.02] [0.00,0.02] [0.00,0.02] 
        
in their 50s −0.013* −0.016*** −0.016*** −0.016*** −0.016*** −0.016*** −0.016*** 
 [−0.02,−0.00] [−0.03,−0.01] [−0.03,−0.01] [−0.03,−0.01] [−0.03,−0.01] [−0.03,−0.01] [−0.03,−0.01] 
Cross-level interactions        
Control (no age) # Old age ratio       Ref. cat. 
20–25 years old # Old age       0.001 
ratio       [−0.00,0.00] 
in their 50s # Old age ratio       0.003* 
       [0.00,0.01] 
Individual-level controls Yes (see Table 2
N 18 21 21 21 21 21 21 
n 23722 26470 26470 26470 26470 26470 26470 

95% confidence intervals in brackets.

*p < 0.05, **p < 0.01, ***p < 0.001.

We use the old age ratio and the unemployment rate as an indicator for the demographic and the economic context. We do not find a direct effect of the old age ratio on support for demanding ALMPs. Second, in contradiction to Hypothesis 5, a higher unemployment rate is positively correlated with stricter sanctions, which suggests that people perceive the need for more demanding ALMPs if the labour market situation is bad. In combination with previous research which has shown that respondents ask for more social protection in times of crisis, this suggests that the public supports more generous benefits in times of crisis but not without applying stricter obligations to the unemployed. Our results also show the importance of the institutional context in shaping attitudes towards demanding ALMPs. We provide some evidence for a congruence between existing policies and attitudes as hypothesized (Hypothesis 4), as the public sets stricter sanctions for not accepting a job in those countries that have set high demands on occupational mobility and strict sanctions for refusing job offers. Yet, this relationship is not significant. Generous welfare states, in contrast, are unrelated to higher or lower support for sanctions.

We argued that the preferential treatment of older unemployed persons is affected by population ageing. An increasing number of older people on the labour market, and hence more intergenerational contact, but also normative changes linked to active ageing policies might reduce the gap between sanctions for not accepting a job between younger and older unemployed. We include a cross-level interaction between experimental treatment and old age ratio in order to test this in a multivariate analysis (Table 1, Model 7). These interaction effects confirm what we find in our bivariate analysis. The preferential treatment of older unemployed as indicated by a negative treatment effect becomes weaker in countries with a larger share of older persons.

The examination of individual-level determinants of attitudes towards demanding ALMPs has not been the focus of our theoretical argument and we therefore only very briefly summarize results here, mainly confirming results from previous studies (Fossati 2018). Table 2 shows the individual-level results of the multilevel model. The unemployed and those with unemployment experience, both indicators for labour market risks, would cut unemployment benefits much less than those in employment or without unemployment experience. Respondents with a higher level of education, in particular those with tertiary education, are less inclined to punish unemployed for not accepting a job offer whereas females and older respondents prefer stricter sanctions. We included political ideology as an indicator capturing (political) values. As expected, we find that people who identify with a right political ideology prefer stricter sanctions than left-leaning people do. In summary, the individual-level analysis shows that the common finding of existing attitudes research, i.e. that both self-interest and values are important determinants of attitudes, also apply to support for demanding ALMPs.

