INTRAGENERATIONAL MOBILITY IN SUCCESSOR STATES OF THE USSR Open Access

Social mobility is widely studied in sociology. Intergenerational mobility is one of the major topics after Pitirim Sorokin worked out its main theoretical framework and opportunities emerged for analyses of the massive data sources. Social mobility is also studied in transitional societies. There the attention has been focused on the issue of how much the state socialist elite was able to facilitate its reproduction through the next generation under the new emerging market society. It is not surprising that intragenerational mobility is not so widely studied in transitional societies, as the time lag is not sufficient for the studies mentioned above. In post-Soviet countries, current independent states emerged as a result of the collapse of the Soviet Union. We are lucky to have data both from the Soviet era and the time after the collapse of the USSR. It allows us to look on intragenerational mobility in the first years of transition in five countries.

In the current article three problems are examined. Firstly, how large was intragenerational mobility during the first half of the 1990s for the cohort of the young adults best located to take advantage of opening market opportunities by their age. When the Soviet Union collapsed, this age cohort was in their late twenties: they had finished their studies and had already had a couple of years of working life and were ready to take risks in the emerging privately organized work-life. Secondly, how much impact had parental social stratum on the intragenerational mobility. thirdly, what factors increased and decreased chances in intragenerational mobility. We use data from the ‘Paths of a Generation’ project on Belarus, Estonia, Latvia, Russia and Ukraine.

Sorokin (1927) in his book Social Mobility made a major distinction between intergenerational and intragenerational mobility, but empirical studies of the mobility started only when after WWII computers provided an opportunity to analyse large data sets. For a long time intergenerational mobility has been a major research topic in mobility studies. In the early research, issues about the extent of mobility dominated. Intragenerational mobility became a topic of attention when age started to be a significant factor in intergenerational studies (Riley 1987; Warren et al. 2002). The problem of the stage of the occupational career as a factor in both generations was recognized as an important issue in intergenerational mobility. Empirically, a classic study of intragenerational mobility can be linked with the availability of data about the same age cohorts in different times. Panel studies enabled the collection of such data, and serious studies were undertaken. Those studies followed very much the same logic as classic intergenerational studies by Blau and Duncan (1967) and Featherman and Hauser (1978). Now cohort comparison of intragenerational mobility on the bases of census data or panels is widespread (Mayer 1986; Muller et al.1988).

Our study is linked with another theoretical paradigm developed on the bases of longitudinal research. Occupational career was widely studied by social scientists before these kinds of data were used for intragenerational mobility analyses. Longitudinal data allowed not only to study patterns of life career, but also to establish clear differentiations of careers. This differentiation corresponded first of all to the white- and blue-collar jobs in the United States. In the European tradition, class analyses were used to group occupations into more detailed units. Besides general patterns of life and occupational career, researchers started to pay attention more specifically to the factors influencing life career. Coleman (1966) included social environment in his very broad study of educational career. In occupational career research, major interest turned towards explaining the factors facilitating or constraining career (Sewell et al. 1969; Sorensen and Kalleberg 1981; Davis 1982). This is also the aim of our analyses.

Those studies were primarily done on Western societies and these theoretical approaches were used as a theoretical framework for analyses of state socialist societies (Zagorski et al.1984; Slomczynski and Krauze 1987; Simkus 1996), and classifications were also adapted from studies carried out in market-based Western societies on local occupational structure. As classifications in American social stratification research were based on the division of labour in occupational structure, it was very difficult to see crucial difference between a market society and a state socialist society. For instance, Kornai's studies (1980) explored the basic economic differences between a Communist and a free market economy.

State socialist society was first characterized by political logic, not only in the economy, but also in all spheres of society. It was especially the case in the outer part of the Soviet empire. In most of those countries the major problem was political loyalty to the regime. Correspondingly, state socialist society tried to control all the processes, especially the formation of and mobility into politically important strata. Political loyalty first and professional qualification second, were the principles of policy and the bases for the promotion of people into the upper echelons of society. In Eastern Europe, the constraints and advantages in educational careers, and especially access to university, were quite open to everybody, but on the elite level, as a rule, the proposal for promotion came from the party–state apparatus. Although career-oriented people were also active, it was still secondary compared to the selection of candidates of party members and elite in general.

