27.1 Introduction

It’s been almost 50 years since the term ‘ageism’ was first introduced in recognition of prejudicial attitudes held towards people because of their age. Since then, research in Europe has consistently shown that more people report experiencing age prejudice and discrimination followed by prejudice based on gender, and then race or ethnic background (Abrams et al. 2011a; Abrams and Houston 2006). This trend is found in most countries in the European region with the exception of Israel and Latvia (Ayalon 2014). On average 35% of people who took part in the European Social Survey (ESS) said they had experienced ‘unfair’ treatment because of their age, but these experiences of ageism differ across Europe, from 54% in the Czech Republic to 17% in Portugal and Cyrus (Abrams et al. 2011a). Despite its relatively high prevalence, ageism is still under researched (Abrams et al. 2015) and remains a relatively accepted form of prejudice, deemed to be an inevitable part of the ageing process (Nelson 2005; Nelson 2017). Differences between countries in the prevalence of experienced ageism could be related to various contextual differences that exist for these countries. Across Europe rates of population ageing, fertility, immigration and emigration differ and are leading to changing patterns of demography (Creighton 2014). Although Europe as a whole must adapt to population ageing, changing demographic patterns across Europe, coupled with substantial socio-economic, cultural and political differences, are likely to impact on people’s experiences of ageing, and could also explain why people in some countries are more or less likely to experience ageism. Cross-national data, such as the ESS, have allowed researchers to identify some of the psychological and sociological factors associated with people’s experiences of ageism.

In this chapter, we review the existing social psychological framework used to explore ageism which informed the design of the Experiences and Expressions of Ageism Module fielded in the 2008/2009 ESS. We also review the subsequent published findings and studies that have utilized the ESS data to explore how the availability of cross-national data has advanced theory and understanding of ageism across Europe. Much of this evidence uses a multilevel approach. Thus, we also explore the benefits and limitations of such an approach for understanding ageism across Europe.

27.2 The European Social Survey

The European Social Survey (ESS) is an academically driven cross-national survey conducted every 2 years across the European region since 2001. The survey measures attitudes, beliefs and behavior patterns in (up to) 30 nations. Each survey consists of a fixed module and two or three rotating modules. Round 4 of the survey fielded in 2008 and 2009 included a rotating module on Experiences and Expressions of Ageism (herein referred to as the Ageism Module). The module was designed by members of the EURAGE group (European Research group on Attitudes to Age) lead by Prof. Dominic Abrams (University of Kent), Prof. Luisa Lima (CIS-IUL at ISCTE-IUL) and Prof. Genevieve Coudin (University of Paris). It contains 55 items which were developed and pilot tested extensively within a framework that was subject to detailed scrutiny, peer review and evaluation by experts in the ESS Central Coordinating Team. The Ageism module provides representative samples from 29 countries and over 55,000 individuals. It uses computer-based personal interviews with national samples between 1215 and 2576 people aged 15 and over. The questions in the module reflect seven key domains including how age boundaries are defined, age identification, the content of age-based stereotypes, perceived social status, perceived threat, experiences of discrimination and the extent of intergenerational relations. The items within these domains are guided by theoretical models from social psychology theories of prejudice, including social identity and self-categorisation theory (Tajfel 1981), stereotype content model (Fiske et al. 2002), intergroup threat theory (Stephan and Stephan 2000) and intergroup contact theory (Pettigrew 1998).

This chapter introduces each of these domains and reviews key findings from published research that has used the Ageism module. The search for published research was conducted using Google Scholar in 2017. First, key search terms ‘Ageism’ and ‘European Social Survey’ were entered. The search, which was restricted to publications after 2008, resulted in 391 hits. Each article (conference proceedings, presentations were excluded) was read to make an initial judgement about its relevance. We were only interested in academic or grey literature (i.e. research published in academic journals or by government or non-government organizations) that used, analyzed or reported findings from the Ageism Module. We stopped the search at the point when all of the articles listed were irrelevant (page 20 of Google Scholar search). This resulted in 36 articles for review which were then categorized according to the primary or most relevant domain. Before we review the findings it important to note that many of the publications adopt a multilevel approach to analyzing ESS data.

