Abstract It is generally assumed that life courses in European societies have become less orderly and more destandardized in recent decades. Focusing on the family sphere, the article examines to what degree patterns of destandardization are stratified by educational attainment across seven European countries. Using data from the Generations and Gender Survey (GGS) and the National Educational Panel Survey (NEPS) (n = 70,228 respondents), the article adds to the methodological discussion of destandardization by implementing both abstract analyses of life course dissimilarity, which focus on the ‘timing’ of events; and specific analyses of common episode orders, which relate to the ‘order’ of events. While European countries differ considerably with respect to dominant life course patterns in early adulthood, a consistent finding is that destandardization is more pronounced among individuals with lower than with higher levels of education. Introduction It is widely understood that since the 1970s, life courses across Europe have become less institutionalized and increasingly individualized (Kohli, 1986; Beck, 1992; McMillan, 2005). In the post-war period, economic growth and mass prosperity fostered orderly and highly predictable life courses based on continuous employment (particularly among men) and nuclear family patterns accompanied by early marriage, low divorce rates, and high levels of fertility. However, in the mid-1960s and 1970s major shifts in demographic behaviour occurred, as living arrangements proved to be more heterogeneous and family life transitions became institutionally less regulated (Van de Kaa, 1987; Lesthaeghe, 2010). In the economic sphere, employment patterns turned out to be less stable (Myles, 1992). As a result of these developments, cohort-specific life courses became less orderly and more heterogeneous. The shifts in these patterns have been particularly pronounced among young adults, as events and status changes in family and work life are more prevalent during this life stage than at older ages (Rindfuss, Swicegood and Rosenfeld, 1987; Furstenberg, Rumbaut and Settersten, 2005). In quantitative life course research, the trends described above are often referred to as a process of destandardization. According to Brückner and Mayer (2005: 32f.), life course destandardization takes place when ‘life states, events and their sequences … become experiences which either characterize an increasingly smaller part of a population or occur at more dispersed ages and with more dispersed durations’. Providing a comprehensive overview of country- and cohort-specific patterns of change in different life domains can be difficult, especially since previous research has suggested that a juxtaposition of highly standardized ‘post-war’ life courses and individualized ‘post-fordistic’ life courses is simplistic (Mayer, 2005). There is evidence that processes of destandardization are less widespread than was initially assumed, are more pronounced in union and family histories than in education and work (Kohli, 2007), and are modified by nation-specific institutional contexts. Conceptual dichotomies and straightforward adaptions of (post-)modernization theory have thus proven inadequate to explain the empirical life course dynamics over the past five decades in western European countries. These theoretical models are even less applicable to developments in central and eastern Europe, where patterns of change before and after the collapse of communism vary considerably between individual countries (Kogan, Clemens and Gebel, 2011). Given the complex nature of life course patterns, it is clear that in addition to engaging in careful empirical description of and applying innovative methodological tools to these patterns, further theoretical refinement is needed. As we observed above, the narrative of a shift towards less orderly and predictable life courses is indicative of either an optimistic stance (i.e. the assumption that institutional control over individual lives is decreasing) or a pessimistic stance (i.e. the assumption that life course risks and instability are increasing, and that individual life paths are disintegrating). Against this background, we believe it is crucial to examine empirically the question of which ‘variant’ of destandardization best fits the experiences of subsequent birth cohorts in European countries. In our research, we analyse cases from different regions of Europe—namely, northern and continental Europe and Mediterranean and eastern Europe—to validate our assumptions across a diverse set of societal contexts. While it is beyond the scope of the current article to systematically analyse national specificities, we will explore the question of whether there is a common pattern of destandardization across European regions and institutional contexts. In this article, we focus on life course destandardization in union and family formation, as across European societies, the family is the institutional sphere in which the greatest life course changes have been occurring over the past 50 years (Kohli, 2007). We build on previous studies that addressed this issue conceptually (Brückner and Mayer, 2005), as well as empirically through the application of sequence analysis (SA) (most notably, Elzinga and Liefbroer, 2007). While these studies pioneered the comparative analysis of complex life courses, they also pointed to certain theoretical and analytical problems that call for closer examination. First, social disparities in destandardization have rarely been analyzed even though the increasing heterogeneity of life courses is closely related to issues of social stratification. Against this background, we will examine the extent to which destandardization patterns differ between educational groups. Previous research on the impact of education on demographic behaviour followed two opposing perspectives. On the one hand, the Second Demographic Transition (SDT) theory (Lesthaeghe, 2010) assumes that individuals with higher levels of education, and particularly highly educated women, have been pioneering demographic changes, most notably, rising levels of cohabitation and postponement of fertility. On the other hand, theoretical accounts that focused on increasing economic strains, the decline in the male breadwinner model, and family instability identified less educated groups as the drivers of non-traditional demographic behaviour (Friedman, Hechter and Kanazawa, 1994; McLanahan, 2004). In addition, a number of authors have argued that increases in non-marital fertility and single motherhood are related to the deteriorating economic status of less educated men and the decreasing supply of marriageable men, and, thus, that the dynamics of life course trends differ by gender (Oppenheimer, 1988, 2000). Empirical research on the educational gradient of cohabitation and single motherhood in European countries tends to support this hypothesis (Perelli-Harris et al., 2010). We therefore assume that the destandardization of family trajectories is an aggregated outcome of disparate life course changes at the upper and the lower ends of the social strata. We anticipate that disentangling these sub-processes will allow us to better identify the socio-structural bases of the processes of family life course destandardization. Secondly, despite recent innovations, the conceptual bridge between life course standardization and SA needs to be strengthened. A growing body of research has been examining the link between dissimilarity measures provided by methods of SA and the structural dimensions of the life course (Robette and Bry, 2012; Halpin, 2014; Elzinga and Studer, 2015). Studer and Ritschard (2016) suggested using three measures of life course dissimilarity (focusing on ‘timing’, ‘duration’, and ‘order’), and Fasang (2012) applied two dissimilarity measures (focusing on ‘timing’ and ‘order’). For the analysis of destandardization, several authors have proposed comparing average dissimilarities of life courses between cohorts (Elzinga and Liefbroer, 2007; Robette, 2010; Aisenbrey and Fasang, 2010; Zimmermann, 2013). We build on these approaches by distinguishing between the two basic life course dimensions of ‘timing’ and ‘order’, and by using ‘average dissimilarities’ as abstract summary measures of levels of life course dissimilarity. To identify the predominant life course trajectories in different social strata, we complement our analysis of average dissimilarities with an analysis of ‘common episode orders’. Theoretical Considerations Destandardization as a Concept of Life Course Change Previous research has shown that over the past half century, major life course shifts have occurred in European countries. Several studies have linked these changes to secular processes of deinstitutionalization and individualization (Buchmann, 1989; McMillan, 2005; Scherger, 2007; Mayer, 2009). In family demography, similar discussions emerged around the SDT theory, which used the concept of an intergenerational value shift (Inglehart, 1997) to explain changes in union formation and fertility (Lesthaeghe, 2010). In a similar vein, it has been argued that changing gender norms led to the dissolution of traditional family arrangements, and of the male breadwinner model in particular (Esping-Andersen, 1999). Since the 1990s, research on growing labour market risks and economic insecurity has also asserted that previously orderly life courses have been disrupted by globalization and more flexible labour markets (Blossfeld et al., 2005); the retrenchment of welfare states (Mayer, 2005); and, more recently, the economic recession that hit Europe starting in 2008. However, there has been no consensus among scholars regarding the long-term changes in life course dynamics. Beyond the common observation that cohort-specific patterns have become less standardized, there are at least three alternative narratives that seek to explain long-term life course dynamics. First, the increase in destandardization can be seen as marking the end of an outlying period of exceptionally high standardization that emerged during the European post-war period. According to this view, the process of destandardization will eventually end after the ‘normal’ low levels of standardization observed in previous periods have been reached. This argument is especially persuasive when applied to the family sphere, given that the ‘golden age of marriage’ in the 1950s and the 1960s has been identified as an historical exception (Festy, 1980). The second perspective posits that destandardization does not represent a new ‘stage of development’, but rather a transitional period that will ultimately be followed by a process of restandardization (Huinink, 2013). According to this perspective, destandardization merely represents a secondary aspect of the transition from one dominant life course regime to another (Lesnard et al., 2016). In line with this assumption, Elzinga and Liefbroer (2007) found two new emerging standards of family formation in Europe: ‘traditional late motherhood’, which is characterized by the postponement of traditional family formation, but not by prior cohabitation; and ‘modern late motherhood’, which is defined by both premarital cohabitation and the postponement of family formation. Billari and Liefbroer (2010) argued that a more protracted and complex life course pattern is evolving that is mainly caused by the disconnection of life transitions. According to Lesnard et al. (2016), early independence from the parental home and delayed family formation have become new standard behavioural patterns. Since empirical research does not unequivocally support either of these two perspectives, we will elaborate a third perspective in which we argue that the dynamics of life course change differ by life course dimension (domain) and social strata. Previous studies have suggested that diagnoses of (de)standardization are sensitive to the dimensions and age spans considered (Scherger, 2007; Konietzka, 2010; Robette, 2010; Zimmermann, 2013), e.g. that the increasing standardization of family formation may be accompanied by the destandardization of retirement processes. Such findings call into question the idea that encompassing long-term trends of destandardization exist; as well as the assumption that life courses tend to follow the subsequent stages of a modern (standardized, stable, continuous) and a postmodern (destandardized, unstable, discontinuous) life course regime (Mayer, 2005). In the next section, we will elaborate in more detail how life course changes are shaped by social stratification, while paying particular attention to education and its effects on life course (de)standardization. Educational Disparities in Destandardization Education is commonly seen as a major individual asset and a key factor in the social stratification processes that affect working life, family formation, and health (Mirowsky and Ross, 2003; Kravdal and Rindfuss, 2008; Kreyenfeld and Konietzka, 2008; Kreidl, Ganzeboom and Treiman, 2014). We therefore consider education as a comprehensive proxy for the cultural and economic resources that have sustained influence on all life spheres, and, thus, on each individual’s position within society over his or her life course. Since it is beyond the scope of this article to take into account specific institutional features of educational systems in Europe, we use a simple and straightforward approach to categorizing educational levels that distinguishes between ‘lower’, ‘medium’, and ‘higher’ education. Individuals who have a basic secondary education degree or less are classified as lower educated. Individuals who have attained some level of upper secondary education are defined as medium educated. Finally, individuals who have a post-secondary or tertiary education degree are categorized as higher educated. Our main focus is on the question in which educational category the de(or re)standardization processes of family life courses are most advanced. Table 1 identifies four analytical spheres that we will elaborate in the following (referring to processes (i)–(iv)). For each field, two competing hypotheses with respect to the relationship between education and de(re)standardization are derived. In other words, we will examine whether de(re)standardization is more pronounced among the lower educated or the higher educated (see also Table 3).1 Table 1. Method of analysis, analytical focus, and main research areas Analytical approach Method of analysis and life course aspect focused on Abstract changes Substantial changes Average dissimilarity in life courses (Hamming distance); focus on the timing of events Prevalence of the most common episode order (proportion of respondents); focus on the order of events or episodes Direction of change Destandardization (i) Increase in average dissimilarity (Hypotheses 1a and 1b) (ii) Decreasing prevalence of the ‘traditional’ episode order (Hypotheses 2a and 2b) Restandardization (iii) Decrease in average dissimilarity (Hypotheses 3a and 3b) (iv) Increasing prevalence of the new episode order (Hypotheses 4a and 4b) Analytical approach Method of analysis and life course aspect focused on Abstract changes Substantial changes Average dissimilarity in life courses (Hamming distance); focus on the timing of events Prevalence of the most common episode order (proportion of respondents); focus on the order of events or episodes Direction of change Destandardization (i) Increase in average dissimilarity (Hypotheses 1a and 1b) (ii) Decreasing prevalence of the ‘traditional’ episode order (Hypotheses 2a and 2b) Restandardization (iii) Decrease in average dissimilarity (Hypotheses 3a and 3b) (iv) Increasing prevalence of the new episode order (Hypotheses 4a