In the Soviet Union, ‘class enemies’ were eliminated before World War II and political loyalty did not play such an important role after the war. The so-called Khrushchev ‘thaw’ even made university education obligatory for promotion in the party apparatus. In this sense, intergenerational mobility in the Soviet Union was determined by the person's education and not by the parents’ social status (Titma et al. 2003). In a heavily tracked educational system, education directly prescribes the social stratum to which a person enters after graduation. Later mobility was supposed to be party–state sponsored as an ideal. In practice, the system did not work entirely this way as the social fabric of society had a huge independent impact.

A major difference from Western market societies was state-sponsored mobility. As state socialist economies were deficit economies, manual labour and lower non-manual jobs were in high demand. Correspondingly, mobility in those strata was very much like that in a market society. Managerial and professional strata were under control of the party–state apparatus and mobility was state-sponsored. Of course, a state-run society functions with large deviations from supposed patterns. Differences between state socialist countries were considerable, and intragenerational mobility had country peculiarity. Eastern Germany certainly had very tight political control over educational and occupational career. The situation in Lithuania was under the control of Lithuanians, and the local elite used the system of elite schools to reproduce itself. In Latvia, Russians were in a majority in all cities and the Russification process was very strong in the educational system and party–state sponsored careers on the top levels were controlled by Moscow. To study intragenerational mobility in Georgia without inclusion of the Soviet Union (nowadays Russia) would not be representative, as larger part of the elite always had positions in Moscow or other major Russian cities (of a nation of five million people, one-fifth made career outside of Georgia, not as immigrants but temporarily for career advances).

Essential starting points for analyses of social mobility in transitional studies are Soviet time patterns and factors predicting social mobility. Transitional societies are overcoming state-sponsored mobility, opening it to individual actions under market competition. There are multiple ways of establishing a market society, but previous socio-economic development had a great impact on transitional societies. For the following analysis, it is important to investigate the basic differences in the transition process itself. All satellite countries of the Soviet Union were independent states, at least after World War I, and did not need to build up the state itself in the transition period. With the exception of Russia and the Baltic countries, all other former Soviet republics were quite new formations. It is extremely difficult to lay foundation to a state with no previous statehood experience, and quite often this led to personal rule supported by the Soviet era party apparatus.

Another theoretically very important process for analysis of intragenerational mobility was the emergence of private property and the privatisation of state economy. One opportunity was to copy the historical process of the emergence of the market economy. This meant opening opportunities to everybody, and enterprising people tried to establish businesses. This approach was in particular followed in Poland, Estonia and Latvia. In these countries, the bureaucracy left by the old regime was politically marginalized and unable to take control over the privatisation process. It was very easy to open a new business in the first years after the formation of the non-communist governments. This created a mass of small businesses and entrepreneurs, who seized an opportunity for setting up a market economy. We call this type of emergence of private property ‘bottom up’ privatisation.

In sharp contrast to this, most former Soviet republics were exposed to the ‘top down’ privatisation of state property. Opening businesses was controlled by the old bureaucracy and massive efforts were needed to overcome all kind of regulations and bribes. Although oligarchic economy did not emerge everywhere, the fact that private businesses emerged not from the participation of ordinary people but through social and political capital was more important. We suppose that social mobility differs considerably as a result of these two different approaches to the implementation of market reforms.

Strategies for building up market economies not only opened new opportunities in the job market, but also considerably influenced the mechanism which distributed these new opportunities. For our analyses of intragenerational mobility, the lineage with intergenerational mobility is also an important aspect. We assume that a ‘bottom up’ transition significantly weakened the age-based distribution of occupational careers and opened social niches, usually available to more mature age groups, to younger age cohorts. At the same time, a ‘top down’ build-up of market economy may facilitate social reproduction and restrict the intragenerational mobility of younger cohorts.