27.2.1 Multilevel Modelling

The ESS data constitute individual responses that are grouped or clustered by the country they are from. This hierarchical structure (i.e. individuals grouped within countries) is conducive to multilevel modelling. Multilevel modelling is a statistical regression approach that accommodates the context in which individuals are living (Raudenbush and Bryk 2002), thus allowing for the simultaneous examination of phenomenon at both the societal (e.g. country)- and individual-level. In order to explain differences between countries, country-level (also known as macro-level) variables are used as predictors of individuals’ attitudes and experiences. These can be socio-cultural, but also socio-structural contexts. The first refers to collective social phenomena such as widely shared social representations, for example, about older people’s social status in society, or societal meta-perceptions (see Vauclair et al. 2016). The latter refers to structural aspects of society such as economic and political systems of the country, for instance a country’s gross domestic product (GDP), level of inequality (as indicated by the GINI index) or a country’s average retirement age. All of which can have an effect on different aspects of expression of prejudice and experiences of ageing.

The benefit of the multilevel approach is the examination of psychological phenomena of expressions of ageism (e.g. age stereotypes, perceived threat, perceived social status of older people) which do not emerge in social vacuum. It is known that groups and whole societies recognize and tend to share similar stereotypes and prejudices (Schaller et al. 2002) which constitute a collective reality that has an impact on older people in terms of their ageing experiences. Thus, the ESS allows for research to be conducted at two levels of analysis and therefore can independently explore effects at the individual and country-level, as well as provide insight into what kind of the combination of these factors are most likely to ameliorate or exacerbate ageism. Some of the factors might be difficult to change (e.g., a country’s affluence), however, other factors might be more malleable. For instance, employment of older people may be raised through ‘age quotas’. Cultural beliefs may also (slowly) change through targeted information, because culture is after all dynamic. Interventions can be targeted by taking into account the different layers of effects that produce negative age stereotyping. Furthermore, country-level findings (e.g., the effect of unemployment rates or the age ratio) can be translated to the regional-level within countries which will be of great interest for policy-makers within countries in Europe. The following sections summarise the empirical studies that were identified as relevant in our review.

27.3 Ageism in Europe

27.3.1 Age Categorisation

In order to understand who is vulnerable to age discrimination it is important to understand how people define age groups. Age categorisation is the process of classifying people as belonging to a certain age group, and by implication, not belonging to other age groups. Age categorisation is highly relevant to the issue of age-based discrimination because ageism can arise in relation to specific age points, for example, the age of retirement or particular age ranges (e.g. the over 50’s), but also in terms of general category labels such as ‘young’ or ‘old’ (Bytheway 2005).

Once categorisation has occurred, stereotypes associated with the category label might then be erroneously applied to the individual. This can result in discrimination, but also put individuals (particularly older adults) in a situation where they fear being ‘judged’ negatively in terms of their age (cf. stereotype threat, Lamont et al. 2015). Socially and psychologically the use of age categorisation can be highly problematic because it may cause people to restrict what is deemed acceptable for particular age groups based on their own or others’ ageist assumptions. For example, people may be judged or they may see themselves as ‘too young’ or ‘too old’ to pursue particular activities or roles. For this reason, the very act of categorising others into different age groups and the way people define those groups has significant implications for people’s choices and actions (see The Risks of Ageism Model, Swift et al. 2017).

However, understanding the process of age categorisation can be tricky. Unlike other forms of prejudice, such as racism and sexism, which are directed at fixed social categories, the boundaries that define ‘old’ and ‘young’ age groups are fluid (Bytheway 2005). In the Ageism module respondents were asked to estimate the age at which people stop being described as ‘young’ and to estimate the age at which people start being described as ‘old’. On average, across the 29 countries, respondents perceived youth to end at 40 years and old age to begin at 62, this may mean that people below 40 years and over 62 years are more vulnerable to age prejudice and discrimination due to their perceived ‘young’ or ‘old’ age, respectively. However, the placement of these age boundaries varies by respondent’s own age, the self-categorization of their age, gender and the country in which they live (Abrams et al. 2011a; Ayalon et al. 2014; Basleven 2010), such that both perceptions of the end of youth and onset of old age increase with respondents’ age. In addition, respondents who self-categorized themselves as belonging to a younger age group (relative to those who are the same chronological age) also perceived old age to start later (Basleven 2010). Women perceived the end of youth and onset of old age to be later than did men, and the end of youth was viewed as early as 34 in Norway, but as late as 52 in Greece, while old age was viewed as starting at 65 in Greece, but 55 in Turkey (Abrams et al. 2011a) (Fig. 27.1).

