Since the global financial crisis in 2009 there has been a renewed interest in the impact of macroeconomic changes on the job trajectories of the workforce. In Germany, sociologists and economists are increasingly painting a gloomy picture about the consequences of macroeconomic changes (Grabka and Frick 2008; Bertelsmann Stiftung 2013; Bosch and Kalina 2015). It is believed that while the older generations witnessed an improvement in job opportunities during the times of the German economic miracle since the 1950s and the expansion of the welfare state in the 1960s and 1970s, recent cohorts not only suffer from declining job entry opportunities but also from more downward mobility, fewer upward job moves, and generally flatter career lines. Based on these changes, it is assumed that the rising middle class is declining again (Herbert-Quandt-Stiftung 2007; Bertelsmann Stiftung 2013; Bosch and Kalina 2015; Nachtwey 2016). Whether this diagnosis is indeed valid or not (see the critique by Groh-Samberg and Hertel 2010; Institut für Sozialforschung und Gesellschaftspolitik 2011; Niehues 2014; Mayer 2016) is still unclear, given that most of the claims and counterclaims are based on relatively short-term (and mostly cross-sectional) empirical evidence contrasting the current labour market situation with what is believed to be true about the changing macrostructure and its effects on job histories in the distant past.

The goal of this study is, therefore, to describe the long-term modernisation process and the changes in the labour market conditions affected by the ups and downs of the business cycle.¹ Further, this study aims to analyse the effects of these macro processes on the entry of men into the labour market in West Germany and their career mobility. Using combined data from the German Life History Study (GLHS) and longitudinal study ‘Working and Learning in a Changing World’ (ALWA: Arbeiten und Lernen im Wandel), we reconstruct the career histories of men born between 1929 and 1976, and analyse how their careers were shaped by macroeconomic changes between 1945 and 2008. As a measure of men’s “goodness of jobs” at entry into the labour market and across the job career we do not use wages but the broader sociological concept of magnitude prestige scale (MPS) suggested by Wegener (1988, 1992). Taking into account the job entry and career processes of men over a long observation window allows us to distinguish between the effects of age (A), period (P) and cohort (C) on the career processes in a more detailed manner. Compared to conventional APC analyses, which often apply methodological constraints to identify APC effects (Glenn 1977; Mayer and Huinink 1990), we suggest a more substantially grounded approach in this study (Sørensen 1986). We propose to measure cohort and period effects using theoretically derived indicators (Blossfeld 1986). Utilising time series data obtained from official statistics we first identify the most relevant macroeconomic changes and their impact on the labour market by performing confirmatory factor analysis. We then apply these macro indicators in a multi-level regression model to assess the historical macro conditions under which the successive cohorts entering the labour market started their job careers in West Germany (cohort effect). Based on men’s further careers, we then examine how the continuously changing macrostructure influences the men’s upward, lateral and downward career mobility (period effect) using a multi-level transition rate model and controlling for the men’s individual resources, such as educational attainment and changing labour force experience (life-course or age effect). This fully dynamic analysis allows us to gain a better understanding of the effect of long-term macroeconomic trends on social inequality – among other trends, such as educational expansion and qualification upgrading of the labour force, as well as demographic trends and their consequences, such as change of manpower and inter-cohort competition at the time of labour market entry and across the work history (Mayer and Huinink 1990) in Germany and across several generations.

Our paper is organised as follows: based on economic and sociological labour market theories we develop hypotheses for our dynamic analysis of cohort, period, and life-course effects on career mobility. Then, we describe the data and methods used to identify cohort, period, and life-course effects. Thereafter, we present empirical evidence for the job trajectories of men in West Germany, and finally we draw conclusions about the development of opportunities at the time of entry into the labour market, career mobility and the long-term trend of social inequality across generations.

Sociological and economic theories agree that education is one of the most important individual characteristic required for all kinds of job rewards (status, prestige, income, fringe benefits, employment opportunities, upward moves, etc.). The sociological status attainment theory (Blau and Duncan 1967; Sewell and Hauser 1975) – a static approach (Blossfeld 1986, p. 209) – emphasises the role of education in getting a first job, as well as for the job observed at the respondents’ age of 30 years. This supply-side theory assumes that all positions in the labour market are freely available to anyone with the necessary qualifications. Thus, it does not involve changes in the structure of the labour market in its approach (Sørensen 1986).

