INTRODUCTION

From the editor

Mobility studies constitute the most technically sophisticated major branch of contemporary sociology, as is well illustrated by the methodological array of this special issue. Perhaps just because of their methodological acumen and evidential scruples, more than any other part of the discipline, they bring forth, frustratingly or fascinatingly, the complex and inconclusive character of social processes, their ‘constant flux’ as Robert Eriksson and John Goldthorpe summed it up in their landmark study. Neither de-industrialization nor globalization seems to have created any new general trend of inter- and of intra-generational mobility. The long shadow of parental background is still there, varying somewhat between countries. Education matters, but not more than decades ago. Occupational careers have, on the whole, not become more fluid and unpredictable. They are still governed by social birth and education, which sometimes, but not always, accumulate inequalities over the life-course.

Editorial credit for this issue goes to Carlo Barone at the University of Trento, Italy, whom I want to thank for a very good job. We have had a serious problem of space to accommodate the full richness of all the contributions. When we finally got the page tally, some cuts had to be made, which the guest editor and the contributors have accepted in good grace. As a modest contribution to austerity, I have refrained from writing a normal editorial.

Göran Therborn

1. Introduction

This special issue of European Societies is devoted to the study of career mobility1 in five European countries (Sweden, Great Britain, The Netherlands, Germany, and Italy) and to its implications for the dynamics of intergenerational reproduction. This introduction illustrates the common research questions addressed by the country-specific contributions and outlines some common findings. The articles are the outcome of a collaborative research project developed within the Equalsoc Network of Excellence, which is funded by the European Union under the 6th Framework Programme. In this context, the national experts from the above-mentioned countries have agreed on a set of research questions concerning the connections between intragenerational and intergenerational mobility. These research topics and their overall significance for social stratification research will be discussed in Sections 2 and 3, while in Section 4 we will provide an overview of the main findings of the project.

A key requirement to participate in this project was access to longitudinal information on the work histories of several birth cohorts. This requirement has considerably constrained the number of countries for which reliable data are available.2 Of course, the five nations under study are not representative of all European nations. However, they cover a wide variety of welfare arrangements and labour market regulations. Moreover, we cover a heterogeneous set of countries with respect to intragenerational mobility, from Italy, an outlier with low career mobility rates, to Great Britain, where career mobility is far more common.

2. Career mobility and intergenerational reproduction: some unsolved issues for social mobility research

Social mobility research has for a long time been a debate about modernisation theory. This debate has focused on a small number of substantive questions concerning the relationship between social origins, education, and occupational destinations, the so-called O–E–D triangle. Education has invariably emerged as the main determinant of occupational attainment in modern countries, although ‘direct effects’ of family background have been extensively documented. However, these direct effects operating over and above the influence of education were supposed to weaken according to modernisation theory, whereas the importance of educational qualifications for labour market success was expected to increase (Blau and Duncan 1967; Treiman 1970; Bell 1973). These two supposed trends, together with predictions regarding the declining influence of family background on educational attainment, would promote increasing social fluidity. A great deal of social mobility research assesses whether these long-term trends associated with the O–E–D triangle are indeed at work (Ganzeboom et al. 1991; Wolbers et al. in this issue). These key research questions have been raised since the first post-war generation of social mobility research (Glass 1954; Lipset and Bendix 1959), and they became the core research focus of the second generation (Blau and Duncan 1967; Sewell and Hauser 1975), which used models of status attainment to test the general hypothesis that industrialisation promotes achievement and reduces ascription. The same hypothesis was addressed (and largely rejected) by the third generation of social mobility studies, which was dominated by log-linear models of occupational mobility (Erikson and Goldthorpe 1992; Breen 2004). Although there is evidence of a slight increase of social fluidity in some countries, social class returns to education look largely stable across cohorts or, if anything, tend to weaken.

Although the main analytical focus of these studies is intergenerational mobility, some have considered intragenerational mobility as well, as is most evident in status attainment research, where the first occupation figures among the intervening variables between origins and destinations. The analogue to this focus in a mobility table framework is the log-linear analysis of three-way tables representing origin, first and last occupation (Erikson and Goldthorpe 1992: ch. 8). However, intergenerational mobility has been considered far more than intragenerational mobility, and the connections between the two have been seldom explored, as discussed by Bukodi and Goldthorpe in this issue.