and 4b) (i) Education and destandardization: According to theories of value change, higher educated individuals are trendsetters and pioneers of alternative lifestyles (Inglehart, 1990). Since these individuals generally have more life options than their less educated counterparts, we could assume that the higher educated tend to have more diverse life course patterns than the lower educated. Accordingly, the degree of destandardization should be larger among the higher than the lower educated (Hypothesis 1a). On the other hand, we could assume that growing socio-economic vulnerability due to declining job opportunities, decreasing social resources, and increasing imbalances in the marriage market lead to greater family instability among the lower educated (Oppenheimer, 2000; Perelli-Harris et al., 2010; Kalmijn, 2011). Hence, when examining the life course patterns of lower educated women and men, we can expect to observe a relatively large degree of destandardization in family sequences (Hypothesis 1b). (ii) Education and ‘traditional family formation’: In line with Hypothesis 1a, we can expect to find that cohort-level destandardization is the outcome of the trend towards rejecting traditional patterns of family formation, which was started by the higher educated during the 1970s and 1980s. Therefore, we can assume that the substitution of the traditional sequence of ‘marriage and family formation’ with a pattern that includes non-marital cohabitation and postponed life transitions (Lesthaeghe, 2010) spread more quickly among the higher educated (Hypothesis 2a). When looking at the lower educated, in contrast, we can expect to find that their access to the resources needed to establish ‘orderly’ sequences of family life has been declining. In particular, the rise of single motherhood and early childbearing has been related to a lack of alternative roles available to lower educated women (Friedman, Hechter and Kanazawa, 1994). We can thus assume that traditional patterns of family formation have been decreasing more among the lower than the higher educated (Hypothesis 2b). (iii) Education and restandardization: Behavioural patterns that were first adopted by specific sub-populations may initially result in increasing dissimilarities in life courses. However, these patterns may eventually turn into mainstream life course stages, thereby redefining behavioural standards (Huinink, 2013). For example, non-marital cohabitation started as a deviant behaviour in the 1970s, before being transformed into a standard stage of union dynamics and replacing marriage as the dominant union type in early adulthood, particularly in north-western Europe (Hoem and Hoem, 1988; Kiernan, 2001; Hiekel, Liefbroer and Poortman, 2014). It thus appears that processes of standardization may supersede processes of destandardization, or may follow them historically. With respect to education, it is possible to argue that restandardization is more prevalent among the higher educated because they adapt to new behavioural patterns in society more quickly, command more resources, and are less restricted by adverse living conditions when considering adopting a new ‘desired’ family life course model (Hypothesis 3a). However, it could also be argued that the sharp increase in non-marital fertility among the lower educated may foster single motherhood as a new modal family trajectory, and that restandardization trends may therefore become more prevalent among the lower educated (Hypothesis 3b). (iv) Education and the emergence of new standards in family formation: New dominant life course trajectories may emerge if new types of behaviour diffuse into mainstream behaviour—as was the case with cohabitation in northern and western Europe. In line with Hypotheses 3a and 3b, we therefore expect to find that the higher educated have been acting as forerunners of cultural change by increasingly adopting a family life course pattern that involves a period of pre-marital cohabitation (Hypothesis 4a). The counterargument is that the lower educated have increasingly adopted non-traditional paths to family formation, in particular single motherhood, as a new modal life course pattern (Hypothesis 4b). Methods and Data In this article, we compare cohort-specific patterns of family formation between ages 15 and 35 years. By distinguishing life course states for each quarter of a year, the study is based on sequences of 80 states for each individual. For the sake of simplicity and comparability of cross-national data, we consider three different dimensions: cohabitation (single vs. cohabiting with partner), formal marriage (not married vs. married), and children (childless vs. at least one child in the household). Using these dimensions, we are able to differentiate eight family statuses. Although this classification system is only capable of mapping basic dimensions of living arrangements, it enables us to assess the extent to which central aspects of family life course changes line up with the theoretical accounts discussed above. Including more dimensions (e.g. living-apart-together relationships, stepfamilies, same-sex couples) would require data that are not available, and would in the case of stepfamilies restrict our analyses to older birth cohorts. Within the scope of this study, we cannot give an in-depth account of country-specific institutional contexts or of cross-national differences in life course patterns. We also do not aim to emulate models that rely on welfare regime typologies (Esping-Andersen, 1999), not least because it is doubtful that such typologies provide a sound theoretical basis for comparative life course research. Still, we expect our analysis to show that family life courses and the impact of education on life courses vary considerably across country-specific institutional settings and path dependencies (Mayer, 2005, 2009). Against this background, our strategy is to include a diverse set of countries to put our assumptions regarding the impact of education on patterns of (de)standardization (as expressed in Hypotheses 1–4) to various tests. The seven countries under study include a northern European country (Norway) and two major western European continental countries (France and Germany). For all three countries, it has been shown that life courses have become significantly more individualized and destandardized during the past decades. In contrast, Italy has been found to follow the more traditional ‘familistic’ Mediterranean model (Ferrera, 1996). Meanwhile, the former socialist countries of Estonia, Hungary, and Russia have been shown to represent the range of transformation pathways that central and eastern European countries followed after the breakdown of communism (Kogan, Clemens and Gebel, 2011). We use nationally representative data from the first wave of the Generations and Gender Survey (GGS) for respondents born between 1935 and 1969. For Germany, due to limitations of data quality of the GGS (Sauer, Ruckdeschel and Naderi, 2012; Kreyenfeld, Hornung and Kubisch, 2013), we use data from the ‘starting cohort six’ of the National Educational Panel Survey (NEPS, Blossfeld, Roßbach, and von Maurice et al., 2011), which provides comparable information for respondents born between 1940 and 1974. Because there were too few cases, we excluded eastern Germans and respondents from Berlin (it was not possible to distinguish between respondents from East and West Berlin). In sum, our analyses rely on 70,228 respondents, which are grouped into birth cohorts covering 5 (youngest cohort) or 10 (all other cohorts) years of birth. The number of respondents per group varies considerably. Some groups had to be excluded from the analysis because of low case numbers (see Table 2, the sizes of the excluded cohorts are in brackets). Table 2. Case numbers by country, cohort, gender, and education Country Cohort Lower educated (ISCED 0-2) ISCED 3 Higher educated (ISCED 4–6) Male Female Male Female Male Female France 1935–1944 264 429 223 211 107 91 1945–1954 284 440 366 396 175 212 1955–1964 210 303 381 408 209 330 1965–1969 85 107 245 273 130 189 Germany 1935–1944 (4) (16) (45) (31) (48) (20) 1945–1954 45 140 488 514 496 311 1955–1964 75 187 602 896 767 551 1965–1974 28 54 297 438 352 250 Italy 1935–1944 388 691 122 108 26 41 1945–1954 515 689 280 271 113 97 1955–1964 519 476 444 495 136 152 1965–1969 237 208 215 320 68 104 Norway 1935–1944 217 258 475 454 265 205 1945–1954 205 220 705 650 406 463 1955–1964 237 304 692 585 382 532 1965–1969 116 105 383 312 280 397 Estonia 1935–1944 153 376 104 159 194 522 1945–1954 71 132 202 289 353 746 1955–1964 37 36 302 295 454 991 1965–1974 (13) (17) 157 126 206 362 Hungary 1935–1944 237 596 270 193 320 434 1945–1954 185 443 472 268 467 724 1955–1964 145 282 466 265 403 714 1965–1974 68 75 209 143 226 339 Russia 1935–1944 176 277 71 160 168 362 1945–1954 135 124 133 272 215 488 1955–1964 72 60 216 278 254 596 1965–1974 (24) (22) 139 134 95 257 Country Cohort Lower educated (ISCED 0-2) ISCED 3 Higher educated (ISCED 4–6) Male Female Male Female Male Female France 1935–1944 264 429 223 211 107 91 1945–1954 284 440 366 396 175 212 1955–1964 210 303 381 408 209 330 1965–1969 85 107 245 273 130 189 Germany 1935–1944 (4) (16) (45) (31) (48) (20) 1945–1954 45 140 488 514 496 311 1955–1964 75 187 602 896 767 551 1965–1974 28 54 297 438 352 250 Italy 1935–1944 388 691 122 108 26 41 1945–1954 515 689 280 271 113 97 1955–1964 519 476 444 495 136 152 1965–1969 237 208 215 320 68 104 Norway 1935–1944 217 258 475 454 265 205 1945–1954 205 220 705 650 406 463 1955–1964 237 304 692 585 382 532 1965–1969 116 105 383 312 280 397 Estonia 1935–1944 153 376 104 159 194 522 1945–1954 71 132 202 289 353 746 1955–1964 37 36 302 295 454 991 1965–1974 (13) (17) 157 126 206 362 Hungary 1935–1944 237 596 270 193 320 434 1945–1954 185 443 472 268 467 724 1955–1964 145 282 466 265 403 714 1965–1974 68 75 209 143 226 339 Russia 1935–1944 176 277 71 160 168 362 1945–1954 135 124 133 272 215 488 1955–1964 72 60 216 278 254 596 1965–1974 (24) (22) 139 134 95 257 Note: Cells with fewer than 25 members were excluded from the analysis (number of respondents in brackets). In Germany, the oldest birth cohort was excluded from the analysis. Source: GGS, NEPS, own calculations We apply the ISCED 1997 classification in creating categories of the respondents’ levels of education. We distinguish three educational levels. The Level 0–2 category encompasses respondents who completed only lower secondary education or the first stage of secondary education. This stage generally equates to Grades 7–9, and in some cases Grade 10. The Level 3 A–C category includes individuals who completed upper secondary education. The Level 4–6 category includes respondents who completed post-secondary and tertiary education. In some cases, missing information had to be corrected or imputed. In some countries, the months of some events were not reported. In these cases, the dummy month June was entered. In many countries, the older respondents in particular did not remember the exact month of events; in these cases, the dummy month of June was used to replace missing information. Cases in which neither the month nor the year of an event was reported, or in which the information was contradictory (e.g. divorce before marriage) were excluded. In each group (defined by cohort and education), less than 5 per cent of respondents had to be excluded. The exclusions had no systematic influence on the results because they were equally distributed among the respondent groups. Applying the concept of destandardization in empirical research requires a methodological tool that reduces rich and diverse data on events and states to formal characteristics of life courses. SA provides a variety of tools that enable us to conduct analyses of life course sequences. SA was first introduced in the form of optimal matching analysis (OMA) into the social sciences by Abbott and colleagues (Abbott and Hrycak, 1990; Abbott and Tsay, 2000). OMA uses pairwise comparisons of sequences of elements (representing life courses), and counts the minimum number of substitutions and deletions/insertions needed to transform one life course into another. Other SA methods for calculating pairwise sequence dissimilarity are available (for overviews, see Aisenbrey and Fasang, 2010; Studer and Ritschard, 2016). Examples of such methods include subsequence measures, which were introduced by Elzinga and Liefbroer (2007), and which have been supplemented with an episode-based subsequence approach by Elzinga and Studer (2015). To analyse the cohort-specific level of destandardization, we adopt the concept of the average dissimilarity of all life courses, defined as the mean of all values in the dissimilarity matrix (Elzinga and Liefbroer, 2007; Robette, 2010; Fasang 2012; Zimmermann, 2013; Fasang, 2014). We apply this concept to assess differences in destandardization levels between groups differentiated by education and gender. We use the Hamming distance (a simple dissimilarity measure, Hamming, 1950, 1980), which counts the numbers of equal and unequal positions in a pairwise comparison of sequences (see Figure 1). We only count perfect matches, i.e. similarity in selected life course areas is not considered as ‘partly equal’. If applied to sequences representing equal age frames, the measure represents the degree of (dis)similarity between respondents at various chronological ages, i.e. it specifically addresses the ‘timing’ of events (or ‘age standardisation’, Konietzka and Huinink, 2003). Prior comparative analyses of more complex dissimilarity measures revealed that they yield very similar results (Robette and Bry, 2012; Halpin, 2014) but are harder to interpret in terms of the type of life course dissimilarity measured. We calculated 90 per cent bootstrap confidence intervals (Efron and Tibshirani, 1993) to assess the reliability of the results. Figure 1. View largeDownload slide Illustration of the comparison of life course states at similar chronological ages in the Hamming distance Figure 1. View largeDownload slide Illustration of the comparison of life course states at similar chronological ages in the Hamming distance In previous research, the rather abstract analysis of average dissimilarities has often been supplemented by analyses of levels of dissimilarity between life course clusters (Elzinga and Liefbroer, 2007; Robette, 2010; Fasang, 2014). While these analyses have proven to be valuable and plausible, within-cluster heterogeneity may distort comparisons between country and gender groups, as well as between cohorts (Halpin, 2010; Anyadike-Danes and McVicar, 2010). Thus, when analysing the substantial changes in family-related life courses, and when seeking to determine the sequence of life course events and episodes, we suggest specifically focusing on the order of episodes. The concept of ‘episode orders’ is used to identify distinct orders of life course episodes (see Figure 2). In the most common representation of sequence data as strings of symbols, such episode orders appear as a series of equal elements that are not disturbed by other elements in between. Accordingly, episode orders are constructed by eliminating all immediate repetitions of states within a given sequence. Figure 2 gives two examples of life courses that display a common episode order, i.e. that belong to the same group defined by a common episode order. Within life course groups defined in this way, the only differences that can occur are in the timing of events, while the prevalence and the order of states are fixed. An advantage of this approach over cluster analyses based on average sequence dissimilarity is that the kind of heterogeneity that may appear within life course groups—and, hence, the