The same is true for the economic human capital theory (Becker 1975; Mincer 1974), a semi-static approach, which postulates that the level of education reflects investments in human capital and the related individual productivity. Hence, well-educated people tend to get better jobs and higher incomes at the time of entry into the labour market compared to those who are not well-educated. Moreover, according to this approach, they might experience more upward moves and fewer downward moves in the later phases of their careers compared to less skilled manpower.

While an individual’s formal school or university education is regarded as relatively stable over his/her job career, the main cause of change in career opportunities is on-the-job training. Training is, however, costly and therefore is not evenly distributed in an individual’s occupational career (Mincer 1974). The human capital theory predicts that people receive training as long as their expected returns exceed their expected costs. Therefore, training is concentrated in the early phase of one’s career, when more time is available to recover training costs. Therefore, career lines are considered as concavely curved, because increasing labour force experience reduces the likelihood of receiving new training and consequent upward moves. Furthermore, labour force experience protects workers against downward job moves, even though decreasing productivity in the later phases of one’s career may result in an increased likelihood of downward mobility (Sicherman and Galor 1990).

In the available literature related to our study there are several variants of the main human capital argument. Because employers are often not certain of an individual’s productive capabilities at the time of hiring, they have to rely on education certificates as observable indicators of the individual’s productivity (Spence 1973). This means that job-person mismatches may occur, and job shifts are used to resolve them (Tuma 1985; Blossfeld 1986). Because unexpected gains owing to mismatches balance unexpected losses, on average upward and downward shifts are equally likely. However, downward shifts are more likely if individuals are over-rewarded, and the likelihood of being over-rewarded increases with increasing job rewards and decreasing individual resources. Conversely, upward shifts are more likely if individuals are under-rewarded, and the likelihood of being under-rewarded increases with decreasing job rewards and increasing individual resources (Tuma 1985).

Another variant of the human capital theory is the semi-static job competition model by Thurow (1975), which considers ideas of the institutional approach to labour market segmentation. This model rejects market imperfections as a possible explanation of career mobility and suggests that individuals compete against one another for jobs based on their relative costs of being trained. Thus, there is a labour queue for each job vacancy, and the most important characteristics that increase the likelihood of being hired are education and training. Thus, according to this theoretical model, the predicted consequences of the effects of education and labour force experience on career mobility are the same as suggested by the traditional approaches.

Overall, according to all the above sociological and economic theories, formal education is expected to have a positive effect on the status attainment and the related benefits at entry into the labour market (hypothesis 1) and on the rate of upward moves in the later phases of a person’s career (hypothesis 2). There is no specific effect of education on lateral moves (hypothesis 3). Furthermore, education protects individuals from downward moves (hypothesis 4). In addition, labour force experience has a negative effect on upward moves (hypothesis 5), because of the declining investments in further training across individuals’ job histories, and a negative effect on downward moves (hypothesis 6) and lateral moves (hypothesis 7), because accumulated (job-specific) training protects individuals from downward moves.

However, all the theories discussed so far only focus on the determinants of the supply side of the labour market and do not consider structural changes in terms of modernisation on the demand side. Therefore, it is necessary to consider these changes in a dynamic approach of career mobility (Blossfeld 1986). In the era of globalisation and tertiarisation, during the last 25 years, labour market structures and labour market conditions have become more dynamic and less predictable (Blossfeld et al. 2011; Blossfeld and Mayer 1991). Market wages and job positions have become more volatile because product markets are increasingly influenced by random shocks and contingent events in the interdependent global economy (Blossfeld et al. 2005, 2006a, b; Blossfeld and Hofmeister 2006). Due to increased global competition, there is also an increase in scientific and technological innovations in the global and national economies, which is accelerating the long-term economic modernisation trend (Blossfeld et al., 2005, 2006a, b; Blossfeld and Hofmeister 2006; Becker and Schömann 2015). Thus, less predictable economic cycles and accelerated technological modernisation are two important macro factors impacting today’s labour markets in all modern economies.