Table 2.
Individual determinants of support for sanctions of the unemployed – multilevel linear regression.
(1)
Experimental groups (age of the unemployed person)  
Control (no age)   
20–25 years old 0.02*** [0.01,0.02] 
in their 50s −0.01* [−0.03,−0.00] 
Individual level     
Female 0.02** [0.01,0.03] 
Age (Ref.: <30years)  
30–59 years old 0.02** [0.01,0.03] 
60+ 0.04*** [0.02,0.05] 
Education (Ref.: ES – ISCED I)   
ISCED II 0.01 [−0.01,0.03] 
ISCED III −0.00 [−0.01,0.02] 
ISCED IV −0.01 [−0.03,0.01] 
ISCED V −0.04* [−0.06,−0.01] 
Employment status (Ref.: Paid work)   
Unemployed −0.06*** [−0.09,−0.03] 
Retired −0.00 [−0.02,0.01] 
In education, housework, other −0.03*** [−0.05,−0.02] 
Unemployment experience −0.03*** [−0.04,−0.02] 
Political ideology (Ref.: Left)   
Centre 0.05*** [0.03,0.06] 
Right 0.08*** [0.06,0.11] 
Don’t know 0.05*** [0.02,0.07] 
Constant 0.39*** [0.35,0.43] 
N 21 
N 26470 
(1)
Experimental groups (age of the unemployed person)  
Control (no age)   
20–25 years old 0.02*** [0.01,0.02] 
in their 50s −0.01* [−0.03,−0.00] 
Individual level     
Female 0.02** [0.01,0.03] 
Age (Ref.: <30years)  
30–59 years old 0.02** [0.01,0.03] 
60+ 0.04*** [0.02,0.05] 
Education (Ref.: ES – ISCED I)   
ISCED II 0.01 [−0.01,0.03] 
ISCED III −0.00 [−0.01,0.02] 
ISCED IV −0.01 [−0.03,0.01] 
ISCED V −0.04* [−0.06,−0.01] 
Employment status (Ref.: Paid work)   
Unemployed −0.06*** [−0.09,−0.03] 
Retired −0.00 [−0.02,0.01] 
In education, housework, other −0.03*** [−0.05,−0.02] 
Unemployment experience −0.03*** [−0.04,−0.02] 
Political ideology (Ref.: Left)   
Centre 0.05*** [0.03,0.06] 
Right 0.08*** [0.06,0.11] 
Don’t know 0.05*** [0.02,0.07] 
Constant 0.39*** [0.35,0.43] 
N 21 
N 26470 

95% confidence intervals in brackets.

*p < 0.05, **p < 0.01, ***p < 0.001.

In this paper, we investigate how an unemployed person’s age influences individuals’ willingness to punish this person if he or she refuses a job offer. We build on previous research on deservingness of older unemployed (Buss 2019; Fossati 2018; Larsen 2018) but provide a comparative perspective. To our knowledge, this is the first experimental study to investigate the role of the recipient’s age for support for ALMPs in a cross-country setting (but see Goerres et al.2018; Naumann et al.2018 for cross-country survey experiments on other topics).

In general, it appears that unemployment benefit recipients are not perceived as a homogeneous group and that people do consider the individual situation when applying demanding ALMPs in form of cutting unemployment benefits. Our results confirm previous research and show that age is an important deservingness cue and older unemployed are punished less than younger unemployed (Buss 2019; Petersen et al.2011). These differences in the willingness to sanction an unemployed person conditional on his or her age are small though and we find that countries vary in this regard. Spain and Portugal are outliers in this respect as support to sanction older rather than younger unemployed is higher there. One possible explanation for this pattern of support might be the difficult labour market situation of the young in the aftermath of the financial crisis and the young might be perceived as less in control of their situation. Yet, other countries with a similar economic and labour market context do not show such a attitude pattern (e.g. Italy), so our data do not provide a conclusive explanation for this finding. Hence, future research in the form of country-case studies might add more detailed explanations for the particularities of the two countries.

We expected that older unemployed individuals would be punished less in countries where age stereotypes are more prevalent and that the preferential treatment of older individuals will decrease in societies where they make up a larger share of the population. Surprisingly little support could be found for the first expectation, whereas older unemployed seem to be treated less as a special group in the need of protection and are not exempted from the obligation to accept job offers if they made up a larger share of the population.

