SOCIAL PERCEPTIONS AND THEIR CHANGES ABOUT LIFE COURSE AND TIMING OF KEY LIFE EVENTS IN EUROPE1

People usually think about age as an objective data indicated in their official documents. However, there is a subjective side to it: the age we associate to an individual’s life-cycle – youth, middle age, and old age – depends on the social-economic context we live in. At the same time an individuals’ perceptions about age become elements of demographic trends, as individual decisions about major life events are influenced by perceptions about the ideal age of these event, i.e., which age is ideal to move from the parental home, start a relationship, give birth to the first child or retire from work. These very personal decisions are greatly influenced by the expectations or even norms posed by the broader society and the narrower community we live in. Social norms associated with age may also change in accordance with more general developments in medical science, expansion of education or policy areas related, for example, to childcare services or retirement. In our paper we examine perceptions about the lifecycle and timing of life and their respective changes in a European comparison. The analysis will use data from the Timing of Life questionnaire module of the European Social Survey (ESS) and compare data from Round 4 in 2006 and Round 9 in 2018. Our analysis triangulates the perception of the timing of life, demographic characteristics of societies and the policy context. Our analysis found that perceptions about ages associated to the lifecycle and major life events have shifted: in all respects people tend to associate a later age to these. The most prominent changes happened in societies of Central East Europe with regard to ages associated with the ideal and latest age of childbearing. There is a slow but explicit convergence of perceptions about age in post-communist countries in Eastern Europe with more long-time democracies in Western Europe. The changes are fuelled primarily by young people’s significantly different perceptions compared with the middle-aged and elderly. The trend of later ages seen as an ideal for key life events especially that of childbearing, is an important driver of further in the aging of European societies.


Introduction
Expectations and norms for the different stages of human life are changing and are impacted by broader social, demographic, and political processes, such as increasing life expectancy due to medical advances and increasing prosperity, the expansion of education, the extension of the age of childbearing, and changes in compulsory education legislation. Societal perceptions about age linked to major life events are norms that change in time and across societies (Billari-Liefbroer 2007). However, the perception of age limits is also of great socio-political importance, as these perceptions influence the main determinants of demographic and labour market processes, such as fertility and labour market activity, which in turn feed into actual processes in the economy and society.
In this paper, we examine perceptions and changes in perceptions of the life-course and major life events across Europe. In our analysis, we use data from the European Social Survey (ESS) Timing of Life Module collected in 2006 and 2018 (Billari et al., 2005). We consider the demographic composition of society, people's perceptions of age and major life events, and the public policy environment regarding age (e.g., age of compulsory schooling, age of majority, age of retirement) as a "triangle" where each element is interacting with the other two.
Scholarly literature discussing the social relevance of timing of life offers two broad approaches: the structure-oriented perspective branch of studies focuses on the institutional environment, while the action-oriented perspective examines the individual determinants of life-course and the timing of key life events. The first approach postulates that timing of life is primarily determined by the social, economic, and political context in which the individual lives (Kohli, 2007). According to this approach, education, the labour market, and housing conditions, as well as the public policies that influence them, play a particularly important role in shaping an individual's life cycle, as they set the parameters for the transition from childhood to young adulthood, and from middle to old age. The second approach emphasises the individual determinants of the life-cycle and the timing of key events in life, arguing that life stage decisions are up to individual choices rather than the social context, thus societal norms are less significant. This approach emphasises the agency of the individual (Buchmann, 1989;Heckhausen-Buchmann, 2019;De Vroom, 2004) and argues that the increasing importance of individual choice has led to a kind of second demographic transition (see, e.g., Van de Kaa, 1987;Lesthaeghe, 1995Lesthaeghe, , 2019. This is also referred to by others as the era of second modernity (Beck, 1982), in which the perceived age boundaries of the life-cycle are expanded, leaving room to deviate from socially and economically determined norms (e.g., religion, family). We treat these two approaches as complementary theoretical frameworks rather than mutually exclusive explanations. The social and institutional environment and individual choices interact and complement each other in creating the space in which age-related norms emerge and influence the choices that individuals make about their key life events.
Research on cross-country comparative analysis of timing of life offers an extremely rich source of literature (Aassve, 2013;Billari et al., 2011;Billari-Liefbroer, 2007;Buergin et al., 2017;Loretto et al., 2000;Radl, 2012;Spéder-Murinkó, 2014;Van Bavel-Nitsche, 2013). Our study is more of a descriptive than explanatory nature; it adds to the existing wealth of empirical knowledge by using the most recent data from the 9th wave of ESS (2018/19) and offers insights to the changes through over a decade in expectations and norms related to the stages and key events of life by comparing two rounds of ESS data. The study and its analysis puts a special focus on describing and understanding changes of perceptions about the timing of life in the Central East European region.
The paper is structured as followed: following the introduction of the context and the methods and data we used, we present the findings of the analysis of perceptions associated with the life cycle comparing two large European regions and the two-time momentums (2006 and 2019). In the second part of the analysis we focus on two key life events and their perceptions in Europe: the birth of the first child and retirement. We conclude by summarizing our findings and raising the policy relevance to our results.