In previous studies there have been various contradicting claims regarding the effects of technological change on the occupational structure and job requirements (Treiman 1970). Industrial sociologists have claimed that new technologies and changes in work organisations are leading to a general up-gradation of skills while on the job (Blauner 1975), a universal downgrading of job requirements (Bright 1966), a polarisation of the demand for skills between different job categories (Kern and Schumann 1970, 1984), and perpetual diversification of skills in occupations (Janossy 1966). At the macro sociological level, the theory of knowledge society suggested by Bell (1973) predicts an overall increase in highly skilled job positions at the expense of a decrease in unskilled jobs; however, Braverman’s (1977) theory of deskilling forecasts a general increase in unskilled occupations in the course of the development of modern capitalist societies. If the arguments of Bright and Braverman are true, modernisation, on average, should worsen job entry opportunities (hypothesis 8) and produce increasing downward mobility (hypothesis 9), as well as decreasing upward mobility (hypothesis 10); if Blauner and Bell are right, modernisation should lead to better entry job opportunities (hypothesis 11) and more upward moves than downward moves (hypothesis 12); if Kern and Schumann are right, modernisation should lead to more upward, lateral, and downward moves (i. e. a higher overall mobility rate) (hypothesis 13). However, the upward mobility rate should be lower than the downward mobility rate (hypothesis 14), because only a few people get the opportunity to take advantage of the creation of jobs that require higher qualifications (e. g. engineers, professionals) and most workers with vocational training are forced to move down to unskilled positions. Thus, modernisation simply means more mobility, with a trend towards the down-gradation of the occupational structure. Despite these conflicting theoretical views, empirical evidence based on representative employment statistics of today’s more advanced service societies show, with an impressive regularity, that the greatest rises in non-manual employment over recent decades have not occurred in relatively low-level clerical, sales, and personal service grades but in professional, administrative, and managerial occupations, owing to the tertiarisation of the occupational structure and the expansion of public employment (Erikson and Goldthorpe 1993; Becker 2007). Further, they show that the major declines in manual employment have been in the less skilled rather than more skilled job categories (Becker and Blossfeld 1991; Blossfeld and Mayer 1991; Erikson and Goldthorpe 1993; DiPrete et al. 1997; Blossfeld et al. 2005, 2006a, b, 2011; Blossfeld and Hofmeister 2006). This empirical evidence is also true for the most recent technological developments in the field of mechanisation, automation, robotisation, and digitalisation, as demonstrated by a comparative Organisation for Economic Co-operation and Development (OECD) study (Arntz et al. 2016; see also: Autor et al. 2003; Oesch 2013). Thus, from our empirical analysis, we aim to prove that the monotonic modernisation trend is closely associated with a general up-gradation of the occupational distribution, little downgrading and an average improvement in job opportunities across cohorts.

In addition to the modernisation process, we observe the effects of regular ups and downs of business cycles on the growth in output, jobs, income and career mobility in capitalist economies. “Business cycles are a type of fluctuation found in the aggregate economic activity of nations that organise their work mainly in business enterprises; a cycle consists of expansions occurring at about the same time in many economic activities, followed by similarly general recessions, contractions and revivals which merge into the expansion phase of the next cycle; in duration, business cycles vary from more than one year to ten or twelve years” (Burns and Mitchell 1946). A critical feature of business cycles is that they are widely diffused over the entire economy at a certain historical point in time (Burns 1951). During periods of economic expansion, sales, production, employment, and incomes generally increase in the economy. During periods of boom, jobs are created, unemployment declines and real wages, as well as career opportunities, increase. Thus, we assume that an improvement in the labour market conditions is associated with better entry opportunities for job starters (hypothesis 15). Further, we expect that during periods of boom, there are more upward and fewer downward career moves (hypothesis 16).

During periods of contraction, sales decline, production slows down, unemployment rises and incomes stagnate or even decline. During periods of slowdown or recession, fewer job vacancies are available for people looking for work or promotion opportunities. Young people may choose to wait until a more suitable job becomes available or accept a job whose status (including benefits such as prestige or wages) is lower than initially expected. Given these two alternatives, most young people are expected to choose the latter alternative and accept a job of lower quality than initially expected (hypothesis 17). Obviously, economic contraction should also lead to fewer upward job moves and is expected to force people to move in the downward direction (hypothesis 18). A dynamic approach to men’s job mobility and status attainment should, therefore, consider the impact of both the continuous path of modernisation and the cyclical development of labour market conditions on an individual’s entry into the labour market and career mobility.