Unsurprisingly, these mainstream approaches were a favourite target of criticism by the proponents of work history analyses, which began to accumulate in the mid-1980s (Blossfeld 1986; Sørensen 1986; Carroll and Mayer 1987). The main substantive critique is that if only origins and destinations are examined, we fail to consider the overall process of career development, which is treated as a black box. Additionally, if career advancement incorporates some relevant mechanisms of intergenerational reproduction, the risk is that we fail to adequately grasp how social origins shape achievement in the labour market. There is also an important methodological objection at play: because standard social mobility studies have often relied on a single cross-sectional survey, they have been forced to make the questionable assumption that age differences can be directly interpreted as cohort effects. However, the dynamics of social change across birth cohorts and historical periods should be carefully disentangled from life course processes to achieve reliable empirical results and sound theoretical conclusions.

These critiques have merit. However, the key challenge that they raise is to empirically demonstrate that the analysis of detailed work histories enables a better understanding of processes of intergenerational reproduction. In our view, this challenge is still open: the proof is yet to be given. As discussed by Bukodi and Goldthorpe in this issue, this novel approach has tended to replace the above-mentioned core substantive concerns of social mobility research with a rather different set of questions (Ganzeboom et al. 1991).3 Of course, focusing on new research issues, such as life cycle destandardisation or the impact of globalisation on work careers (Blossfeld et al. 2006), is both perfectly legitimate and highly interesting. However, if we fail to connect the analysis of career dynamics to the role of social origins in occupational attainment, the heuristic value of the ‘new’ approaches to the ‘old questions’ remains unclear. We must still work to understand if and how intergenerational inequalities are shaped by career moves.

In turn, this general research question – which originally inspired our collaborative project – may be conceptualised within the framework of the O–E–D triangle, and thus broken down into two more specific questions. In this regard, it may be noted that research on the role of education in occupational success has focused predominantly on labour market entry or on what position an individual holds at the beginning of his or her career. Little is known about returns to educational qualifications at later stages. This may be problematic if some portion of these returns arises after the beginning of one's career and would make later jobs more strongly associated with education than the first job, as argued by human capital theory (Sicherman 1990; McGuinness 2004). However, signalling and job screening models predict the opposite: that the effects of education are more restricted (Goldthorpe 2009) and should be more pronounced at the beginning of work careers than they are later on, once employers have access to direct, first-hand information about actual worker productivity. Then credential requirements leave room for actual performance to determine how a person's career might progress. These kinds of arguments may justify a positive image of career mobility, understood as a meritocratic competition in which individual efforts and skills play a major role. Hence, our first more specific question concerns the evolution of E–D over one's career. In other words, the articles in this special issue examine how returns to education change after the first job.

Similarly, the existence of direct effects of social origins on a person's first occupation is well documented, but little is known about the evolution of these effects over the life course. This gap poses a problem because the main mechanisms that are supposed to be at work in this effect may operate beyond the point of labour market entry. For instance, an individual might inherit a family business after having gained work experience in one or more other positions. Additionally, the social networks provided by family may support career advancement. Furthermore, parental background may shape personality attributes, including the ‘big five’,4 and influence social attitudes (e.g., orientation toward risk-seeking) and other cultural resources that work as incentive-enhancing preference traits, thus promoting or hindering career advancement (Bowles et al. 2001; Jackson 2007; Goldthorpe 2009). If the distribution of resources and attributes valued by employers is influenced by ascriptive factors, the positive image of career mobility as the locus of a meritocratic competition between individuals is seriously undermined. Our second research question concerns the evolution of the direct effects of origins over the life course.

In short, it is currently unclear whether the influences of education and that of social origins (net of education) are confined to labour market entry or whether, on the contrary, they remain a factor operating throughout the life course. If the latter hypothesis is valid, we would like to determine the overall result of these two influences. That is to say, is social fluidity (i.e., the overall O–D association) strengthened or weakened over the career? Answering this question should help to clarify to what extent intragenerational mobility affects the mechanisms of intergenerational reproduction.