In transitional societies the destabilization of the state (not to speak of building a new state) and economy re-evaluates the opportunities and responsibilities of social actors in society. First of all, this destabilization devalues the established occupational careers and over-evaluates risk-taking and self-efficacy (Pavalko 1997). Market opportunities open in mass and the state-run economy vanishes relatively quickly. It is well known that younger people are more ready than older people to take risks in life careers. As many studies have shown, taking risks is more typical for men than for women. In transitional societies, women can also be hurt by previous traditions of undervaluing their human capital, including education (Janicka 1995). In many transitional societies, including the former East Germany and the European part of the Soviet Union, women were more educated than men in the late 1980s (Janicka 1995; Titma and Tuma 2001). As studies have shown, gender segregation is widespread in many labour markets (Rosenfeld 1983; Bielby and Baron 1986; Tienda et al. 1987; England et al.1988). The trend of segregation is similar: manager and manual labour positions are mostly occupied by men, intellectual labour and service positions by women (Charles 1992; Charles and Grusky 1995; Petersen and Morgan 1995; Warren et al. 2002). The measure of segregation varies greatly by country (Nermo 2000). A socialist state economy produced higher gender segregation than a market economy. There may not only be occupational shifts but also changes in the gender composition of the strata.

A process of intragenerational mobility in the transitional society can be less dependent from obtained education than in the Soviet period. A market society re-establishes the parents’ role and choices in the educational process. An open labour market raises the number of available choices and the increases the role of parents in guiding the next generation. This means that the parents’ education and position have an influence on their offspring's education rather than their career (Warren et al. 2002). The likelihood of some growth of social reproduction during the transitional period is high, as state-sponsored mobility vanishes.

For our analyses we used longitudinal data from the ‘Paths of a Generation’ project. This allows us to look at intragenerational mobility in the early years of the transition from 1991 until 1998. For this analysis, we are using data from Belarus, Estonia, Latvia, Russia and Ukraine. In studying intragenerational mobility it is useful to apply comparative analyses. The comparison of the mentioned countries enables to study to what degree different approaches of building up a market society influence intragenerational mobility. The cases of Estonia, Poland and East Germany, which all had the experience of transition from socialism to a market economy, ensure that structural changes in economy caused an increase in mobility (Domanski 1997, 1998; Titma et al. 1998; Mayer et al. 1999).

All post-Soviet states experienced considerable economic losses during the first years of transition. However, Estonia and Latvia moved rapidly towards a market economy and in 1994 the economic growth started. Entrepreneurial opportunities were available for everybody, and in 1997 the private sector had a dominant role in the economy. We expect that young adults in the Baltic countries had considerably wider upward mobility than in other countries under consideration. In Russia and Ukraine, the economy was still shrinking in 1998. Privatisation from the top left people out and created an oligarchic type of economy. Belarus tried to go ahead with a command economy and was very slowly opening doors to private business even in 1998.

Our first hypothesis states that new market opportunities created by ‘bottom up’ transition increase more upward intragenerational mobility among young adults than ‘top town’ transition or a slow move towards a market society.

Intragenerational mobility usually depends on opening vacancies, but in a transitional society new jobs are the widest base for intragenerational mobility. Official data show that with a market economy, a structural change of labour opportunities in different industries also comes (Domanski 1998; Eamets and Philips 2000). Another reasoning of the importance of industry comes from the Soviet past. Industry was the basic structural unit of the command economy. With the collapse of the command economy, many previously attractive industries lost their importance. So, we assume that industry can be an important factor for raising the chances of intragenerational mobility.

The second hypothesis is that the change in the industrial structure of economy is an important factor in intragenerational mobility in the early stage of transition.

As many studies have shown, social background affects not only early years and educational career, but also later life. Many authors (Rona-Tas 1994, Szelenyi and Szelenyi 1995; Eyal et al. 1998) have found that certain groups of elite were able to pass their social position to the next generation in transitional societies. Parental social capital can also be transferred to the next generation. All this can justify raising a hypothesis that intragenerational mobility in a transitional society can be influenced by parental social position. However, as mentioned above, a major part of the intergenerational reproduction for our respondents took place during the Soviet era (educational and early occupational career). We still believe that in a transitional society the parents’ generation has an impact on intragenerational mobility.