common features of all life courses that constitute one group—is exactly known. A disadvantage of this approach is that a wide variety of episode orders may exist—in our study, up to about 100 different life course groups per country—that cannot be meaningfully analysed as a whole. Figure 2. View largeDownload slide Illustration of the identification of life course groups with common episode orders Figure 2. View largeDownload slide Illustration of the identification of life course groups with common episode orders Empirically, however, no more than the three most common episode orders per country are needed to provide sufficient substantial insights into the dominant life course patterns as these orders represent between one-half and three-quarters of all of the life courses in the countries examined. A drawback of this approach is that heterogeneity and changes in life courses that take place beyond the dominant patterns are disregarded. To sum up, we propose applying a 2-fold measure of destandardization. On the one hand, by focusing on the timing of events, analyses based on the Hamming distance capture (de)standardization on an abstract level. On the other hand, analyses based on common episode orders focus on the order of life course episodes and substantial changes in the dominant life course patterns. In such analyses, (de or re-)standardization is indicated by the disappearance or emergence of one or a few dominant life course patterns. Through the empirical analysis of both ‘average dissimilarities’ and most common ‘episode orders’, we are able to reduce complex and multi-faceted life course data to a manageable level. Results We start by presenting calculations of average dissimilarities of family life courses. Figure 3 displays the cohort-specific levels of dissimilarity for each country, separated by educational level and gender. Figure 3. View largeDownload slide Average dissimilarity of family-related life courses between ages 15 and 35 years by country, birth cohort, gender, and education (ISCED 1997) Note: Average dissimilarity measured by the normalized Hamming distance (1 = maximum dissimilarity (no similar elements), 0 = complete similarity of life courses (all elements are similar)). Because of low respondent numbers, no results are displayed for the 1935–1944 cohort in Germany and for the lower educated born in 1965–1974 in Russia and Estonia. *Values for Italian men (not displayed in the figure, ISCED 4–6) 0.34 (cohort 1955–1964), 0.29 (cohort 1965–1974). Figure 3. View largeDownload slide Average dissimilarity of family-related life courses between ages 15 and 35 years by country, birth cohort, gender, and education (ISCED 1997) Note: Average dissimilarity measured by the normalized Hamming distance (1 = maximum dissimilarity (no similar elements), 0 = complete similarity of life courses (all elements are similar)). Because of low respondent numbers, no results are displayed for the 1935–1944 cohort in Germany and for the lower educated born in 1965–1974 in Russia and Estonia. *Values for Italian men (not displayed in the figure, ISCED 4–6) 0.34 (cohort 1955–1964), 0.29 (cohort 1965–1974). Regarding destandardization (see Hypothesis 1), we find that across cohorts, family life courses have become more destandardized among the lower than the higher educated in all countries except Germany. This finding clearly supports the view (expressed by Hypothesis 1b) that destandardization is driven by social groups with fewer resources. With respect to the process of restandardization as measured by average dissimilarities (Hypothesis 3), we find some support for the suggestion that restandardization is more prevalent among the higher than the lower educated strata (supporting Hypothesis 3a). This result applies to both men and women in France and to men in Italy. For the youngest cohort of these groups, the average level of life course dissimilarity has decreased among the higher but not the lower educated. In contrast, for all countries no evidence is found in support of Hypotheses 1a (which posits that the degree of destandardization is larger among the higher educated) and 3b (which posits that the degree of restandardization is larger among the lower educated). Turning to the analysis of ‘common episode orders’, we find similar patterns for Estonia, France, Germany, Norway, and Russia. Across all cohorts in those countries, the most prevalent life courses are (i) ‘traditional’ life trajectories, in which cohabitation is tied to marriage and precedes family formation; (ii) ‘modernised’ life trajectories, in which cohabitation precedes marriage, and marriage precedes family formation; and (iii) ‘singles’ life courses, in which respondents between ages 15 and 35 years are not partnered and remain childless. The ‘traditional’ order of family-related episodes has become less common over time in all of the countries studied (Figure 4, Supplementary Tables S2, S3), while modernized life courses have been spreading in all of the countries except Hungary and Italy (Figure 5, Supplementary Table S2). In Hungary and Italy, the prevalence of life courses defined by a postponement of events has increased considerably (Figure 6, Supplementary Table S3). Since in all of the countries the third most common life course sequence has been far less prevalent than the other two (Supplementary Table S2), the results are not displayed in the figures. Figure 4. View largeDownload slide Proportions of respondents with a ‘traditional’ life course pattern (simultaneous marriage and cohabitation, followed by childbearing) between ages 15 and 35 years, by country, birth cohort, gender, and education (ISCED 1997) Note: Because of low respondent numbers, no results are displayed for the 1935–1944 cohort in Germany and for the lower educated born in 1965–1974 in Russia and Estonia. Figure 4. View largeDownload slide Proportions of respondents with a ‘traditional’ life course pattern (simultaneous marriage and cohabitation, followed by childbearing) between ages 15 and 35 years, by country, birth cohort, gender, and education (ISCED 1997) Note: Because of low respondent numbers, no results are displayed for the 1935–1944 cohort in Germany and for the lower educated born in 1965–1974 in Russia and Estonia. Figure 5. View largeDownload slide Proportions of respondents with a ‘modernised’ life course pattern (marriage after start of cohabitation, followed by childbearing) between ages 15 and 35 years, by country, cohort, gender, and education (ISCED 1997) Note: Because of low respondent numbers, we do not display results for the 1935–1944 cohort in Germany and for the lower educated born in 1965–1974 in Russia and Estonia. Figure 5. View largeDownload slide Proportions of respondents with a ‘modernised’ life course pattern (marriage after start of cohabitation, followed by childbearing) between ages 15 and 35 years, by country, cohort, gender, and education (ISCED 1997) Note: Because of low respondent numbers, we do not display results for the 1935–1944 cohort in Germany and for the lower educated born in 1965–1974 in Russia and Estonia. Figure 6. View largeDownload slide Proportions of respondents with a ‘postponement’ life course pattern (without cohabitation, marriage, or childbearing) between ages 15 and 35 years, by country, cohort, gender, and education (ISCED 1997) Figure 6. View largeDownload slide Proportions of respondents with a ‘postponement’ life course pattern (without cohabitation, marriage, or childbearing) between ages 15 and 35 years, by country, cohort, gender, and education (ISCED 1997) With respect to educational differences in destandardization, as measured by the prevalence of dominant life course patterns, our empirical results confirm for the eastern European countries Hypothesis 2b, which assumes that the decrease in the prevalence of traditional family formation is larger among the