Methods and data
In our analysis we use 2006 and 2018 rotating modules of the European Social Survey data about the Timing of Life in 19 European countries 2 which participated in both survey rounds (R3 and R9). The list of countries offers an equal coverage of east, west, north, and south of the continent, and includes a broad variety of societal developmental traits that characterize the diversity of European countries. The total size of the sample analysed was 68,398. We applied weighting offered by the ESS methodology: design weight allowing unbiased estimations. It is important to note that each respondent was randomly assigned to answer only questions about either men or women (split-voting method). Therefore, the sample size of the responses to the questions was halved. A total of 13 variables mapping perceptions about life-cycle and timing of key life events were used in our study.

Perceptions and norms related to life-course
The first part of our analysis offers a triangulation of the ages attributed to the beginning of three life cycles (adulthood, middle age, and old age), demographic characteristics (life expectancy) and public policy framework related to the turning point of the life cycle: retirement age. The perception of the life cycle is influenced both by individual attitudes and by the structural and social environment in which the individual lives, such as the demographic composition of society, public policies, and legal frameworks affecting the lives of different age groups. (Settersten, 2018) Figure 1. attempts to summarize this pattern by combining perceptions about life cycle boundaries (circles representing averages for the beginning of adulthood, middle age, and old age), statistics on life expectancy, and the legal retirement age for men and women in 19 European counties (indicated by diamond shaped symbols).
The perception about the age associated with boundaries of the life cycle differs for women and men to some extent in most countries. However, there is a much greater difference across countries in how on average age boundaries of the life cycle are viewed. In particular, there are differences regarding the beginning of middle age and old age. Perceptions on when middle age begins vary by up to a decade with the lowest in Spain and Italy (32 and 35 years) and the highest in Norway, Ireland and Cyprus (47). There is similarly-wide gap in the belief of when old-age starts, where country averages are fluctuating between 64 and 73 years). Countries of the Central East European region seem to have a shared understanding about the beginning of the life cycle: they perceive the beginning of middle age at 41 to 43 on average. The perception about when old age starts is relatively homogeneous in the region too: on average, people think that old age begins relatively early, at the age of 64 to 67 years on average compared to other countries in Europe, where over 70 years is the typical perception. However, as a 'rule of thumb', in countries with an average life expectancy of under 80 years -as is typical for men living in the CEE region -people tend to think that old age starts earlier (64 -65 years) than in countries with a higher life expectancy, where people usually tend to associate the beginning of old age with a later age. We assume that the reason for the marked difference between populations in Central East Europe and Western Europe concerning their perceptions of old age lies in differences in intra-generational division of work rather than in differences in demographic features or dissimilarities in pension systems (Gál & Medgyesi 2017). Broadly speaking, in poorer and more traditional post-communist societies of the CEE region, women entering old age very often get involved in supporting their children's families, care for their grandchildren and help with household duties of their children. In the northern countries of Europe people entering old age are more likely to begin a new life stage with engaging in hobbies, travelling, and spending their time with various activities for their own pleasure.
Looking at the changes between 2006 and 2018/19, we see a relative stability across all countries in their perceptions of age associated with the life cycle: on average the age associated with the beginning of adulthood increased by 0.4 years; the perception about middle age has not changed, while there is a more explicit increase (1.4 years) in what Europeans, on average, think about the beginning of old age. However, there is a clear regional difference in the extent of this increase: compared to Western Europe, in countries of the CEE region, especially in Hungary and in Poland, all age boundaries have shifted upwards to a much greater extent than in other countries in Europe. Since the perception of age is strongly dependent on the actual age of the respondent, it is also worth examining the changes in age-related norms by looking at them within the same age groups (cohorts). Figure 2 shows perceptions of when old age begins and changes in these perceptions by cohort, comparing data for CEE and Western Europe in 2006 and 2018/19 for men and women, respectively 3 . Across all age groups, people in Western Europe generally place the start of old age at a later age than in Eastern Europe, though the East-West gap appears to be closing among younger cohorts. While in the older age groups there is a difference of about 5 years, the difference is significantly less in younger cohorts (2.5 years). It is also evident that the gap has continued to close over the past 12 years, with the averages for young people's perceptions in Eastern Europe in 2018 (purple line) approaching those of young people in Western Europe in 2006 (red line). In short, we see a clear trend of convergence of people's perception concerning life cycles between the countries of the CEE region and the rest of Europe. Although the age of the respondent is a very strong predictor of the perception of the beginning of old age, in all age groups for both men and women it still increased between 2006 and 2018. This may be interpreted as a significant change in the norms associated with the perception of when people become old. 3 We grouped countries based on the fact that all age-related analyses show a consistent difference between postcommunist Central Eastern European countries and the rest of Europe. The CCE group includes the following countries: Bulgaria, Estonia, Hungary, Slovenia, Poland and Slovakia; The remaining group of countries, referred to as Western Europe includes Austria, Belgium, Switzerland, Cyprus, Germany, Finland, France, Spain, the UK, Ireland, The Netherlands, Norway, Portugal and Sweden.