Fig. 27.1
figure 1

Estimated end of youth and onset of old age

Analysis by Ayalon et al. (2014) using the full ESS data confirmed that 14% of the total variance in the perceived end of youth was associated with differences between countries, while only 5.7% of the total variance in the perceived onset of old age was associated with differences between countries. In addition to the previous study, the authors found that having better subjective health and living with a spouse or partner was also associated with perceiving the end of youth to be later, while higher levels of education, better subjective health, higher life satisfaction and sharing a residence with a spouse or partner were all associated with perceiving old age to be later. Taking a multilevel approach, Ayalon et al. (2014) also examined how contextual factors, such as country-level differences in life expectancy or inequality, are associated with each age-boundary. The study revealed that respondents in countries with a higher life expectancy, higher levels of inequality and a higher percentage of individuals between the ages of 15–64 were more likely to perceive the end of youth to be later. Whereas, respondents in countries with higher education and higher levels of inequality, were more likely to perceive old age to be later. However, official retirement age and fertility rates were not associated with either age boundary.

Together, this research suggests that people have different constructions of age depending on their vantage point, depending on personal, shared and interdependent social contexts of the individual, but also on some culturally shared social contexts (e.g. education, life expectancy and inequality). Understanding how age boundaries are socially constructed within societies is important because they can shape individual’s social construction of what age is, and what it means to belong to a particular age group. In this way, age categorisations can impact on individuals via their own behaviours and life choices, but also by constructing attitudes towards various age groups (Swift et al. 2017).

27.3.2 Age Identification

One factor that is likely to influence the salience of age in social contexts, and therefore the likelihood of age categorisation processes and the subsequent application of stereotypes is age identification. Age identification is the extent to which people identify with an age category and can be an important measure of the extent to which age, and belonging to an age group informs social identity and who we are (Tajfel and Turner 1979; Tajfel 1981). Social Identity Theory states that individuals are motivated to gain positive distinctiveness for their ingroups by comparing them favourably with other groups, which can result in holding prejudice views against outgroup members. Shared age identity is assumed to denote an awareness of similarities, a feeling of solidarity towards others in the same position but this, by implication can also be a potential antagonist against those who are different. Therefore, age identity could also be an important factor in shaping generational conflict, for instance, over the distribution of resources.

The meaning of age identification can be ambiguous unless it clarifies what age people identify with. For instance, research conducted with individuals aged 65 and over revealed that those with more positive perceptions of ageing (measured by self-ratings of cognitive and physical functioning tasks) typically have a more youthful age identity, i.e. participants felt younger. These individuals also had a higher level of satisfaction with their current age, perceived the onset of old-age to be later and showed more willingness to live to 100 years of age (Uotinen et al. 2003). Therefore, it cannot be assumed that age identification measures identification to chronological age, as this research demonstrates individuals can identify with their subjective age – the age they feel, which might be different from their chronological age.

Accordingly, in the ESS, respondents were first asked to ‘self-categorize’ themselves by picking a box from a card with nine possible options laid out horizontally. The first three boxes (A, B and C) were described as ‘young’, the next three (D, E and F) as ‘middle’ and the final three (G, H, and J) as ‘old’. Respondents were asked ‘which box best describes the age group you see yourself as belonging to? If you see yourself as very young, pick the first box. If you see yourself as very old, pick the last box. Otherwise pick one of the boxes in between. Fitzgerald et al. (2010) analysis of this item (for UK only) revealed that the oldest ‘young’ category and the two younger ‘middle-aged’ categories were most popular accounting for half of the responses from respondents aged 18 to 59. Interestingly, around half of those in their 40s (49%) and 50s (52%) chose the mid-point of the “middle” category on the scale (E). Perhaps this reflects a desire to be perceived as precisely in mid-life, aware that ‘young’ is no longer an option but not ready yet to accept being labelled ‘old’. Most notably, 25% of respondents aged 70 and over defined themselves as ‘middle aged’ and only a small minority (14%) choose the highest “old” label (J). Importantly, the analysis by Fitzgerald et al. (2010) suggests two things; (1) people of the same age do not necessarily identity with the same age group, and (2) many people chose to identify with an age group that is ‘younger’ than what might be assigned according to chronological age.