It should be noted that modernisation theory does not deny the existence of ascriptive inequalities in contemporary societies. However, it is supposed that a secular trend exists for them to decline. Not only is the overall conditioning of social origins seen as declining across cohorts, but it also should be increasingly channelled via education. Thus, social origins may still affect labour market entry as well as career advancement, but this influence should weaken over time in favour of an increased role of educational qualifications. Unfortunately, empirical evidence concerning these supposed trends is either controversial (as regards an individual's first occupation) or largely missing (with reference to career mobility dynamics). The contributions in this special issue begin to fill this gap by assessing cohort and period variations at the time of labour market entry and at later stages. This task involves the methodological problem of disentangling period, cohort, and life cycle effects, as discussed in the next section.

3. When does career mobility occur? The debate about occupational maturity

The analysis of social change in career mobility processes presents some specific difficulties that do not arise when we focus only on the first occupation. In particular, when career development is given consideration, it becomes problematic to disentangle life cycle, cohort and period effects5 due to the mathematical impossibility of observing two individuals at the same point in time that are of the same age but were born on different dates. This is a critical problem for social mobility research given its substantive focus on long-term trends and, more generally, on social change. As already mentioned, this difficulty is indeed a common target of critics of standard social mobility studies who instead propose work-history analysis as an alternative method. For instance, cohort differences in social fluidity are difficult to interpret as ‘true’ cohort effects because individuals belonging to different birth cohorts have careers of different lengths that have developed over different periods.

Although a number of possible solutions to this identification problem have been proposed, none has reached undisputed consensus. These solutions typically impose some restrictions on the specification of the three effects, for example, by introducing some assumptions regarding their functional form or by postulating that one of them equals zero. In the case of social mobility studies, the latter solution is opened by the so-called assumption of occupational maturity (Goldthorpe 1987: 52–3), which asserts that individuals reach a stage in their careers after which occupational changes conductive to significant upward or downward mobility become relatively uncommon. In other words, it is suggested that there exists an initial stage that is ‘critical’ for career promotion (or demotion). After this stage, job mobility is still possible but should mostly involve horizontal moves that are not much consequential for individuals’ locations in the occupational hierarchy. If this assumption is valid, and if we restrict our analysis to those individuals who have reached occupational maturity, life cycle effects on occupational attainment will not bias our assessment of historical trends in career mobility.

In short, the assumption of occupational maturity may represent a simple, viable solution to the identification problem that arises in analysing social change in the context of career mobility processes. As discussed by Härkönen and Bihagen in this issue, there are some substantive reasons why this stage of occupational maturity may indeed exist. However, although this assumption is often invoked, there is scant evidence concerning its validity. Hence, another purpose of our research project has been to assess the validity of this assumption. More generally, a common substantive aim of the contributions in this special issue has been to gain a better understanding of the timing of career mobility. In the previous section, it was argued that we do not know exactly if and how the O–E–D triangle evolves over an individual's career. We also do not know when career mobility is most significant for the process of social stratification. Is there a critical stage for career progression after which individuals’ location within the occupational hierarchy in unlikely to change? The longitudinal analyses presented in this special issue provide some provisional answers to this question.

4. Commonalities and differences in career mobility dynamics across five European societies

In this section, we will briefly outline the main findings of our collaborative project with reference to the research issues that we have just outlined. Several shared conclusions emerge from our contributions despite the different methodological approaches employed by the authors. Indeed, a distinctive feature of this special issue is that it covers the full variety of analytical approaches available for use in studying career mobility, from status attainment models and log-linear techniques to sequence analysis, event history models, and growth curve models (which are meant to offer the combined advantages of the former two approaches). Furthermore, also the measurement of occupational position comprises several options, including social class, social standing, social status and income scales. Of course, it is clear that results based on different kinds of data, different variable specifications and different statistical techniques cannot be compared too strictly.6 Moreover, the analyses for Britain and The Netherlands are restricted to men, whereas those for Sweden, Germany, and Italy cover both sexes. Hence, rather than engaging in detailed cross-national comparisons, we will simply try to identify the main similarities and differences between the substantive conclusions reached by our contributors.

As stated above, what follows is intended simply to show only the overall picture that emerges from our research project. The authors of the country-specific contributions have addressed several additional theoretical, methodological and empirical issues that we will not present here. In fact, each article contains a rich field of information on domestic career dynamics that extends far beyond the simplified overall image of career mobility processes presented below.