Perceptions and norms related to key life events
In the second part of our study, we address the perceptions about key events of life. Major life events such as graduation from school, starting a household, getting married, parenthood, and retirement also mark life stages. Unlike phases of the life cycle (young, middle and old ages), these are more tangible and therefore more clearly associated with a particular age. Moreover, there is a chance to compare people's perceptions about these events, the social norms (expressed in policies related to some of the key events, such as schooling and retirement) and the actual statistical data on when these events occur. This allows for comparing the relationship between norms, perceptions, and reality at the societal level. In what follows, we focus on two important life events: the birth of the first child, and retirement. We examine the perception about what the ideal age is for these events for both men and women, when these events are considered too early and too late, and how this relates to reality, i.e., when these life events occur according to national statistical data.

Becoming a parent
The birth of the first child is one of the most important events in life. Expectations of the immediate and wider social environment are very strong, and often becoming a stressful external pressure on the individual. The mother's age at the birth of her first child is also a key indicator of socio-demographic processes, as the timing of the first child's birth has a great impact on the total number of children a mother will have and thus influencing the reproduction rates at a societal level. According to data from Eurostat, the average age of women at childbearing is 29 years in the European Union, but it varies considerably between countries, ranging from 26 years in Bulgaria to 31 years in Italy. The long-term trend is apparent: women are becoming parents at an increasingly later age and the shift is especially prominent in countries of the CEE region. Perceptions and norms associated with the age of parenthood influence greatly not only the age when young people decide to become parents, but through their fertility behaviour, they impact demographic processes at the level of society.
Looking at the ESS data we find that norms and expectations regarding the age at which the first child is ideally born has a visible gender difference: Expectations about when is the earliest (19 -22) as well as the ideal (24 -27) age for women to become parents is 2 years less compared to men. The differences in the perception of the upper age limit (the age at which someone is too old to have children) are, unsurprisingly, much greater. For women it varies between 40 (in Slovakia) and 45 (in Austria), while for men, the averages are 4 to 8 years higher. The median age of 50 seems to be the symbolic limit in terms of perceptions for men having their first child.
Perceptions about the ideal age of becoming a parent is surprisingly homogeneous across the continent: people consider the age between 25 to 27 as ideal for women to become parents irrespective of the country they live in. However, the ideal age is rather distant from realities in contemporary societies: the actual age when women give birth to their first child is 2 to 4 years later than what the society perceives as ideal. The gap in societies of the CEE region is typically 2 years, while in Norway, Finland, Austria it is 4 years.