There are some limitations of our study, both methodological and theoretical, that could explain why no effect of ageism on our dependent variable was found. Methodologically, one should keep in mind that we operationalized age stereotypes as the perception of age-related discrimination. However, the perception of age discrimination also depends on identifying ageist behaviour as a problem biasing our indicator of age-related discrimination. This poses critical questions with the common practice of measuring actual discrimination with perceptions of discrimination in the ageism literature. Theoretically, we argued that strong age stereotypes would increase older individuals’ deservingness. This argument neglects that strong age stereotypes and age discrimination might also indicate a low level of identification with older individuals. As such, the mere existence of age stereotypes could lead to harsher sanctioning of older individuals. Moreover, the respondents, although not holding age-stereotypes themselves, might take potential ageist behaviour of the employers and the resulting lower chance of older unemployed individuals in the hiring process into account when evaluating the sanctioning of unemployed. Still, the idea that older unemployed individuals might have less control over their situation, might be cancelled out in our experiment as the question gives full control to the unemployed who deliberately refuses a job offer (Jensen and Petersen 2017).

Nonetheless, if our finding that ageism does not affect deservingness would be confirmed in future studies, we would need to rethink where deservingness conceptions come from (Jensen and Petersen 2017). Possibly, discrimination in terms of deservingness is not built on stereotypes, but instead on prejudice. People might assign deservingness based on a ‘feeling’ (prejudice) rather their ‘knowledge’ (stereotypes) about a certain group (see also Petersen et al.2012).

The paper contains some important policy implications. As societies age, support for exempting older individuals from active labour market policies declines. Moreover, in situations of high unemployment, willingness to sanction those turning down a job offer is higher. Hence, opportunities to introduce or strengthening demanding ALMPs will most likely increase in the future and find public approval. Moreover, the congruence of institutional rules and public support for demanding ALMPs suggests that policy-makers do not have to expect a backlash in popular support when introducing ALMPs. The implications of population ageing for ALMPs are both worrying and encouraging. It is encouraging that older people are perceived as more equal in society, though it remains to be seen whether this also applies to other areas of the welfare state and not only to demanding ALMPs. What is worrying though are the societal implications of further strengthening demanding ALMPs (see also Fervers 2019; Greer 2016). Stricter sanctioning to refuse a job offer bears the risk of forcing individuals into jobs with unfavourable working conditions. This risk might materialize in particular among disadvantaged groups in the labour market, such as older workers, facing discrimination in hiring and labour market integration processes. Moreover, this might negatively affect older workers’ ability to continue working until retirement age. Therefore, tailor-made policies remain necessary to cope with their continuing position of disadvantage in the labour market. Further, we would suggest that policy-makers complement demanding ALMPs for older jobseekers with enabling ALMPs, such as targeted training and reintegration programs, and also to allow exit routes for those who have little chances of finding a job due to qualification or health reasons.

No potential conflict of interest was reported by the authors.

Elias Naumann is JFK Memorial Fellow at the Minda de Gunzburg Center for European Studies at Harvard University. Previously, he was Postdoc researcher at the Collaborative Research Center “The Political Economy of Reforms” at the University of Mannheim where he also gained his PhD in 2014. His research interests include political sociology, comparative political economy and public attitudes towards the welfare state.

Wouter De Tavernier is a Postdoc Researcher at the Center for Social and Cultural Psychology, KU Leuven (Belgium). His research interests include employment of older workers, retirement, pensions and ageism. Before, he worked as a researcher at the Centre for Comparative Welfare Studies, Aalborg University (Denmark) where he also obtained his PhD, and at the Center for Sociological Research, KU Leuven.

Laura Naegele is a Post-Doctoral Researcher at the Institute for Gerontology, Department of Ageing and Work at the University of Vechta, Germany. Her research topics include ageism in the labour market, ageing workforces and age management/competence management, with a focus on small and medium-sized enterprises (SMEs). Previously, she worked for the Bertelsmann Foundation out of their offices in Gütersloh (Germany) and Brussels (Belgium).

Moritz Hess is a Postdoc Researcher at the Research Centre on Inequality and Social Policy, University of Bremen (Germany). His research foci are older workers, retirement, ageism and comparative welfare state research. Before he worked at the Institute of Gerontology at TU Dortmund and the Mannheim Centre for European Social Research.