In the ESS, respondents were then asked whether they had a weak or strong sense of belonging to the age group they had chosen, using a scale from 0 (‘very weak sense of belonging’) to 10 (‘very strong sense of belonging’). Analysis of this item, has tended to find the relationship between age, self-categorized age and age identification to be a curvilinear ‘U’ shape, with younger and older age groups showing stronger identification to their respective age groups (Abrams et al. 2011a, b; Fitzgerald et al. 2010; Marques et al. 2014a). This suggests that age identification is more important for younger and older age groups, or that age identity is more salient for these age groups, perhaps due to transitional changes occurring around them (e.g. becoming a young adult or moving into retirement). It could also suggest that these age groups are more likely to be motivated to maintain a positive distinction and therefore, more likely to hold ageist attitudes against one another, which in turn means people belonging to these age-groups may be more vulnerable to age-discrimination from other age groups.

However, social identity also can have important positive consequences for self-esteem, health and wellbeing (Whitbourne and Sneed 2002; Marques et al. 2014a) and the literature thus far, has mostly been concerned with exploring how older adults are able to maintain a high level of age identification, self-esteem and wellbeing in later life, despite the existence of ageism and other negative perceptions associated with ageing (e.g. paradox of wellbeing, Swift et al. 2013; Whitbourne and Sneed 2002). Research suggests a positive age-identity is one mechanism by which older adults are able to maintain positive self-esteem, sense of self and wellbeing. The other mechanisms include having a developed set of coping mechanisms, having the ability to select emotionally rewarding social partners (see socio-emotional selectivity theory, Carstensen 1992) and being able to shift goals and standards to what is achievable in later life (e.g. accommodation; Whitbourne and Sneed 2002).

Research exploring the associations of age identity on older people has been mixed. For instance, older individuals who perceived themselves as “old” also rated their health as poorer than older individuals who perceived themselves as younger (Stephan et al. 2012). But, in a study of 60 people aged 64 and over, Garstka et al. (2004) showed that when older people are faced with age discrimination, increasing their identification with their age group may be an important “fighting” strategy to increase well-being. However, the impact of social identification on self-esteem, health and wellbeing is dependent particularly on the perception of the social status associated with the group (Ellemers 1993; Ellemers et al. 1988; Tajfel and Turner 1979). In line with this, research using the ESS data suggests that these mixed effects of age identification are because age identity can confer a positive or negative image of ageing, depending on the perceived social status of older people (Marques et al. 2014a).

27.3.3 Social Status

The status of a group is determined by its relative place or position in society (Parsons 1951). In the ESS, social status refers to prestige, social standing or position ascribed to individuals that mark their position in a given social system. Evidence form the ESS demonstrates that people in their 40s are perceived to have the highest status, followed by people in their 20s and people over 70 are perceived to have the lowest status (Abrams et al. 2011a). Understanding social status is important because ageism is partly driven by societal representations of older people as being inferior to middle-aged people in terms of power and social status, wealth, respect and influence (Garstka et al. 2004). These perceptions determine how we deal and interact with members of these groups, and how we feel and identify with our own social groups.