Figure 3
Country clusters in Europe based on women's considered ideal, too early, too late and average real age at the birth of their first child, 2018 -2019 Figure 3 presents the results of a (hierarchical) cluster analysis of these four data points (perception of the minimum, ideal and maximum age of becoming a mother and official statistics about the same life event). The left side of the figure shows the dendrogram of how countries were grouped, while the right side of the figure shows the values of the first two principal components used for clustering. In simple terms, the horizontal axis condenses the perception about the minimum, the ideal ages of becoming a mother and the statistical data on the same life event, while the vertical axis shows the inverse of the perception about the maximum age for giving birth to the first child. Thus, the further to the right a country is positioned on the graph, the perceptions and reality of women becoming parents are linked to a later age, and the higher it is positioned the earlier the maximum age of giving birth is perceived.
Among the four country groups, Slovakia occupies a unique position because the upper limit (maximum) age for becoming a parent is uniquely low compared to other European countries. Countries of the CEE region, except for Slovakia, are grouped into a separate cluster since the actual as well as the perceived ages associated with the ideal, minimum, and maximum ages for childbearing are low compared to the other countries. Countries with a slightly above average ideal age (27 -28 years) for childbearing (the Netherlands, Spain, Ireland, Cyprus, and Switzerland) have also been grouped into a separate cluster. Indeed, in these countries the average age of the mothers at giving birth to their first children is higher (30 -31 years). Finally, the country cluster making up most Western European societies present values in years close to the averages for all the four studied aspects (except for Austria, where the upper threshold for the birth of the first child is set at 45, a remarkably late age on average).
It is also worth looking at the data by cohorts since norms associated with childbearing age may significantly differ by the age of the respondent. Figure 4 shows the average age perceived as ideal for women to become mothers in a regional and time-wise comparison. In the chart representing 19 countries, the solid line shows data from 2018 -2019 (ESS R9) and the dashed  Based on Figure 4 we see four distinctive effects.
• Cohort effect: the younger someone is, the higher s/he perceives the ideal age for childbearing (and the upper limit for becoming a parent, which is not shown in the figure). Based on data from ESS R9 (2018 -2019), older generations (born between 1930 and 1940) consider the ideal age for women to have their first child two years earlier than younger cohorts.
• Time effect: the perception about the ideal age of women's childbearing age has significantly increased -by about 1 year -between 2006 and 2018 -2019 in both Western and Eastern Europe in all cohorts, and for both sexes.
• Gender effect: regardless of age, region and the time of the survey, men consistently think that the ideal age for women to have their first child is two years earlier than what women think. • Effect of region: the effect of region is weaker but also consistent. Women are considered to have their first child at a younger age in Eastern Europe than in Western Europe. However, an important and significant trend of convergence between East and West is also uncovered by the data. When looking at older generations we see a clear difference in perceptions of people in Eastern and the Western Europe, however, this difference fades away as we approach to the youngest cohorts.
These findings have important socio-political and demographic implications. The changing norms about the age associated with childbearing affects people's actual fertility behaviour and tends to lead not only to later childbearing of women, but also to fewer children due to natural biological constraints. Taking all these factors into account, it is likely that changing perceptions of the age of fertility lead to the further ageing of societies.

Expectations and norms regarding retirement age
Retirement is another key life event for which data also shows significant cross country differences and changes over time. Similarly to childbearing, perceptions and norms related to ageing and retirement have a direct impact on society and economy, where policy frameworks and norms interact dynamically. Figure 5 presents result of a (hierarchical) cluster analysis in which countries were grouped according to their perceptions of the beginning of old age, and the ideal, earliest, and latest age for retirement. The left side of the figure illustrates the clustering process, and the right side shows the relative positioning of countries according to the first two principal components.

Figure 5
Country clusters in Europe based on the beginning of old age and norms on retirement, 2018 -2019 Source: Own calculation based on ESS R9 (2018/19) data. The cluster analysis has grouped European countries into four clusters: Scandinavian countries form the first group; the United Kingdom and the continental countries of Northern Europe (and Cyprus) construct the second; Germany and the countries of the CEE region are clustered into the third group; and finally, the Mediterranean countries and Belgium belong to the fourth group.
What are the characteristics of the specific country groups and why were they matched together?
• The average age of when the life cycle reaches old age is quite similar in the Nordic countries and in Western Europe, but norms associated with the age of retirement are significantly higher in the Nordic countries. This implies both the age of the minimum (55 years), the ideal (64.5 years), and the maximum (69 years). Remarkably, there is virtually no difference in the ideal retirement age for women and men in these countries.
• The average age of when the life cycle reaches old age is slightly higher in the UK and the continental countries of Northern Europe (71 years) than the European averages, but in contrast to the first country group, the norms associated with retirement age are lower, while there are clear gender differences in these perceptions too (the minimum is considered to be 52 years for women and 54 years for men, while the ideal age is considered 62 years for women and 64 years for men; the maximum age for retirement is considered to be 62 for women and 66 years for men).
• The third group is formed by Mediterranean countries and Belgium. The perception about when old-age begins is relatively high (70 years), but people consider that labour market activity could end as early as 49 for women and 53 for men in this country group. Another interesting characteristic of this group is that the perception about the ideal and the maximum age for retirement are alike with 60 years for women and 62 years for men. There is a slight gender difference (2 years on average) in the perceptions associated with old age and pension in this country group.
• Countries in the CEE region together with Germany (which obviously includes the former GDR) differ most significantly from other European countries in their perception about old-age. In these countries, people consider that old age begins by about 4 to 5 years earlier (65 and 66, respectively) than what people in other European countries think. Age norms associated with retirement are also low and are similar to those of the Mediterranean countries. Another important feature of this group is the marked difference in perceptions between women and men: on average, people believe that women should retire 3 -4 years earlier than men.
These perceptions are also significant for social policy making, as they can influence or limit pension policies' room for manoeuvre. A pension reform that sets the retirement age significantly higher than what the population considers as the maximum age for labour market activity presents serious political risks. Yet, in European societies which are characterized by the demographic process of increasing life expectancy, declining birth rates, and an ageing population, raising the retirement age is the most important social policy measure to control the country's dependency rate. According to OECD forecasts, almost all European countries will be forced to increase the retirement age significantly in the coming decades. (OECD, 2011) Therefore, people's ideas and societal norms associated to retirement age and their changes across time is indeed a significant factor for policy makers.
Similarly, to age norms associated with parenthood, perceptions about the ideal age for retirement could be analysed along cohorts and regions. Figure 6 shows the perception of the ideal retirement age for women and men by cohort, for Eastern and Western Europe respectively, and its change between 2006 and 2018 -2019.