1

We are aware that it is a strong assumption that the answer ‘should lose small part of unemployment benefit’ means a cut of 25%. Yet, from the construction of the scale the value should be somewhere between 0% and 50% and we decided to work with this assumption in order to be able to keep the substantial meaning of the three other response options. As a robustness check we re-run our analysis with an ordered logit model instead of a linear regression which does not substantially change our results.

2

One should be aware that combining these two variables might hide some interesting variance as the questions might capture two different aspects of support for sanctions. Sanctioning someone for not accepting a job because of lower pay mainly has consequences for the individual whereas employers and perhaps the society might benefit from lower costs. However, forcing unemployed to accept jobs below their education might lead to a suboptimal allocation of the workforce and hence to overall economic losses in productivity. Yet, our decision to combine these two variables is motivated by three rather methodological considerations. First, the psychological and survey methodological literature strongly recommends using multi-item measures (instead of single-item measures) as they are more reliable (Moore et al.2002). Second, combining the two variables adds variation to the measurement scale, increases the number of observed values from 4 to 8 (on the 0 to 1 scale of the latent support for sanctions) and hence comes closer to representing a continuous variable. This is the assumption for using a linear regression model. Third, the correlation between the two variables is high (r = 0.69) and our overall results do not differ if we use the single-item dependent variables (see Appendix D).

3

In the Appendix, we provide the detailed distribution of answers across experimental groups (Table A2).