The third hypothesis states that parents’ social strata predict intragenerational upward mobility.

The Soviet Union, as well as the other socialist countries, had deep gender segregation in their labour force, which also reflected the gender composition of the social strata. Western market economies had a more balanced distribution of males and females in the social strata. Transitional countries came from the Soviet type of gender segregation, where managerial positions and manual labour was far more typical for men and intellectual work for women. Labour market development makes corrections to a Soviet type of gender segregation and produces gender flows in such vastly different directions than were previously quite unknown. The early stage of transition favoured men who were more ready to take risks. Rough rules of privatisation also gave the advantage to men, especially among young adults. At the same time, women were more educated and in the long run should have gained from it.

The fourth hypothesis states that gender strongly predicts intragenerational mobility in the early stage of transition.

At first, transitional societies suffered setbacks and losses in the standard of living of the majority of the population. The significance of monetary rewards rose hugely at this period of transition. Studies show that salaries were the most quickly changing economic indices in the beginning of transition. It was a considerable factor for many people to change jobs. The question of whether income was also a factor that influenced intragenerational mobility also rises. According to Domanski (1995) in Eastern Europe there are two new elements which might influence mobility: growing income inequality and new rules of income distribution. Domanski (1994) also notes that the average income of the intelligentsia went up and that of manual workers and farmers clearly declined. To test this we used income at the beginning of the transition as an independent variable to predict intragenerational mobility.

The fifth hypothesis states that income predicts intragenerational mobility.

Data for this article come from the longitudinal study ‘Paths of a Generation’ (PG). The project began as a study of life careers of the cohort of 1983–1985 (wave 1) graduates of secondary schools in 15 regions of the USSR (Titma and Tuma 1995). The same respondents were re-surveyed in 1987–1989 (wave 2), again in 1992–1994 (wave 3), and yet again in 1997–1999 (wave 4). The respondents were typically 17–18 years old and about to graduate from secondary school when the study began, 26–28 when the societal transformations began in 1991, and 32–34 during the most recent survey in 1997–1999. Although, there were 15 regions in the survey, in the present analysis we use five of them to compare three areas with different features. The first area is western- oriented and developed Estonia and Latvia, where the changes have been rapid. The second is Belarus with a functioning command economy. The third consists of two industrial centres with a collapsed economy in the former Soviet heartland, Kharkov in Ukraine and Sverdlovsk in Russia.

Occupational origin (1991, wave 3) and destination (1997–1999, wave 4) are defined on the bases of ISCO. The occupations of respondents are converted from their original codes in the three-digit version of ISCO 88 (ILO 1988) into a modified version of the class scheme of Erikson and Goldthorpe (1993). To consider the specifics of the PG sample, the top of the E-G scheme is disaggregated and the bottom is compressed to give more detail on the occupational groups of the PG respondents. The occupational groups of manual work come directly from the five-class E-G: scheme: skilled workers (V + VI), unskilled workers (VIIa), and agricultural workers (VIIb). A routine non-manual (III) group comes directly from the seven-class E-G scheme. Higher and lower professionals are also represented in the full E-G scheme. As in Titma et al. (2003), in the present study the top class in the E-G scheme (I) is divided into managers (Ia) and professionals (Ib). Because of the different political influences of professionals and managers during the Soviet era, this change reflects more vividly the social realities of the studied countries. Entrepreneurs are a new stratum, which arose after the soviet era (Domanski 1998, 1997) and is added to the managers’ category in 1998.

If respondent's current occupation in 1991 is missing it is replaced with the most recently held occupation. The distribution of occupational groups is quite similar in different regions. In both waves 3 and 4, professionals and skilled workers are the most numerous categories (between one-fifth and a quarter of the sample); the least represented categories are unskilled and agricultural workers in 1998 and managers in 1991.