Finally, let us consider the effects of the changing labour market structure (modernisation) on job opportunities across generations. Successive generations start their careers in different structural contexts. The main features of these contexts can be described by the historic level of modernisation and labour market conditions (the changing state of the business cycle). We assume that the modernisation process is closely linked to new job requirements, which differ from the respective structure of people’s qualifications at a given time (Janossy 1966). Thus, in periods of accelerated modernisation, the gap between the structure of qualifications of the labour force and the structure of skill demands of jobs tends to widen quickly. A continuous matching of both structures may be achieved through two basic adjustments. First, people in the labour force may acquire new qualifications through further education, lifelong learning, or on-the-job training (Blossfeld et al. 2014; Becker and Schömann 2015). Second, new labour market entrants, who have a more up-to-date and relatively high educational attainment, may fill the newly created and more demanding jobs, and at the same time obsolete jobs may disappear when the older generations retire from work (generational replacement). In reality, both adjustments occur simultaneously. However, according to Janossy (1966), re-training of employees is limited by their age because the return on investments in training steeply diminishes with age for workers and companies alike. In other words, as the modernisation process advances, younger cohorts of entrants start their careers at higher average levels of job attainment (hypothesis 19). Because qualification is a key factor, it is particularly the qualified and highly qualified young labour market entrants who are expected to get access to the newly created skilled jobs (Blossfeld et al. 2005), whereas the unskilled job starters are penalised by the modernisation process (hypothesis 20). Overall, the mechanism of generational replacement in the occupational structure is supported by the macro processes of educational expansion and the up-gradation and tertiarisation of job requirements across cohorts. The generational replacement of less qualified older workers by more qualified younger workers is a particularly effective mechanism of occupational change, especially in times of globalisation and tertiarisation of economic sectors.

However, it is important to address how the experience of cohorts at their time of entry into the labour market influences their subsequent career trajectories (Blossfeld 1985). In a pyramidal job inequality structure (Sørensen 1975, 1983; Blau 1977), there is a ceiling effect, and thus a very favourable placement in the first job position makes it harder to find an even better position. Hence, we expect a decreasing rate of upward mobility (hypothesis 21) and an increasing rate of downward mobility (hypothesis 22) the more prestigious the job is at the time of entry into the labour market.

We conducted empirical analysis and assessment of the hypotheses in three steps. First, we described the career profiles of men from successive birth cohorts and compared how these profiles changed in the process of socio-economic change in Germany from 1945 to 2008. Second, we focused on the entry of men into the labour market and examined how their first jobs were influenced by the modernisation process and the business cycle. Third, for the historical period 1945–2008 in West Germany, we conducted a multi-level event-history analysis of the career processes after entry into the labour market.

4.1 Changing career profiles of successive birth cohorts

The theoretical section emphasises how increasing skill requirements of jobs in the course of modernisation, and their correlation with changes in occupational structures (tertiarisation) and the decreasing number of unskilled positions, have altered the career profiles of each successive younger cohort of men. In Germany, this process of economic modernisation took place in parallel with continuous educational expansion (up-gradation of qualifications) across cohorts. If we first look at the average prestige scores of the first jobs across birth cohorts, it is obvious that the mean occupational status has more or less continuously increased for each successive birth cohort (Fig. 2). This finding supports hypothesis 11 claiming that the younger cohorts of entrants start their careers at higher average levels of job attainment compared to the previous or older cohorts as the modernisation process advances.

This pattern of cohort differentiation also continues in the later stages of men’s age-specific career profiles (Fig. 2 ). Although the youngest cohorts’ average prestige levels became increasingly similar in later phases of their job careers, the early age-specific average career profiles tended to be increasingly steeper. Thus, compared to those of previous cohorts, each younger cohort reached a higher average prestige level at a younger age. At this point of our analysis it has to be emphasised that, for well-educated graduates and occupational beginners in particular, the continuing strong institutional linkage between the educational system and the labour market in Germany is an additional explanation for this trend of better job opportunities in terms of average status (Müller and Shavit 1998; Mayer et al. 2010). Our job entry and job career analysis separates the effects of individual educational attainment and macro-structural factors.