1. Education displays marked effects on first occupation, but it only moderately mediates the overall influence of social origins. In Italy, three-quarters of the total effect of parental occupation on children's occupations persist once educational qualifications are controlled for. The direct effect of origins net of education is very pronounced in all Italian cohorts. In Sweden and The Netherlands, some 40 to 50 percent of the overall effect of family background operates through education, so direct effects account for about half of the total influence of origins. Additionally, the results for Britain point to rather strong direct effects of social origins, more for social status than for children's earnings. However, in Germany the impact of father's occupation on current occupational status works predominantly through education, and the direct effect of origins appears to be rather small. As explained by Hillmert, these results should be understood in the light of the peculiarities of the German educational system, which promotes particularly strong schooling inequalities while simultaneously strengthening the labour market value of educational qualifications. Taken together, these results undermine the idea that contemporary societies can be described as education-based meritocracies. The influence of social origins remains largely unaccounted by educational achievement. However, it remains to be seen whether there at least exists a trend across cohorts toward growing importance for education.

2. The influence of education on the first occupation is not increasing over time. If anything, it has tended to decline in recent cohorts, and cohort specificities prevail over secular tendencies. The Swedish team reports evidence of diminishing returns to educational credentials. Returns to education have increased in Britain for all qualifications, except the tertiary level if we consider social status. However, this is not the case when we analyse occupational attainment based on earnings scales. Rather, there is empirical evidence of a specific effect for the intermediate cohort of individuals born in 1958. In The Netherlands, the influence of education on the social standing of the first job is strong, but stable across cohorts. In Germany, the marked effect of educational qualifications on the social standing of one's first job increases for the initial cohorts (1929–1931) and peaks for both men and women born in the 1940s, but then declines for recent cohorts. Hence, there is little evidence of any secular trend towards the increased importance of education for occupational attainment. Only in Italy is it possible to detect a significant, progressive increase in the role of education at the time of labour market entry. This finding may be understood in the context of the extremely slow educational expansion that Italy has experienced, which has probably prevented (or at least attenuated) credential inflation until recently. Interestingly, this trend of growing returns on education has come to a halt for the last Italian cohort (1958–1967), which has experienced much more rapid educational growth. Overall, it seems problematic to claim that the five countries under examination are progressively moving towards an education-based meritocracy, at least insofar as we consider entry into the labour market. However, we should also discuss whether this conclusion holds when career mobility enters the picture.

3. Early disadvantages in the labour market have a marked and lasting effect on occupational attainment at later stages in all countries under examination. This statement applies to all cohorts under examination, and little change over time is apparent in this respect. Italy is perhaps the ‘extreme’ national case in this regard: downward mobility is virtually absent, and upward mobility flows are highly uncommon and are largely confined to short-range moves. Hence, it is entirely likely that one will spend the whole career at the same occupational level as that of the first job. Some increase in career fluidity across cohorts is detectable in Italy, but the magnitude of the change appears to be quite small. Career mobility is more widespread in Sweden, but Härkönen and Bihagen conclude that career advancements over the life course are unlikely to overcome the initial differences in occupational prestige that exist between individuals at the point of labour market entry. Furthermore, the Swedish team finds only limited differences across cohorts in patterns of career mobility, even though men in the intermediate birth cohort (1955–1964) stand out for their slower career progression. This cohort specificity is attributed to the difficulties that Swedish industries have faced since the early 1970s, when most members of this cohort began their careers. These results suggest that early disadvantages in the labour market can have a lasting impact on later achievement, which is intriguingly similar to those findings reported by Bukodi and Goldthorpe for the British case. Although the overall amount of intragenerational mobility looks to be sustained in Britain, at least relative to Italy and Sweden, the general patterns at work in this country and the overall rates of career progression at play are broadly similar for the three cohorts under examination. The social status at the time of the first occupation has a marked and enduring effect on later occupational attainment in Britain as well. These effects are not declining across cohorts. Furthermore, Bukodi and Goldthorpe report consistent evidence of cohort-specific effects: men born in the late 1950s whose early work-life histories took place in a period of severe economic difficulties exhibit distinctive patterns of occupational attainment as compared to both the earlier and the later cohort. In The Netherlands, the social standing of the first job greatly affects one's occupational position at later stages. Little change across cohorts occurs in this respect, although in the last cohort there is some decline in the effect of the first job on the job attained 10 years later. The status attainment model estimated for different German cohorts indicates that the influence of the first occupation on the current occupation is always strong. It exhibits trendless fluctuations over time. Drawing on various analyses, Hillmert concludes that ‘in the course of a career, it becomes increasingly unlikely to change the individual occupational status; in other words, the status order becomes consolidated’. Overall, these findings suggest that the potential for career mobility to reshape the disparities produced at the time of labour market entry is quite limited. The lack of sustained cohort-specific trends towards increased career mobility undermines arguments about the supposed increasing fluidity (or destructuration) of work careers.