lower educated (Figure 4). In contrast, our results confirm for Italy and (to a lesser extent) for Germany the alternative Hypothesis 2a, which assumes that the higher educated drive the decline in traditional family life courses. The results for Norway and France, where traditional family formation has been vanishing at similar speeds and to similar degrees in all groups, do not support Hypotheses 2a or 2b. Regarding the emergence of new standards of family life courses (Hypothesis 4), the evidence shows that in France, western Germany, and Norway, the adoption of the ‘modernised family life course pattern’ has been increasing more quickly among the higher than the lower educated (Figure 5). This finding clearly supports Hypothesis 4a, which posits that the process of restandardization (via establishing a new standard life course pattern) is stronger among the higher than the lower educated. In Italy, restandardization has been primarily based on large increases in life courses characterized by no major familial transitions before age 35 years among both men and women. Since this pattern has become particularly prevalent among the higher educated (Figure 6), Hypothesis 4a is again supported. In Estonia, Hungary, and Russia, the results are less clear-cut, as no new standard of family life courses is found among the lower or the higher educated. An overview of the results is given in Table 3. A closer look at country differences reveals that no coherent trends are shared by country clusters. An overview of the most important similarities and dissimilarities between countries is given in Supplementary Table S4. Table 3. Synopsis of the main results regarding educational differences Phenomenon Expected to be more pronounced among … … higher educated … lower educated (i) Destandardization (average life course similarity decreasing) Hypothesis 1a Hypothesis 1b Not supported Supported for all countries except Germany (ii) Decreasing prevalence of traditional family formation Hypothesis 2a Hypothesis 2b Supported for Italy Supported for eastern European countries (iii) Restandardization (average life course similarity increasing) Hypothesis 3a Hypothesis 3b Supported for women in France and Germany and for men in Italy Not supported (iv) Emerging new standards of family formation Hypothesis 4a Hypothesis 4b Supported for France, Germany, Italy, Norway Not supported Phenomenon Expected to be more pronounced among … … higher educated … lower educated (i) Destandardization (average life course similarity decreasing) Hypothesis 1a Hypothesis 1b Not supported Supported for all countries except Germany (ii) Decreasing prevalence of traditional family formation Hypothesis 2a Hypothesis 2b Supported for Italy Supported for eastern European countries (iii) Restandardization (average life course similarity increasing) Hypothesis 3a Hypothesis 3b Supported for women in France and Germany and for men in Italy Not supported (iv) Emerging new standards of family formation Hypothesis 4a Hypothesis 4b Supported for France, Germany, Italy, Norway Not supported Conclusions Starting with the premise that processes of life course destandardization do not follow a predefined script in modern societies, the article highlighted social disparities in the destandardization of family life courses. Empirically, we analysed educational stratification in life course trajectories in a number of European countries. In terms of methods, we diverged from commonly used techniques of SA by analysing average dissimilarities of the Hamming distance (which is a simple measure of sequence dissimilarity that focuses on the timing of events) and the most common episode orders (which, by ignoring the duration of episodes, focus on the sequence of events). In our view, this mix of tools proved suitable for empirically examining the conceptual idea of life course destandardization, while also bypassing certain shortcomings associated with comparisons of clusters across cohorts—in particular, the phenomenon of undesired within-cluster variation. Our analyses provided consistent evidence that destandardization is an empirically real phenomenon that has shaped the family life courses of successive birth cohorts in European societies. These findings are in line with classic accounts of social change provided in life course research and family demography (Kohli, 1986; Van de Kaa, 1987; Beck, 1992). However, the differences we found between countries, educational groups, and men and women also suggest that theories that assume that long-term trends in life course change coincide with the shift from a modern to a postmodern society need some refinement (Mayer, 2001, 2005). In light of the empirical evidence we presented on life course change since the 1970s, we conclude that theoretical accounts should be framed in a more specific manner. First, our results support the assumption that after the demise of the historically unique post-war golden age of marriage, destandardization signifies a return to low levels of standardization across Europe, especially among the lower educated. Secondly, the alternative account of destandardization as a transitory stage between an ‘old’ and a ‘new’ standard—i.e. paving the way to restandardization—was found to apply to the higher educated in northern and western Europe, among whom a new common life course pattern has been emerging. At the same time, we found no evidence of restandardization with respect to the dimension of average life course dissimilarity. Thirdly, regarding educational differences, the evidence suggests that primarily the lower educated promoted processes of destandardization. This finding supports the theoretical perspective that links destandardization to deprivation rather than to deliberate choices and less institutional control. Evidence supporting this assumption was found across countries, which is remarkable given the large differences between countries regarding substantial life course changes. Most of these differences are related to the types and the extent of the new patterns of family formation that have been emerging. As indicated above, cohabitation has become common in northern and western European countries, whereas in Italy and (to some degree) in Hungary the ‘postponement pattern’ has prevailed. The findings we have presented are tentative in the sense that they depend on a set of predefined life course states. Future research should seek to include living-apart-together relationships, as individuals we classified as single may have been in intimate relationships. This is especially likely to be the case for the 15–35 years age group and for the higher educated (Liefbroer, Poortman and Seltzer, 2015). Accordingly, it may be possible to detect higher degrees of destandardization among the higher educated (especially in the Italian case) than our results suggest. We also have to acknowledge that our findings may be specific to the period of youth and young adulthood (ages 15–35 years), while later life course phases may be marked by other or even opposing trends. Despite these limitations, our analyses have shown that across the countries under observation, the destandardization of the family life course is a phenomenon that varies considerably by educational attainment. We can therefore conclude that while general theoretical accounts provide an anchor narrative of long-term change in life course patterns, they fail to take into account the dimension of social stratification within life course dynamics. Significant educational disparities in the process of destandardization imply not only that the term ‘destandardisation’ refers to a wide range of phenomena but also that current trends in destandardization are related less to value changes und individualization than to social disintegration. Against this background, our results point to the need for specific research perspectives