Figure 6
Changes in the perceptions of the ideal retirement age by gender in Eastern and Western European countries, 2006-2019 Source: Own calculation based on ESS R3 (2006) andESS R9 (2018/19), aggregated data from 20 countries.
Several noteworthy effects can be distinguished in terms of age norms regarding retirement: • The perception about the ideal age of retirement has increased in both broadly defined regions and for both genders over the 12 years studied. However, the increase is somewhat greater in the CEE region (2 years for men and 3 years for women compared to Western Europe (1.5 -2 years).
• There is no significant cohort effect in age norms associated with retirement. There is only a minimal difference between what older and younger people consider to be the ideal age for retirement.
• There is a marked gender difference: perceptions about when is it ideal to retire for women is two years lower than men. The gender gap has not changed over time, but the perception of ideal age of retirement has increased by about 2 years for both men and women.
• There is a regional effect: people in Eastern Europe consider the ideal age for retirement to be earlier than in Western Europe, but the gap has narrowed in the past 12 years. In other words, the age considered to be ideal to retire has increased more in Eastern Europe than in other parts of the continent.

Conclusions
Based on the ESS data allowing the comparison of attitudes and norms related to age and timing of life in 2006 and over a decade later in 2018/19, it enabled us to pinpoint some important changes. We saw that perceptions about ages associated to the life cycle and major life events have shifted: in all respects people tend to associate a later age to these. Compared to 2006 in 2018/19 people in Europe generally think that events such as becoming an adult, becoming a parent, becoming old is delayed by several years. One of the consequences of these processes is the emergence of a new stage of life referred to as 'young adulthood' (post-adolescence), wedged between traditional childhood and adulthood. The perception about the beginning of old age has increased in all European countries, however this process is particularly pronounced in the countries of the CEE region. The most important trend, however, in our understanding is the slow but clear process of convergence of societies in the CEE region and Western Europe. While perceptions about the life cycle and timing of life are also changing in Western Europe, the dynamics are more pronounced in Eastern Europe, mainly due to the 'almost Western European norms' of younger cohorts.
We found a profound trend of changing norms associated to childbearing, too. These have pronouncedly increased in most societies for both men and women, and in all cohorts. However, this trend is also fuelled primarily by young people's changing perceptions. This is a welldocumented trend in developed societies. However, what we may add to the broader scholarly knowledge is the explicit convergence of countries of the CEE region and other European countries. Potential reasons behind the changing (increase of) age norms associated to adulthood and childbearing include the disintegration of the traditional lifestyles and increasing awareness about gender roles and related inequalities. An important finding in this respect is the closing of the gap between broad regions of Europe in terms of young people's attitudes and norms associated to life cycles and timing of life. Young people across Europe think quite similarly about age norms associated with childbearing, whereas this is no longer the case for middle-aged people, and certainly not for the elderly.
Norms and expectations with relation to the timing of retirement are also converging, but difference remains significant in this respect. At the two extremes are the Nordic countries and the post-communist societies of the CEE region. What we see from the analysis of the ESS Timing of Life module is that people's attitudes and norms related to the human life cycle and timing of major life events adopt to the realities of aging societies, while they also support the same process: with young people associating later age to the ideal and the maximum age of childbearing, they also tend to delay the decision about becoming parents. This trend is further supported by the increased diversity of family and partnership models. As a result, earlier stigma attached to having a child after the age of 40 has been fading away. However, the increase of the age-norms associated with retirement age as well as increasing life expectancy shifts labour market activity towards later ages, and thus offers a chance of preserving a sustainable dependency ratio.