4. Less educated individuals enjoy slower career progression, so their disadvantages tend to increase. However, this dynamic of cumulative inequality may be attenuated by ceiling effects that constrain the chances of further career advancement enjoyed by highly educated individuals. In The Netherlands, more educated workers attain better jobs after 10 or 20 years of work experience, even if the prestige of the first job is controlled for. These results are highly stable across different birth cohorts. In Germany, education has a strong positive influence on occupation at age 30 among individuals who started their ca reers in jobs with similar prestige. Again, no trend is apparent in this regard. Additionally, Hillmert finds consistently negative effects of initial occupational level on upward career mobility. This result points to the existence of ceiling effects. In Britain, more educated individuals tend to progress more rapidly, both in terms of social status and in the earnings hierarchy, and again, little change across cohorts is detected. In Sweden, a similar tendency exists, but it is not monotonic: men with compulsory education display slower career progression than men with academic secondary education or lower tertiary degrees. However, the latter progress more than university graduates on the career ladder. The career patterns of women are generally similar to those of men. In Italy, more educated workers enjoy greater chances of upward career mobility, although this effect is small and perfectly stable over time. This result is quite unsurprising once we consider that opportunities for career mobility are severely constrained for all social groups in this country. Moreover, for both Italy and Sweden, estimates of growth curve models provide evidence of a negative correlation between the level of one's first job and the probability of further career advancement. This result suggests the existence of ceiling effects that attenuate the chances of career promotion for individuals who begin their careers at the top of the occupational hierarchy. Also the results reported by the Dutch team point to the existence of this kind of ceiling effect.

5. The direct effects of family background are not reduced over one's career; if anything, they are strengthened. In Sweden, the influence of social background net of education remains roughly stable over one's career. However, men from the upper service class exhibit a more rapid career progression than men from lower classes and thus progressively enhance their competitive advantage. In Germany, Italy, and The Netherlands, the conditioning of social origins operates predominantly at labour market entry and is weakly modified later on. In other words, career mobility does not strengthen or weaken the influence of social origins. However, Hillmert reports that counter-mobility plays an important role in intragenerational mobility in Germany: when an individual experiences a status loss relative to the position of her parents, her chances of occupational promotion are substantially increased. This finding suggests that work-life mobility contributes to the crystallisation of intergenerational inequality. Finally, in Britain, family background significantly affects the level of the occupation achieved at occupational maturity, even when the first occupation is controlled for. This result emerges in analyses focusing on both earnings and the social status associated with jobs. Overall, our results indicate quite unequivocally that direct forms of ascriptive inequality are not attenuated by career mobility processes.

Bison's contribution focuses specifically on the interplay between origins and education as determinants of differential patterns of career mobility. Rather than considering the occupational position that a person holds at a given point in time, Bison analyses the whole career trajectory following a sequence analysis perspective. This approach usually relies on optimal matching techniques, which have been criticised because their algorithms for computing the distances between trajectories are largely conventional and somewhat arbitrary. Bison discusses this problem and proposes a possible solution based on a lexicographic index. This novel tool is used to analyse class attainment over the life course in contemporary Italy and ultimately identifies 15 distinct career patterns. Then Bison analyses how education and origins affect the most significant patterns. He finds that education is crucial for access to skilled white-collar jobs, but he indicates that the chances of upward career mobility from these positions to the service class depend much more on social origins than on education. A high level of schooling is what most reduces the chances of moving from white-collar positions to self-employment. Education has a great deal of influence also on one's likelihood of rapidly entering the middle classes from a blue-collar job, but social origins also significantly affect these career trajectories. For instance, family background is particularly important for career trajectories involving access (or a return) to self-employment. Hence, the relative balance between origins and education as determinants of career mobility varies considerably depending on the specific career trajectory under examination. These results indicate the ‘added value’ of a class approach compared to continuous measures of social status or social standing, as the former captures non-vertical dimensions of occupational position that can affect patterns of career mobility.