to be developed in the future. First, life course research should take into account more systematically the variation in the life course patterns of different social strata. At the same time, the role of education needs to be better specified. One question that arises is to what extent changing educational gradients are caused by selection effects, i.e. the changes across cohorts in the relative sizes of the lower and the higher educated groups. A second question that emerges is whether educational disparities affect not only the destandardization of family life courses but also of work lives and careers. Thirdly, country-specific changes should be analysed more carefully within a comparative research framework. Finally, future research would profit from including covariates and using methods of regression analysis for sequences that researchers are currently developing (for example, in Studer, Struffolino and Fasang, 2016). Okka Zimmermann is a postdoctoral researcher at Technische Universität Braunschweig, Department of Social Sciences. Her research interests are life course research, sequence analysis, social change, comparative research, family sociology, and family demography. Her doctoral thesis was accepted in 2017 with distinction (‘summa cum laude’) by Technische Universität Braunschweig. Dirk Konietzka is a University Professor of Sociology at Technische Universität Braunschweig, Department of Social Sciences. His research interests are social stratification, life course research, and family demography. Among others, he has published in Work, Employment and Society, Journal of Youth Studies, European Sociological Review, and Zeitschrift für Soziologie. Together with M. Kreyenfeld he edited ‘Childlessness in Europe. Contexts, Causes, and Consequences’, Springer (2017). Footnotes 1 Hypotheses with respect to life course standardization typically address phenomena on the cohort level (or on the level of sub-groups within cohorts). Although standardization as a construct is generally placed on higher aggregate levels, life course patterns are the outcome of individual behaviour. Hence, hypotheses on life course change should refer to individual actors, their resources, and choices within opportunity structures. Supplementary Data Supplementary data are available at ESR online. Acknowledgements This paper uses data from the National Educational Panel Study (NEPS): Starting Cohort Adults, doi:10.5157/NEPS:SC6:7.0.0. From 2008 to 2013, NEPS data was collected as part of the Framework Program for the Promotion of Empirical Educational Research funded by the German Federal Ministry of Education and Research (BMBF). As of 2014, NEPS is carried out by the Leibniz Institute for Educational Trajectories (LIfBi) at the University of Bamberg in cooperation with a nationwide network. The paper uses data from Generations & Gender Survey (http://www.ggp-i.org/data), a cross-national longitudinal survey of 18-79 year olds in 19 European and 4 non-European countries. Funding The research was supported by the German Research Foundation (DFG), Grant No. KO 2867/4-1. References Abbott A., Hrycak A. ( 1990). Measuring resemblance in sequence data. American Journal of Sociology , 96, 144– 185. Google Scholar CrossRef Search ADS Abbott A., Tsay A. ( 2000). Sequence analysis and optimal matching methods in sociology: review and prospect. Sociological Methods and Research , 29, 3– 33. Google Scholar CrossRef Search ADS Aisenbrey S., Fasang A. E. ( 2010). New life for old ideas: the “second wave” of sequence analysis bringing the “course” back into the life course. Sociological Methods and Research , 38, 420– 462. Google Scholar CrossRef Search ADS Anyadike-Danes M., McVicar D. ( 2010). My brilliant career- characterizing the early labor market trajectories of British women from generation X. Sociological Methods and Research , 38, 482– 512. Google Scholar CrossRef Search ADS Beck U. ( 1992). Risk Society, Towards a New Modernity . London: Sage Publications. Billari F. C., Liefbroer A. ( 2010). Towards a new pattern of transition to adulthood? Advances in Life Course Research , 15, 59– 75. Google Scholar CrossRef Search ADS Brückner H., Mayer K. U. ( 2005). Destandardization of the life course: what it might mean? And if it means anything, whether it actually took place? Advances in Life Course Research , 9, 27– 53. Google Scholar CrossRef Search ADS Blossfeld H.-P. et al. ( 2005). Globalization, Uncertainty and Youth in Society . London: Routledge. Blossfeld H.-P., Roßbach H.-G., von Maurice J. (Eds.) ( 2011). Education as a Lifelong process—the German National Educational Panel Study (NEPS). Special issue of Zeitschrift Für Erziehungswissenschaft, vol. 14. Wiesbaden: Springer VS. Buchmann M. ( 1989). The Script of Life in Modern Society. Entry into Adulthood in a Changing World . Chicago; London: University Press. Efron B., Tibshirani R. J. ( 1993). An Introduction to the Bootstrap . Boca Raton; London; New York; Washington, DC: Chapman & Hall/CRC. Google Scholar CrossRef Search ADS Elzinga C. H., Liefbroer A. C. ( 2007). Destandardization of family-life trajectories of young adults. European Journal of Population , 23, 225– 250. Google Scholar CrossRef Search ADS Elzinga C. H., Studer M. ( 2015). Spell sequences, state proximities and distance metrics. Sociological Methods and Research , 44, 3– 47. Google Scholar CrossRef Search ADS Esping-Andersen G. ( 1999). Social Foundations of Postindustrial Economies . Oxford University Press. Google Scholar CrossRef Search ADS Fasang A. E. ( 2012). Retirement patterns and income inequality. Social Forces , 90, 685– 711. Google Scholar CrossRef Search ADS Fasang A. E. ( 2014). New perspectives on family formation: what can we learn from sequence analysis? In Blanchard P., Bühlmann F., Gauthier J.-A. (Eds.), Advances in Sequence Analysis: Methods, Theories and Applications . New York; Heidelberg; Dordrecht; London: Springer, pp. 107– 128. Ferrera M. ( 1996). Southern model of welfare in social Europe. Journal of European Social Policy , 6, 17– 37. Google Scholar CrossRef Search ADS Festy P. ( 1980). On the new context of marriage in Western Europe. Population and Development Review , 6, 311– 315. Google Scholar CrossRef Search ADS Friedman D., Hechter M., Kanazawa S. ( 1994). A theory of the value of children. Demography , 31, 375– 401. Google Scholar CrossRef Search ADS PubMed Furstenberg F. F.Jr, Rumbaut R. G., Settersten R. A.Jr. ( 2005). On the frontier of adulthood. Emerging themes and new directions. In Settersten R. A.Jr., Furstenberg F. F.Jr., Rumbaut R. G. (Eds.), On the Frontier of Adulthood: Theory, Research, and Public Policy . Chicago, IL: University Press, pp. 3– 25. Google Scholar CrossRef Search ADS Halpin B. ( 2010). Optimal matching analysis and life-course data. The importance of duration. Sociological Methods and Research , 38, 365– 388. Google Scholar CrossRef Search ADS Halpin B. ( 2014). Three narratives of sequence analysis. In Blanchard P., Bühlmann F., Gauthier J.-A. (Eds.), Advances in Sequence Analysis: Methods, Theories and Applications . New York; Heidelberg; Dordrecht; London: Springer, pp. 75– 103. Hamming R. W. ( 1950). Error-detecting and error-correcting codes. Bell System Technical Journal , 29, 147– 160. Google Scholar CrossRef Search ADS Hamming R. W. ( 1980). Coding and Information Theory . Englewood Cliffs: Prentice Hall. Hiekel N., Liefbroer A. C., Poortman A.