In our recent paper, we argue that in order to understand the impact of age identification on older people’s self-reported health, we have to understand the role played by the perceived social status of old age group in society (Tajfel and Turner 1986; Marques et al. 2014a). For instance, older people’s subjective evaluation of their social status is negatively associated with their self-rated health, depression and long-standing illness (Demakakos et al. 2008; cf. Hu et al. 2005). According to Wilkinson (1996) the perceptions of one’s place in the social hierarchy are linked with negative emotions (i.e., feeling of shame) and may influence fundamental health outcomes via neuroendocrine mechanisms. Belonging to a subordinate category potentially confers a negative social identity which can have negative consequences for self-esteem and well-being. On the other hand, belonging to a high-status category creates feelings of pride and positive outcomes for one’s self-esteem. Therefore, the effects of age identity on self-esteem and subjective-health (i.e. self-rated perception of health) are very much dependent on the social standing or position of that group in society. Our analysis of a subsample of respondents aged 70 and over revealed that perceived status moderates the negative effect of age identification on older people’s subjective health. Firstly, if age identification is high, there is a greater impact of social status on subjective health. Secondly, age identification is negatively related to subjective health in countries where older people are perceived to have lower status (Marques et al. 2014a). In countries where old age is perceived as signifying low status, just acknowledging that one is old is likely to present real risks to one’s health, a possibility that is rarely considered in health policy but should be considered by public policy addressing health issues (WHO 2002). The research provides an important theoretical contribution by demonstrating that the perceived social status of people over 70 is an important determinant of the extent to which age identification has negative consequences on older people’s subjective health and clarifies some previously mixed evidence concerning the effects of age identification.

Although, across ESS countries there is a general consensus that people aged 70 and over are considered to have lower social status, there are considerable differences depending upon the perceivers’ age and also between European countries. For instance, the perceived social status of people aged 70 and over increases with respondents age (Abrams et al. 2011b; Marques et al. 2014a), which is likely to be due to increasing motivation to gain positive distinctiveness for the next age group they will belong to. There are also considerable cross-country differences in the amount of status afforded to older people. For instance, in the ESS 16.27% of the total variance in the perceived social status of people over 70 was associated with differences between countries (Vauclair et al. 2014b). Our research findings also show that societal perceptions of older people’s status are positively related to cross-country differences in Gross Domestic Product (GDP), state pension age, the societal emphasis on autonomy values and levels of social inequality indicated by the GINI index (Abrams et al. 2011b). Furthermore, a separate analysis revealed that both societal modernisation (a sociological construct referring to profound economic, social and cultural transition from ‘pre-modern’, ‘traditional’ societies to ‘modern’ societies, indicated by the Human Development Index) and levels of employment of people aged 65 and over are positively associated to the perceived status of people over 70. However, these two factors also interact with each other. The perception of older people’s status is boosted in less modern societies if there is a relatively high employment rate of people aged 65 and over indicating that their active contribution to the economy is credited with more positive representations (Vauclair et al. 2014b). The finding suggests that maintaining the employment of older people may be an important factor determining their personal living standards and providing more positive representations of the status of older people across society. Moreover, low social status of people over 70 can be particularly problematic for older people who identify with their age, as this can have a stronger negative effect on health.

27.3.4 Age Stereotypes

Research suggests that there are multiple representations and stereotypes of ageing, which vary in the extent to which they are negative and positive. For instance, research from the US, UK and across Europe has continued to suggest that older people are stereotyped as frail, ill, dependent and incompetent (Levy 2009; Marques et al. 2014b, Coudin and Alexopoulos 2010), but also wise, experienced and more moral than younger adults (Levy 1996; Swift et al. 2013; Abrams et al. 2011a). Several studies have explored age-related stereotypes across age and cultural groups. Many of the negative and positive representations of ageing can be captured within the stereotype content model, which has been supported by over 10 years of national and international research. It proposes that younger and older age groups can be evaluated along two basic dimensions of competence and warmth (otherwise referred to as friendliness; Fiske et al. 2002; Cuddy et al. 2005). Work conducted in the UK and across Europe, (e.g., Abrams et al. 2009, 2011a; Vauclair et al. 2010) has shown repeatedly that older people are stereotyped with a mixed representation of high warmth (positive), but low competence (negative). This mixed representation results in feeling pity for older people, thus society holds ‘benevolent’ or patronising ‘doddery but dear’ views of older people (Cuddy et al. 2005) –older people are liked but often patronised and not given power or voice because of their perceived low status and declining competence. In contrast, younger targets are characterised by high competence, but relatively low warmth. This is also a mixed perception, but results in feelings of envy and underpins a more hostile form of prejudice (Fiske et al. 2002). Thus, the stereotype content model does provide a framework that proposes that different societal representations of younger and older adults informs the expressions of prejudices and types of discrimination they are likely to experience. In the ESS, the stereotype content items were presented as meta-perceptions, i.e. they asked ‘to what extent do you think other people think people over 70 are viewed as…competent, friendly’. Therefore, they represent a view and perspective that is thought to be widely shared by others.