6. Occupational maturity exists but varies significantly across individuals and cohorts. In Sweden, career advancement is more rapid at the beginning of one's career and then slows down. In the earliest birth cohort (1925–1934), the chances of upward mobility flattened around age 30, but in the youngest cohorts, those including individuals born in the 1950s and 1960s, this threshold moved towards age 40, probably as a result of educational expansion. In Italy, the limited volume of intragenerational mobility is concentrated within the first 15 years of occupational careers, with little change across cohorts. Goldthorpe and Bukodi report that in the early years of working life job moves tend to be relatively frequent before falling off, usually at some point between the ages of 30 and 40. Afterwards, they are less likely to involve any significant change in occupational level. Also in Germany career mobility steadily declines throughout the life course, although no specific point in time can be identified for such a consolidation. In The Netherlands, only a minority of men reach the highest occupational ranks immediately with their first job; the majority experience upward mobility, mostly in the early years of their working careers. Interestingly, recent cohorts and tertiary graduates reach the peak of their careers more rapidly.

5. Concluding remarks

The results that we have briefly summarised return quite a coherent picture of career mobility dynamics in contemporary European societies. Unfortunately, this picture looks discouraging. Career mobility processes do not seem to correct the socially biased allocation of individuals to their first jobs. We already knew that educational participation and the transition to the labour market are affected by marked ascriptive inequalities, but the contributions to this special issue indicate that we cannot expect any significant degree of equalisation based on career mobility. The competition between workers in market economies cannot be easily depicted as a chance for unexploited talents and individual efforts to emerge. Indeed, several results reported by the contributors support the opposite conclusion that work-life mobility may enhance the dynamics of cumulative disadvantage in several respects, as extensively discussed by Hillmert in his contribution. On the one hand, the direct conditioning of social origins on the level of the first job is not attenuated during later stages of one's career. On the other, we find evidence that the influence of education remains largely stable over the career. The overall effect of family background on the level of one's first job is not attenuated over the life course.

All of the articles in this special issue search extensively for evidence of any systematic change across cohorts, but they enjoy little success in this regard. To be sure, some significant differences between cohorts can be detected in terms of the patterns and determinants of occupational attainment. However, they relate to specific historical events or conjunctures rather than pointing to any sustained secular trend. There are in fact few indications that the five countries under study are moving toward an education-based meritocracy. At the same time, we also do not see indications of any sustained, progressive weakening of structural influences on occupational attainment, whether these factors are ascriptive or achieved. There is little support for claims regarding the growing unpredictability of work careers in ‘globalised’, ‘post-modern’ societies.

Of course, future research should move to a fully harmonised comparative analysis to assess cross-national differences in career mobility patterns in greater detail. However, we would be inclined to stress the noticeable convergence of substantive conclusions reached for different countries, which seems even more remarkable and robust given the wide variety of data and analytical procedures employed by the contributors. A challenging task will be to develop a systematic, micro-level explanation for these commonalities. The studies that are presented in this special issue are admittedly of a largely descriptive nature. They address some relatively new research issues and try to establish a set of empirical regularities. To the extent that their conclusions are confirmed by future research, they will certainly call for more formalised explanatory efforts.

Acknowledgements

We would like to thank our colleagues who kindly accepted to act as reviewers for the contributions included in this special issue. A special thanks to Colin Mills for his invaluable comments and suggestions to the authors since the beginning of the project. Finally, we want to thank John Goldthorpe for promoting this collaborative research project and all of the participants in the project for their comments on this introductory essay.

Notes

¹In this contribution, we will use the terms ‘intragenerational mobility’, ‘career mobility’, and ‘work-life mobility’ as synonyms.

²For an updated discussion of the issue of memory bias in retrospective data on career mobility, see Manzoni et al. (2010).

³Carroll and Mayer's (1987) article may be cited as an important exception.

⁴The five traits are openness, conscientiousness, extraversion, agreeableness, and neuroticism.

⁵Life cycle effects refer to the impact of changes associated with aging and with concomitant physiological, psychological and socio-economic transformations. Cohort effects refer to the influence of those specific characteristics of individuals born in the same cohort or those who have any other past event in common than their birth. Period effects refer to the influence of events that simultaneously affect the conditions of individuals in different cohorts at a given point in time.

⁶However, an interesting side-product of our research project concerns the sensitivity analyses conducted in order to assess to what extent our substantive conclusions are robust to different measurement strategies.