-R. ( 2014). Understanding diversity in the meaning of cohabitation across Europe. European Journal of Population , 30, 391– 410. Google Scholar CrossRef Search ADS Hoem B., Hoem J. M. ( 1988). The Swedish family: aspects of contemporary developments. Journal of Family Issues , 9, 397– 424. Google Scholar CrossRef Search ADS Huinink J. ( 2013). De-standardisation or changing life course patterns? Transition to adulthood from a demographic perspective. In Neyer G. et al. (Eds.), The Demography of Europe . New York: Springer, pp. 99– 118. Google Scholar CrossRef Search ADS Inglehart R. ( 1990). Culture Shift in Advanced Industrial Society . Princeton: Princeton University Press. Inglehart R. ( 1997). Modernization and Postmodernization: Cultural, Economic, and Political Change in 43 Societies . Princeton: Princeton University Press. Kalmijn M. ( 2011). The influence of men’s income and employment on marriage and cohabitation: testing Oppenheimer’s theory in Europe. European Journal of Population , 27, 269– 293. Google Scholar CrossRef Search ADS PubMed Kiernan K. ( 2001). The rise of cohabitation and childbearing outside of marriage in Western Europe. International Journal of Law, Policy and the Family , 15, 1– 21. Google Scholar CrossRef Search ADS Kogan I., Clemens N., Gebel M. (Eds.) ( 2011). Making the Transition – Education and Labor Market Entry in Central and Eastern Europe . Stanford: Stanford University Press. Google Scholar CrossRef Search ADS Kohli M. ( 1986). The world we forgot: a historical review of the life course. In Marshall V. W. (Ed.), Later Life. The Social Psychology of Aging . London: Sage, pp. 271– 303. Kohli M. ( 2007). The institutionalization of the life course. Human Development , 4, 253– 271. Google Scholar CrossRef Search ADS Konietzka D., Huinink J. ( 2003). Die Destandardisierung einer Statuspassage? Soziale Welt , 54, 285– 311. Konietzka D. ( 2010). Zeiten des Übergangs. Sozialer Wandel des Übergangs in das Erwachsenenalter . Wiesbaden: Springer VS. Kravdal Ø., Rindfuss R. R. ( 2008). Changing relationships between education and fertility: a study of women and men born 1940 to 1964. American Sociological Review , 73, 854– 873. Google Scholar CrossRef Search ADS Kreidl M., Ganzeboom H. B. G., Treiman D. J. ( 2014). How Did Occupational Returns to Education Change Over Time? California Center for Population Research On-Line Working Paper, 2014-017. Los Angeles: University of California. Kreyenfeld M., Konietzka D. ( 2008). Education and fertility. In Hamm I., Seitz H., Werdig M. (Eds.), Demographic Change in Germany. The Economic and Fiscal Consequences . Heidelberg: Verlag, pp. 165– 187. Kreyenfeld M., Hornung A., Kubisch K. ( 2013). Der deutsche Generations and Gender Survey: Einige kritische Betrachtungen zur Validität der Fertilitätsverläufe. Comparative Population Studies – Zeitschrift Für Bevölkerungswissenschaft , 38, 29– 58. Lesnard L. et al. ( 2016). Do transitions to adulthood converge in Europe? An optimal matching analysis of work–family trajectories of men and women from 20 European countries. European Sociological Review , 32, 355– 369. Google Scholar CrossRef Search ADS Lesthaeghe R. ( 2010). The Unfolding Story of the Second Demographic Transition. PSC Research Report, 10-696. Ann Arbor: University of Michigan's Population Studies Center. Liefbroer A. C., Poortman A., Seltzer J. A. ( 2015). Why do intimate partners live apart? Evidence on LAT relationships across Europe. Demographic Research , 32, 251– 286. Google Scholar CrossRef Search ADS PubMed Mayer K. U. ( 2001). The paradox of global social change and national path dependencies. In Woodward A. E., Kohli M. (Eds.), Inclusions and Exclusions in European Societies . London: Routledge, pp. 89– 110. Mayer K. U. ( 2005). Life courses and life chances in a comparative perspective. In Svallfors S. (Ed.), Analyzing Inequality: Life Chances and Social Mobility in Comparative Perspective . Stanford: University Press, pp. 17– 55. Mayer K. U. ( 2009). New directions in life course research. Annual Review of Sociology , 35, 413– 433. Google Scholar CrossRef Search ADS McLanahan S. ( 2004). Diverging destinies. How children fare under the second demographic transition. Demography , 41, 607– 627. Google Scholar CrossRef Search ADS PubMed McMillan R. ( 2005). The structure of the life course: classic issues and current controversies. Advances in Life Course Research , 9, 3– 24. Google Scholar CrossRef Search ADS Mirowsky J., Ross C. E. ( 2003). Education, Social Status and Health . New York: Aldine de Gruyter. Myles J. ( 1992). Is there a post-fordist life course? In Heinz W. R. (Ed.), Institutions and Gatekeeping in the Life Course . Weinheim: Deutscher Studienverlag, pp. 171– 185. Oppenheimer V. K. ( 1988). A theory of marriage timing. American Journal of Sociology , 94, 563– 591. Google Scholar CrossRef Search ADS Oppenheimer V. K. ( 2000). The continuing importance of men's economic position in marriage formation. In Waite L. J., et al. (Eds.), The Ties that Bind: Perspectives on Marriage and Cohabitation . New York: Transaction Publishers, pp. 283– 301. Perelli-Harris B., Sigle-Rushton W., Kreyenfeld M., Lappegård T., Keizer R., Berghammer C. ( 2010). The educational gradient of childbearing within cohabitation in Europe. Population and Development Review , 36, 775– 801. Google Scholar CrossRef Search ADS PubMed Rindfuss R. R., Swicegood C. G., Rosenfeld R. A. ( 1987). Disorder in the life course: how common and does it matter? American Sociological Review , 52, 785– 801. Google Scholar CrossRef Search ADS Robette N. ( 2010). The diversity of pathways to adulthood in France. Advances in Life Course Research , 15, 89– 96. Google Scholar CrossRef Search ADS Robette N., Bry X. ( 2012). Harpoon or bait? A comparison of various metrics in fishing for sequence patterns. Bulletin De Methodologie Sociologique/Bulletin of Sociological Methodology , 116, 5– 24. Google Scholar CrossRef Search ADS Sauer L., Ruckdeschel K., Naderi R. ( 2012). Reliability of Retrospective Event Histories Within the German Generations and Gender Survey. BiB Working Paper, 1/2012. Wiesbaden: Bundesinstitut für Bevölkerungsforschung. Scherger S. ( 2007). Destandardisierung, Differenzierung, Individualisierung. Westdeutsche Lebensläufe im Wandel . Wiesbaden: VS. Studer M., Ritschard G. ( 2016). What matters in differences between life trajectories: a comparative review of sequence dissimilarity measures. Journal of the Royal Statistical Society: Series A (Statistics in Society) , 179, 481– 511. Google Scholar CrossRef Search ADS Studer M., Struffolino E., Fasang A. E. ( 2016). New tool for old questions: the sequence-analysis multistate model to study relationships between time-varying covariates and trajectories. In Ritschard, G. and Studer. M. (Eds.), Proceedings of the International Conference on Sequence Analysis and Related Methods, Lausanne, June 8-10, 2016, pp. 429– 432. Van de Kaa D. J. ( 1987). Europe’s second demographic transition. Population Bulletin , 42, 1– 59. Google Scholar PubMed Zimmermann O. ( 2013). Temporary destandardisation of partnership formation and continuous standardisation of fertility in three GGS countries. Demográfia English Edition , 56, 62– 88. © The Author 2017. Published by Oxford University Press. All rights reserved. For permissions, please e-mail: email@example.com
European Sociological Review – Oxford University Press
Published: Feb 1, 2018
It’s your single place to instantly
discover and read the research
that matters to you.
Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.
All for just $49/month
Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly
Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.
Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.
Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.
All the latest content is available, no embargo periods.
“Hi guys, I cannot tell you how much I love this resource. Incredible. I really believe you've hit the nail on the head with this site in regards to solving the research-purchase issue.”Daniel C.
“Whoa! It’s like Spotify but for academic articles.”@Phil_Robichaud
“I must say, @deepdyve is a fabulous solution to the independent researcher's problem of #access to #information.”@deepthiw
“My last article couldn't be possible without the platform @deepdyve that makes journal papers cheaper.”@JoseServera