The literature search found three papers that explored the role of stereotypes. The first described here, explored the extent to which competence and warmth moderate determinants of job satisfaction. Using data from Round 4 and 5 of the ESS, the research conducted by Shiu et al. (2015) was primarily interested in exploring how extrinsic rewards (e.g. job security and opportunity for career advancement) and intrinsic rewards (e.g. task discretion and work pressure) relate to job satisfaction. It also hypothesized that each of these relationships should be moderated by societal views of older people’s competence, such that in countries where there is a shared meta-perception that older adults are less competent, there would be a stronger effect of job security on job satisfaction, a stronger negative effect of work pressure on job satisfaction and a weaker effect of opportunity for advancement on job satisfaction. It also hypothesized that societal views of older people as warm and friendly should moderate the relationship between work pressure and job satisfaction, such that the negative relationship between work pressure and job satisfaction should be stronger for older workers in countries where people over 70 are perceived as more warm and friendly. Based on Gouldner’s (1960) norm of reciprocity, the authors suggest meta-perceptions of older people to be warm and friendly places expectations and work pressures on older workers to be more sociable, leading to lower job satisfaction. The multilevel analysis revealed that country-level competence and warmth (aggregated ESS data from Round 4) moderated the effect of work pressure on job satisfaction in the predicted direction, but also found that the relationship between opportunity for advancement and job satisfaction is stronger in countries where older people are perceived as less warm. No other hypotheses were confirmed and there was no support for moderating effects of competence on job security, opportunity for advancement or task discretion.

However, relatedly, the second paper found that perceptions of people aged 70 and over as competent are significantly predicted by participation of older people in paid and volunteer work (Bowen and Skirbekk 2013). The final paper explored the extent to which personal meta-perceptions predict experiences of ageism, using data from respondents aged 70 and over. It revealed that personal meta-perceptions of negative age stereotypes and specific intergroup emotions (pity, envy, contempt) are associated with higher perceived age discrimination. However, at the country-level, only paternalistic meta-perceptions (i.e. warmth and pity) were consistently associated with greater perceived age discrimination (Vauclair et al. 2016).

In sum, given the theoretical connection between stereotypes, prejudice and discrimination, it is important to establish the role of personal and societal meta-perceptions (i.e. those perceptions and views held by individuals and those aggregated to the country-level). The research reveals the importance of countering age-stereotypes at each of these levels, thus effective counter-ageism strategies need to be directed towards and change individual’s perceptions but also collective societal perceptions (Vauclair et al. 2016).

27.3.5 Feelings of Prejudice

The ESS includes both direct and indirect measures of prejudice. Two direct measures ask respondents to rank on a 0–10 scale “how negative or positive you feel towards people in their 20s / 70s”. In general, respondents reported positive feelings towards older and younger age groups, however, these feelings do vary across Europe, with respondents in Turkey reporting the most negative feelings towards people in their 70s and respondents from Latvia and Finland reporting the most positive feelings. A difference score created by Ayalon (2013) indicated in general more positive feelings towards older than younger adults, with only three countries reporting more positive feelings towards younger (Turkey, Greece and Croatia), but an intraclass correlation coefficient of 2.3% revealed little cross-country variability. On the surface, this finding is inconsistent with the common assumption in the literature that older people are the foremost targets of ageism (see Nelson 2017).

Usefully, the ESS data allows us to make comparisons between the degree to which people hold age-prejudices with the reported experiences of discrimination (i.e. the degree to which people are targeted by discriminatory behaviors or practices because of age), and reveals a discrepancy; despite widely shared positive feelings towards younger and older people, there is a relatively high prevalence of age discrimination. However, direct measures of prejudice should be interpreted with some caution as people may not be willing to admit having feelings of prejudice towards older or younger people (Abrams 2010). To overcome this, the ESS also included two indirect measures of prejudice by assessing the extent to which people are internally and externally motivated to be unprejudiced (Plant and Devine 1998).