Acock
,
A.
(
2005
) ‘
Working with missing values
’,
Journal of Marriage and Family
67
:
1012
28
. doi:
Ayalon
,
L.
(
2013
) ‘
Feelings towards older vs. younger adults: results from the European social survey
’,
Educational Gerontology
39
(
12
):
888
901
. doi:
Ayalon
,
L.
and
Tesch-Römer
,
C.
(
2018
) ‘Introduction to the section: ageism – concept and origins’, in
L.
Ayalon
and
C.
Tesch-Römer
(eds.),
Contemporary Perspectives on Ageism.
International Perspectives on Aging, vol
19
.
Cham
:
Springer
, pp.
1
10
.
Bloemen
,
H.
(
2016
) ‘
Job search requirements for older unemployed workers
’,
IZA World of Labor
235
.
Bonoli
,
G.
(
2013
)
The Origins of Active Social Policy: Labour Market and Childcare Policies in a Comparative Perspective
,
Oxford
:
Oxford University Press
.
Bryan
,
M.
and
Jenkins
,
S.
(
2016
) ‘
Multilevel modelling of country effects: a cautionary tale
’,
European Sociological Review
32
(
1
):
3
22
. doi:
Buss
,
C.
(
2019
) ‘
Public opinion towards targeted labour market policies: a vignette study on the perceived deservingness of the unemployed
’,
Journal of European Social Policy
29
(
2
):
228
40
. doi:
Buss
,
C
,
Ebbinghaus
,
B.
and
Naumann
,
E.
(
2017
) ‘Making deservingness of the unemployed conditional: changes in public support for the conditionality of unemployment benefits’, in
W.
van Oorschot
,
F.
Roosma
,
B.
Meuleman
and
T.
Reeskens
(eds.),
The Social Legitimacy of Targeted Welfare: Attitudes to Welfare Deservingness.
Cheltenham, UK
:
Edward Elgar Publishing
, pp.
167
86
.
Butler
,
R.N.
(
1969
) ‘
Age-Ism: another form of bigotry
’,
The Gerontologist
9
(
4
):
243
6
. doi:
Carriero
,
R.
and
Filandri
,
M.
(
2018
) ‘
Support for conditional unemployment benefit in European countries: the role of income inequality
’,
Journal of European Social Policy
, online first.
Cavaillé
,
C.
and
Trump
,
K.-S.
(
2015
) ‘
The two facets of social policy preferences
’,
The Journal of Politics
77
(
1
):
146
60
. doi:
Cowgill
,
D.O.
and
Holmes
,
L.
(
1972
)
Aging and Modernization
,
New York
:
Appleton-Century-Crofts
.
Dordoni
,
P.
and
Argentero
,
P.
(
2015
) ‘
When age stereotypes are employment barriers: a conceptual analysis and a literature review on older workers stereotypes
’,
Ageing International
40
(
4
):
393
412
. doi:
Ebbinghaus
,
B.
and
Naumann
,
E.
(Eds.). (
2018
).
Welfare State Reforms Seen from Below – Comparing Public Attitudes and Organized Interests in Britain and Germany.
London
:
Palgrave Macmillan
.
Eichhorst
,
W.
and
Konle-Seidl
,
R.
(
2008
) ‘
Contingent convergence: a comparative analysis of activation policies
’,
IZA Discussion Papers
3905
.
Fervers
,
L.
(
2019
) ‘
Economic miracle, political disaster? Political consequences of Hartz IV
’,
Journal of European Social Policy
29
(
3
):
411
27
. doi:
Fossati
,
F.
(
2018
) ‘
Who wants demanding active labour market policies? Public attitudes towards policies that put pressure on the unemployed
’,
Journal of Social Policy
47
(
01
):
77
97
. doi:
Fraile
,
M.
and
Ferrer
,
M.
(
2005
) ‘
Explaining the determinants of public support for cuts in unemployment benefits spending across OECD countries
’,
International Sociology
20
(
4
):
459
81
. doi:
Gallego
,
A.
and
Marx
,
P.
(
2017
) ‘
Multi-dimensional preferences for labour market reforms: a conjoint experiment
’,
Journal of European Public Policy
24
(
7
):
1027
47
. doi:
Goerres
,
A.
,
Karlsen
,
R.
and
Kumlin
,
S.
(
2018
) ‘
What makes people worry about the welfare state? A three-country experiment
’,
British Journal of Political Science
, online first.
Greer
,
I.
(
2016
) ‘
Welfare reform, precarity and the re-commodification of labour
’,
Work, Employment and Society
,
30
(
1
):
162
73
. doi:
Halvorsen
,
R.
and
Jensen
,
P.H.
(
2004
) ‘
Activation in Scandinavian welfare policy: Denmark and Norway in a comparative perspective