In multivariate models, occupation in 1998 is used as a dependent variable and is recoded as described above. Independent variables in models are occupation in 1991, gender, education, father's education, father's occupation, parent decision-making position, income and industry. Occupation in 1991 is inserted into analysis by the modified Erikson–Goldthorpe schema described above, the same way as the dependent variable.

Education, measured in 1991 was used in most regions. In the case of Belarus, the variable was based on the third wave education, but fourth wave data substituted it where education in the third wave was missing to avoid unduly decreasing the sample size. In the data education was fixed by stages but we used in analyses the number of years a full-time student typically needed to complete his or her education. In this sample, secondary education was the most typical educational category (a quarter to a third of respondents) by size; the next group was higher education (a fifth to a quarter). As it can be seen, in our sample there is the more educated segment of this generation. Father's education in 1991 is also measured in levels and recoded into years the same way as respondent's education. The largest number of respondents’ fathers had 8 years of education.

Father's occupation in 1991 is fixed in the ISCO scale and transformed into five categories: managers, professionals, routine non-manual workers, workers and peasants. A half of the fathers were workers, nearly a fifth agricultural workers, both professionals and routine non-manual workers made up over a tenth and the smallest group was managers – under a tenth. The concept of the ‘nomenclature’ is operationalized by whether a parent has ever held a decision-making position in a public institution: the communist party or a state organization (Titma et al. 2003). In our data 5 percent of respondents reported that at least one parent had ever been a ‘public decision-maker’.

Income in 1991 was changed from national and other currencies into dollars. As the economic situation was different in the observed countries, the incomes differ greatly. The incomes are the biggest in Estonia and the least in Belarus. As income was a significant factor in multivariate models but the change by one dollar was too small to be visible, we transformed the income into tens of dollars to make the difference visible.

Economic sectors or industries are also included in this analysis. Agriculture, horticulture, forestry and fishing are bundled in this study as agriculture. Power engineering, water supply, mining and processing of oil shale and minerals, machine building, production of building materials and construction are coded as industry and construction. Light and food industry, service industry, commerce, catering, transportation, communications are assembled in the category service industry. Information, statistics, banking, finances, insurance, education, cultural affairs, medicine, science, art, public defence and state security, state institutions, social organizations, political parties make up the sector of state and social institutions. State and social institutions is the sector which embraces the biggest number of respondents, over a quarter in Estonia and Latvia and over a third in Russia and Ukraine. The smallest share of respondents belongs to the category of industry and construction in Estonia and Latvia and in agriculture in Belarus, Russia and Ukraine.

First we give an overview of mobility patterns in the chosen regions between 1991 and 1998. We start with the outflow mobility for a 7-year interval. As background, outflow mobility is referred by region.

As can be seen from Table 1, the proportion of managers in 1998 was the highest in Estonia and Latvia (13%) and the smallest in Belarus (7%). This means that young adults were twice as likely to move to the position of manager in a ‘bottom up’ privatisation than in a stable command economy. The share of professionals was the highest among the Belorussian respondents but there were no great difference compared to other countries. At same time, Belarus had substantially more semi-professionals than other countries. As Kharkov and Sverdlovsk are two industrial centres of Russia and Ukraine, it is not surprising that the proportion of workers there was larger than in the Baltic. In Belarus, the share of rural people according to census data was far higher and the small number of peasants among the respondents is the result of attrition from the original sample.

Figures in Table 1 suggest that it is possible to identify the occupational groups which were most likely to be stable through the first years of the transition. In Belarus, the proportion of those respondents which maintained occupation from 1991 to 1998 is higher than in other regions. In Estonia, Latvia, Russia and Ukraine there is the same extent of stability among young adults than in East Germany among the whole population (Mayer et al. 1999). The stability was highest among managers and professionals in the Belarus. Estonia and Latvia had lower level of stability, as the number of managers almost doubled in the seven years of the transition. New strata of private entrepreneurs emerged. In general, outflow mobility was the highest in the Baltic but closely followed in Russia and Ukraine with ‘top down’ privatisation. Due to outflow, Belarus had a much more stable young adult population in all occupational groups.