In an extensive exploration of how individual meta-perceptions, societal meta-perceptions and norms of intolerance predict older people’s experiences of ageism, Vauclair et al. (2016) revealed that in countries in which people think that it is important to be unprejudiced towards other age groups, older people reported experiencing less age discrimination. Moreover, social norms of intolerance of age prejudice had a larger statistical effect on perceptions of age discrimination compared to other societal meta-perceptions, including the perceived status, warmth and competence of people aged 70 and over. The analysis suggests that social norms that promote intolerance to prejudice are an important factor to improve older people’s experiences of ageing.

27.3.6 Experiences of Discrimination

In order to understand the extant and prevalence of ageism it is important to ask people whether they have experienced age discrimination. The Ageism module contains three items to explore this issue, one of these measures can be compared to reported experiences based on gender or ethnicity. The other two items allow us to understand experiences of specific forms of discrimination, such as blatant discriminatory acts (e.g. bad treatment) and more subtle behaviours (e.g. patronizing behaviour or neglect). Importantly, these items coupled with advances in statistical analysis, allow us to explore experienced of age discrimination across countries (Trusinova 2014) and age (Bratt et al. 2018). A recent study used these three items to (1) investigate measurement invariance for perceived age discrimination across countries and across age and (2) to investigate levels of perceived age discrimination across age (Bratt et al. 2018). The study concluded that the three items have sufficient measurement invariance across countries and across age to merit conclusions about levels of perceived age discrimination. The analysis uncovered substantial perceived age discrimination not only among older people, but even more so among younger people. The consistency and international scope of the evidence from the ESS strongly challenges the conventional wisdom that ageism is primarily a problem for the old. Rather, it affects all ages and the relationship between them. The study concluded that more research is required to understand ageism toward younger people and to consider developmental differences and changes over the lifespan.

Using a stepwise linear regression approach Van den Heuvel and van Santvoort (2011) revealed that gender, education, income, belonging to an ethnic minority, life satisfaction, subjective health status, trust in other people and the perceived seriousness of age discrimination in the country, are all related to experiences of age discrimination in a subsample of ESS respondents aged 62 years and over. Interestingly, using ESS data from Round 4, Meeusen and Kern (2016) investigated the generalizability of five types of prejudice (prejudice directed towards immigrants, homosexuality, age-groups, unemployed, and gender) and revealed they were positively related, such that if an individual holds negative attitudes towards one target group, they are also likely to hold prejudicial attitudes towards other target groups as well. This could indicate that any counter-prejudice and discrimination interventions are likely to have spill-over effects for reducing prejudicial attitudes towards other social groups not specifically targeted by an intervention.

27.3.7 Perceived Threat

Some theories of prejudice also contend that negative attitudes towards social groups are associated with the perception that these groups pose various types of threat (e.g. Intergroup Threat Theory, ITT Stephan and Stephan 2000; Riek et al. 2006). These include symbolic threat, the threat to people’s values, culture and way of life and economic threat, the extent to which economic outcomes of one group might be burdened or dependent on those of a different group. Evidence from the ESS demonstrates that older age groups pose greater threats to the economy by being a burden on health care and welfare resources by being perceived to contribute little to the economy (Abrams et al. 2009, 2011a).

Perceptions of threat change with age. Younger people are more likely to perceive older adults as being a threat to the economy (Abrams et al. 2011a, b). However, older people themselves perceive other older people as being a burden on health care resources. Perceptions of threat also depend on the cultural context. Higher state pension age is associated with lower perceived threat to health care resources, GDP, autonomy values and inequality are associated with reduced perception of economic threat (Abrams et al. 2011b).