’,
European Societies
6
(
4
):
461
83
. doi:
Heisig
,
J.P.
and
Schaeffer
,
M.
(
2019
) ‘
Why you should always include a random slope for the lower-level variable involved in a cross-level Interaction
’,
European Sociological Review
35
(
2
):
258
79
. doi:
Heisig
,
J.
,
Schaeffer
,
M.
and
Giesecke
,
J.
(
2017
) ‘
The costs of simplicity: why multilevel models may benefit from accounting for cross-cluster differences in the effects of Controls
’,
American Sociological Review
82
(
4
):
796
827
. doi:
Hofäcker
,
D.
,
Hess
,
M.
and
Naumann
,
E.
(
2015
) ‘Changing retirement transitions in times of paradigmatic political change: toward growing inequalities’, in
C.
Torp
(ed.),
Challenges of Aging: Pensions, Retirement and Generational Justice.
Basingstoke
:
Palgrave Macmillan
, pp.
205
26
.
Inglehart
,
R.
and
Welzel
,
C.
(
2005
)
Modernization, Cultural Change, and Democracy. The Human Development Sequence
,
Cambridge
:
Cambridge University Press
.
Jensen
,
C.
and
Petersen
,
M.B.
(
2017
) ‘
The deservingness heuristic and the politics of health care
’,
American Journal of Political Science
61
(
1
):
68
83
. doi:
Koster
,
F.
(
2018
) ‘
Age stereotypes and attitudes towards welfare state arrangements for the old: a multilevel analysis across twenty-nine countries
’,
Journal of Sociology & Social Welfare
45
(
2
):
67
90
.
Kulin
,
J.
and
Meuleman
,
B.
(
2015
) ‘
Human values and welfare state support in Europe: an east–west divide?
’,
European Sociological Review
31
(
4
):
418
32
. doi:
Kumlin
,
S.
and
Stadelmann-Steffen
,
I.
(
2014
)
How Welfare States Shape the Democratic Public: Policy Feedback, Participation, Voting, and Attitudes
,
Cheltenham
:
Edward Elgar Publishing
.
Laenen
,
T.
(
2018
) ‘
Do institutions matter? The interplay between income benefit design, popular perceptions, and the social legitimacy of targeted welfare
’,
Journal of European Social Policy
28
(
1
):
4
17
. doi:
Langenbucher
,
K.
(
2015
)
How Demanding are Eligibility Criteria for Unemployment Benefits, Quantitative Indicators for OECD and EU Countries.
OECD Social, Employment and Migration Working Papers, No. 166.
Paris
:
OECD Publishing
.
Larsen
,
C.A.
(
2008
) ‘
The political logic of labour market reforms and popular images of target groups
’,
Journal of European Social Policy
18
(
1
):
50
63
. doi:
Löckenhoff
,
C.E.
,
de Fruyt
,
F.
,
Terracciano
,
A.
,
McCrae
,
R.R.
,
De Bolle
,
M.
,
Costa
,
P.T.
,
Aguilar-Vafaie
,
M.E.
,
Ahn
,
C.-k.
,
Ahn
,
H.-n.
,
Alcalay
,
L.
,
Allik
,
J.
,
Avdeyeva
,
T.V.
,
Barbaranelli
,
C.
,
Benet-Martínez
,
V.
,
Blatný
,
M.
,
Bratko
,
D.
,
Cain
,
T.R.
,
Crawford
,
J.T.
,
Lima
,
M.P.
,
Ficková
,
E.
,
Gheorghiu
,
M.
,
Halberstadt
,
J.
,
Hřebíčková
,
M.
,
Jussim
,
L.
,
Klinkosz
,
W.
,
Knežević
,
G.
,
de Figueroa
,
N.L.
,
Martin
,
T.A.
,
Marušić
,
I.
,
Mastor
,
K.A.
,
Miramontez
,
D.R.
,
Nakazato
,
K.
,
Nansubuga
,
F.
,
Pramila
,
V.S.
,
Realo
,
A.
,
Rolland
,
J.-P.
,
Rossier
,
J.
,
Schmidt
,
V.
,
Sekowski
,
A.
,
Shakespeare-Finch
,
J.
,
Shimonaka
,
Y.
,
Simonetti
,
F.
,
Siuta
,
J.
,
Smith
,
P.B.
,
Szmigielska
,
B.
,
Wang
,
L.
,
Yamaguchi
,
M.
and
Yik
,
M.
(
2009
) ‘
Perceptions of aging across 26 cultures and their culture-level associates
’,
Psychology and Aging
24
(
4
):
941
54
. doi:
Mau
,
S.
(
2003
)
The Moral Economy of Welfare States: Britain and Germany Compared
,
London, New York
:
Routledge
.
Moore
,
K.
,
Halle
,
T.
,
Vandivere
,
S.
and
Mariner
,
C.
(
2002
) ‘
Scaling back survey scales: how short is too short?
’,
Sociological Methods & Research
30
(
4
):
530
67
. doi:
Naegele
,
L.
,
De Tavernier
,
W.
and
Hess
,
M.
(
2018
) ‘Work environment and the origin of ageism’, in
L.
Ayalon
and
C.
Tesch-Römer
(eds.),
Contemporary Perspectives on Ageism.