Inflow mobility is interesting as it gives a picture of mobility patterns with regard to certain occupational groups.

The young adults who were managers in 1998 in Russia and Ukraine had the largest variety of backgrounds in terms of social strata, but mobility to the managers’ stratum is the highest in the Baltics, excluding the case of professionals (Table 2). Of course, professionals with university education are a 75% stable contingent. If we look at semi-professionals, we can see a surprising distinction in Belarus, where 90% remained in the same group for seven years compared to a much less stable situation in other countries. This effect can be explained by the still functioning command economy that facilitated life-long careers for women in lower white-collar strata. This conclusion is supported by data from the routine non-manual occupational group, which is 70% stable in the Belarus. The most mobile were unskilled workers in every region.

The results on mobility are presented more generally through gross mobility rates. Table 3 gives a better overview of mobility.

It is possible to see in Table 3 that, in the Baltic countries Russia and Ukraine, about two-thirds of the respondents retained their occupational group. In Belarus, three-quarters of respondents had not changed their occupational group between 1991 and 1998. We can conclude that ‘bottom up’ and ‘top down’ shock therapies initiated more mobility among young adults as the still functioning command economy allowed. Mobility trends in the three chosen regions of the former Soviet Union differ considerably. Those trends reflect changes in upward and downward mobility percentages for strata. Though upward mobility exceeded downward mobility in every region, in Belarus it was much lower compared to other regions. Belarus differed even more radically by downward mobility from Estonia, Latvia, Russia and Ukraine. Considering upward mobility in particular, it can be seen in Table 2 that, in Estonia and Latvia, Russia and Ukraine, about a fifth of respondents were upwardly mobile, in Belarus this proportion was 15 percent. Regarding downward mobility, in Table 3 it is clear that 9 percent of Belorussians moved downward, while in the other states the respective indicator was 14 percent. Predictably, a command economy also produced greater stability of respondents in the strata that they entered before the transition. Gender was a very important factor in intragenerational mobility, as can be seen in Table 3. young adult men were more upwardly mobile in all regions. Women of the same age were more downwardly mobile than men, Belarus being an exception. Mayer et al. (1999) reported the same results on the East German population. Stability is greater among women. They had been more likely than men in two Baltic countries to stay in the occupational group which they entered. Belarus even had a higher number of women who remained in the same occupational stratum as in 1991. Upward mobility was highest among Estonian men (28%) and lowest among Belorussian women (11%). The gender difference was largest in the Baltic countries (11%) and smallest in Russia and Ukraine (4%). This shows that Estonia and Latvia are moving towards Central European-like developments where the difference between men and women in upward mobility is increasing compared to the Soviet era (Mayer et al. 1999). The most downwardly-mobile respondents were women in Russia and Ukraine (16%), while Belorussian women were the least likely ones to move downward (8%).

Multivariate models allow us to study which factors are facilitating mobility. The results of a multinomial logistic regression analyses by the strategy of the transition are presented in two models. This basic analysis varies very little in the context of testing the significance of strata in 1991 to dependent variable strata in 1998, but the change of other independent variables as predictors of intragenerational mobility is important. To avoid the problem of too many independent variables in the analyses, we used two separate models. Appendix 1 presents both models of intragenerational mobility. The first model includes 1991 strata with the impact of father's stratum, gender and education. The second includes 1991 strata with father's education, parent decision-making position, income and industry.¹

Mobility to certain occupational groups is described on the bases of both models for three different scenarios of transition in five countries. As the hypotheses are based on certain independent variables, the logic of discussion follows variables presented through both models. We start with managers as the stratum which, in the Baltics, was enlarged most.

As we can see, in the course of a ‘bottom up’ development of the market economy in Latvia and Estonia, the odds are very significant for those being managers in 1991 to be managers in 1998 (B=4.22 in Model I; B=2.99 in Model II). In the other two types of transitional societies, the position of a manager in 1991 did not have any significance on being a manager in 1998. However, odds were high for young adults to stay in their managerial position (B=6.19 in Model 1; B=5.14) in Russia and Ukraine. In all three regions, the odds are low against being a manager if the starting point in 1991 was routine non-manual. It is only in the two Baltic countries where the respondents who were professionals in 1991 were at least twice (B=0.92 in Model I; B=1.38 in Model II) more likely to become managers in 1998 than those who had a lower position in 1991.