Perceived threat in the form of the economic contribution of older people has been associated with older people’s participation in entrepreneurship and innovation. Using a subsample of the Eurobarometer of respondents aged between 50 and 74, Kautonen (2012) found that the perceived economic contribution of people over 70 (a macro-level contextual variable derived by aggregating the ESS data) was significantly, negatively related to the probability of an individual thinking about becoming an entrepreneur. In other words, people over 50 seem to be inclined towards considering entrepreneurship in countries where older people are perceived to contribute less to the economy. Kautonen (2012) suggests that a positive perception of the economic contribution of people aged 70 and over in a country may lead to reduced ageism in the workplace and, therefore, a higher demand of the services of older workers in the labour market, both of which decrease the relative attraction of starting a business.

27.3.8 Intergenerational Relations

An important indicator of social inequality and prejudice is the degree to which social groups share the same goals and values and the degree to which they understand each other. The ‘Common ingroup identity’ model (Gaertner and Dovidio 2000) shows that prejudice can be reduced when people perceive they belong to one common group, or belong to different groups that share an overarching community. ESS respondents were asked how they saw people in their 20s and people over 70. On average, across the European Region, respondents were more likely to view people in their 20s and 70s as two separate groups within the same community (49%), 29% thought that people in their 20s and 70s should be seen as individuals, 10% though they belonged to one group, while 12% thought people in their 20s and 70s belonged to two separate groups who were not part of the same community.

A large number of studies in social psychology (see Pettigrew 1998) have shown that experiences of positive contact between members of different groups lead to positive inter-group attitudes and relations. Specifically, the theories about inter-group contact suggest that positive personal relationships, especially friendship with members of other groups, are important determinants for reducing prejudice between different groups. Thus, the degree and type of contact are relevant indicators of exclusion and discrimination of different age groups in a given society (c.f. Schneider 2004). To get a sense of whether respondents have meaningful voluntary interactions with people from different age groups, ESS respondents were asked how many friends, other than family members, they had under the age of 30 and over the age of 70 – age boundaries which most people would regard as clearly distinguishing between young and old. Across the ESS 52% of respondents said they did not have a friend over the age of 70, compared to 31% who said they did not have a friend under the age of 30. The number of respondents who reported having friendships with people under 30 and people over 70 varied by respondents’ age group. Those aged between 15 and 24 were more likely to have other friends under 30, 80% stated they had no friends over the age of 70. People over the age of 70 were more likely to have friendships with those also over the age of 70, but over half of these respondents (70%) said they had no friends under the age of 30 (Abrams et al. 2011a).

Dykstra and Fleischmann (2016a) explored predictors of cross-age friendships. As predicted by the ‘meeting principle’, individuals who operate in settings where there are opportunities for meaningful interactions with people belonging to a different age group, (e.g. age-diverse work places, religious communities and having close ties with different generational family members) are more likely to have cross-age friendships. In addition, individuals with more favourable attitudes towards other age groups are more likely to have cross-age friendships, which supports intergroup contact theory. In a separate analysis Dykstra and Fleischmann (2016b) contend that having greater capacity to ‘actively age’ i.e. greater independence, health, security, might create greater opportunities for meaningful intergenerational interactions. However, the active ageing index, developed by Zaidi et al. (2013) which assesses experiences and potential for active ageing across four domains of employment, social participations, independent living and capacity for active ageing (as a country-level predictor) was not found to be associated with the development of intergenerational friendships (Dykstra and Fleischmann 2016a). This suggests that to improve contact between generations interventions should focus on proving opportunities to interact within people’s immediate environments.

27.4 Conclusions

The ESS data on ageism provide a framework for understanding people’s attitudes to age, and their experiences of age discrimination. The data provide evidence based indicators that can inform policy on issues surrounding age equality and anti-discriminatory legislation for individual countries but also at the European region level. They also provide an insight into avenues for tackling some of the issues that affect older people and all of us as we age, such as negative attitudes to age, isolation and exclusion. There is great value in cross-national data such as the ESS. Thus far, the ESS data have been important for (a) understanding people’s experiences of growing older, (b) understanding processes that contribute to age discrimination, such as how age groups are perceived throughout the life course and (c) understanding how people’s experiences influence their perceptions of age, or how their perceptions of age influence their experiences. Moreover, multilevel approaches to the analysis of ESS data have made significant theoretical and practical contributions to disciplines concerned about ageism and issues associated with ageing, such as, social psychology, sociology, social gerontology, social policy and business studies.