International Perspectives on Aging, vol
19
.
Cham
:
Springer
, pp.
73
90
.
Naumann
,
E.
,
Buss
,
C.
and
Bähr
,
J.
(
2016
) ‘
How unemployment experience affects support for the welfare state: a real panel approach
’,
European Sociological Review
32
(
1
):
81
92
. doi:
Naumann
,
E.
,
Stoetzer
,
F.
and
Pietrantuono
,
G.
(
2018
) ‘
Attitudes towards highly skilled and low-skilled immigration in Europe: a survey experiment in 15 European countries
’,
European Journal of Political Research
57
(
4
):
1009
30
. doi:
Ostrom
,
E.
,
Gardner
,
R.J.
and
Walker
,
J.R.
(
1994
)
Rules, Games, and Common-Pool Resources
,
Ann Arbor
:
University of Michigan Press
.
Petersen
,
M.B.
,
Slothuus
,
R.
,
Stubager
,
R.
and
Togeby
,
L.
(
2011
) ‘
Deservingness versus values in public opinion on welfare: The automaticity of the deservingness heuristic
’,
European Journal of Political Research
50
(
1
):
24
52
. doi:
Petersen
,
M.B.
,
Sznycer
,
D.
,
Cosmides
,
L.
and
Tooby
,
J.
(
2012
) ‘
Who deserves help? Evolutionary psychology, social emotions, and public opinion about welfare
’,
Political Psychology
33
(
3
):
395
418
. doi:
Pierson
,
P.
(
1993
) ‘
When effect becomes cause: policy feedback and political change
’,
World Politics
45
(
4
):
595
628
. doi:
Posthuma
,
R.A.
and
Campion
,
M.A.
(
2009
) ‘
Age stereotypes in the workplace: common stereotypes, moderators, and future research directions
’,
Journal of Management
35
(
1
):
158
88
. doi:
Sachweh
,
P.
(
2018
) ‘
Crisis experiences and welfare attitudes during the Great Recession
’,
Acta Sociologica
47
:
5
.
Stegmueller
,
D.
(
2013
) ‘
How many countries for multilevel modeling? A comparison of frequentist and bayesian approaches
’,
American Journal of Political Science
57
(
3
):
748
61
. doi:
Svallfors
,
S.
(
2010
) ‘Public attitudes’, in
F.G.
Castles
,
S.
Leibfried
,
J.
Lewis
,
H.
Obinger
and
C.
Pierson
(eds.),
The Oxford Handbook of the Welfare State
,
Oxford
:
Oxford University Press
, pp.
241
251
.
van Oorschot
,
W.
(
2006
) ‘
Making the difference in social Europe: deservingness perceptions among citizens of European welfare states
’,
Journal of European Social Policy
16
(
1
):
23
42
. doi:
van Oorschot
,
W.
,
Roosma
,
F.
,
Meuleman
,
B.
and
Reeskens
,
T.
(eds). (
2017
)
The Social Legitimacy of Targeted Welfare: Attitudes to Welfare Deservingness
,
Cheltenham, UK
:
Edward Elgar Publishing
.
van Prooijen
,
J.-W.
(
2018
)
The Moral Punishment Instinct
,
New York
:
Oxford University Press
.
Voss
,
P.
,
Bodner
,
E.
and
Rothermund
,
K.
(
2018
) ‘Ageism: the relationship between age stereotypes and age discrimination’, in
L.
Ayalon
and
C.
Tesch-Römer
(eds.),
Contemporary Perspectives on Ageism
,
Cham
:
Springer
, pp.
11
31
.
Walker
,
A.
and
Maltby
,
T.
(
2012
) ‘
Active ageing: A strategic policy solution to demographic ageing in the European Union
’,
International Journal of Social Welfare
21
(
2
):
S117
30
. doi:
Weishaupt
,
J.T.
(
2010
)
From the Manpower Revolution to the Activation Paradigm. Explaining Institutional Continuity and Change in an Integrating Europe
,
Amsterdam
:
Amsterdam University Press
.
Wlezien
,
C.
(
1995
) ‘
The public as thermostat: dynamics of preferences for Spending
’,
American Journal of Political Science
39
(
4
):
981
1000
. doi:

Author notes

*

Present address: Center for Social and Cultural Psychology (CSCP), KU Leuven, Leuven, Belgium

This is an open-access article distributed under the terms of the Creative Commons Attribution-NonCommercial 4.0 International (CC BY-NC 4.0) license, which permits unrestricted use, distribution, and reproduction in any medium, provided the use is non-commercial and the original work is properly cited. For a full description of the license, please visit https://creativecommons.org/licenses/by-nc/4.0/legalcode.

Supplementary data