Following research by Weber (1922) and Coleman (1986), the aim of this study was to contribute to a better understanding of how the long-term modernisation process and the ups and downs of the business cycle in the course of modernisation have affected the entry of men into the labour market in West Germany, and their career mobility, since the end of World War II (Mayer 2004). To achieve this, we reconstructed the individual job histories of men from successive birth cohorts in an observation window of more than 50 years and linked them with the most important macro changes in the labour market. Our longitudinal analysis shows that birth cohorts are one of the most important ‘cultural carriers’ of change in the labour market. Furthermore, our dynamic multi-level analysis provides a better understanding of how long-term macroeconomic trends have influenced processes of career mobility and social inequality in West Germany across several generations.

Utilising two longitudinal event-history datasets (from the GLHS and the ALWA study), we linked the careers of individual men with time series describing long-term developments at the macro level. In our analysis, we applied a substantial approach to the identification of life-course, period and cohort effects, and found that the modernisation trend and changing labour market conditions have a strong impact on both the men’s job entry into the labour market and their further career trajectories. Our results show that generational replacement is a particularly important mechanism for the modernisation of the labour market. New labour market entrants who have a more up-to-date and relatively high level of educational attainment tend to match the requirements of newly created jobs, which are more prestigious and demanding. Thus, as the modernisation process advances, younger cohorts of entrants start their careers at increasingly higher average levels of job attainment. Because upgraded qualification is the key to getting these jobs, it is particularly the highly qualified young labour market entrants who are the winners in the modernisation process, while the unskilled school leavers are penalised by the modernisation process. Thus, one conclusion of our analysis is that the educational system must reduce the proportion of school leavers who leave without any educational certificate or further vocational training.

Furthermore, we demonstrated that the progress of modernisation is an important ‘vehicle’ for status attainment. However, modernisation not only engenders upward job opportunities, as supposed by theorists of the knowledge society (Bell 1973), it also leads to increasing lateral and downward moves. Thus, it seems that the polarisation thesis of Kern and Schumann (1970, 1984) describes the period effect of the modernisation process quite adequately. Some men, especially the more qualified ones, can take advantage of the creation of better jobs (engineers, professionals) during their careers, while others are forced to move down to less prestigious job positions during their careers. Therefore, we conclude that the idea of a linear trend of increased upward mobility in the modernisation process is as misleading as the claims of monotonic trends towards individualisation, flexibilisation, destandardisation, and destabilisation of careers (Mayer et al. 2010; Brückner and Mayer 2005; Mayer and Blossfeld 1990). They are misleading because they all treat occupational careers as holistic entities, which seem to be independent from individuals’ scopes and structural life-course conditions in historical periods and for successive birth cohorts (Mayer 2004). Overall, we found that highly qualified men adapted to the newly created jobs quite easily and gained from these societal changes (Becker and Hadjar 2015). Based on our results, it is evident that the claim of a descending middle class after the period of long-term prosperity in the 1960s and 1970s is not supported. In particular, if we look at the career profiles of successive birth cohorts of men since World War II (Fig. 2), there is a clear improvement in men’s average prestige at labour market entry with an increasingly steeper career profile. If we include women in our analysis, we expect quite a positive picture across cohorts because women have not only outperformed men in the educational system in Germany but could also gain from their better qualifications at the time of entry into the labour market (Blossfeld et al. 2015). However, the scenario of labour market entry and further stages of the career may be different for migrants and East Germans.

Investments in human capital – that is, skills and credentials – still remains one of the most important decisions in the life-course of individuals in regard to being prepared for work and life in the fast-changing world. Furthermore, it must be noted that the value of their human capital depends on the dynamic linkages between micro and macro levels of work organisation, stratification, and mobility in work history (Haveman and Cohen 1997, p. 108). Finally, we must admit that we could not control for the corporate action of firms and the ecological dynamics of firm populations at the meso level (Hannan 1988; Windzio 2001). This is also true for the size of the firm employing the respondents and having an impact on their career (e. g. Baron and Bielby 1980; Carroll and Mayer 1986; Mayer and Carroll 1987; Spilerman 1977), since we have no time-varying information at all on the “life-course” of the firm – i. e. on the growth and shrinkage of their number of employees (Wachter and Bender 2006; Hannan and Freeman 1989). Furthermore, it must be noted that the datasets taken into account do not comprise information on the individuals’ preferences and aspirations affected by economic cycles, modernisation, and labour market conditions. By controlling for resources and measuring social action it is possible to take into account the individuals’ preferences and decisions.