Gender is a very significant predictor of managerial position in the Baltics, also significant in Russia and Ukraine and not significant but highly predicting in Belarus. in the Baltics, education was a very significant predictor (B=0.25) in ‘bottom up’ transition. In other strategies of transition, it was not significant. Father's education is as significant predictor of intragenerational mobility (B=0.08) in the Baltics, as it was in other regions. Education was a significant factor in reducing odds of being a manager in the Soviet era (Titma et al. 2003), but had no importance in the first seven years of transition in intragenerational mobility. This means that educated father's negative effect in promotion into a managers’ position has now vanished. Higher income in 1991 is strongly significant and raises odds to be a manager in a ‘bottom up’ strategy of transition. ‘Bottom up’ strategy opened wider opportunities of being a manager, even after the seven years of transition. It is also obvious that respondents working in industry and construction in the Baltics in 1991 had higher odds of being managers in 1998 compared to people working in state and social organisations.

To conclude our description of factors predicting mobility to the manager position after the seven years of transition, we can say that the ‘bottom up’ strategy creates opportunities to be a manager. Those who are mobile to a manager's position in the context of a ‘bottom up’ strategy are well predicted by independent variables.

Professionals in 1991 have predictably high odds to be professionals after 7 years (in Belarus B=2.56, significant only in Model I; Russia and Ukraine 2.45 in Model I; 2.20 in Model II; in Estonia and Latvia 2.05 in Model I; 1.29 in Model II). Routine non-manual position holders in 1991 have very low probability to be professionals in 1998 in all countries, as by definition a professional is a person with a university diploma. The other feminine stratum, semi-professionals, has lower odds to be professionals in the Baltics, but not in other regions. in the Baltics, odds are very low also for young adults in blue collar positions to be mobile to professional strata in the seven years of transition. This raises the question of whether there are some clear borders between professionals and lower strata during the first years of transition. In other regions, there are no significant odds between professionals and other occupational groups. Being a manager in 1991 is not a significant predictor for professionals. This means that, in the Baltics, professionals can rise to a managerial position, but no relation can be established for the transition from manager to professional strata.

There are very high odds between education and professionals in the Baltics (B=0.34) and significant odds in the Russia and Ukraine (B=0.27). Gender is also a significant factor of becoming a professional in the Baltics. Although two-thirds of professionals are women and one-third is men in the Baltics, there are higher odds for men to be professionals in 1998. This is because the referent category, routine non-manual worker, is even more feminine in its composition than the professionals category. Father's education is a highly significant predictor of professional position in the Baltics, as it was in the Soviet era for intergenerational mobility. It was also a significant predictor in other regions in intergenerational mobility to the professionals category (Titma et al. 2003), but is not a predictor in intragenerational mobility to the ranks of the professionals during transition. Higher income also significantly increases the chances of staying in a professional position over the seven years of transition in the Baltics (B=0.04).

Being a semi-professional in 1991 dramatically increases the chances of finding oneself in the same stratum in 1998 in all regions (in the Baltics B=3.90 in Model I; 2.32 in Model II; in Belarus 4.83 in Model I, 4.30 in Model II; in Russia and Ukraine 3.76 in Model I, 2.87 in Model II). As in the case of the other occupations in 1998, routine non-manual workers also have the lowest significant probability of being semi-professionals in 1998. Skilled workers have a significant possibility for upward mobility to the semi-professional stratum in the Baltics (B=1.83). Though semi-professional category is a highly femin ine one, routine non-manual category is even more feminine and, as the latter is the referent category for men in the Baltics, there are significantly higher odds to be semi-professionals (B=1.07). If the respondent's parent has ever been in a decision-making position, it ensures him/her the position of semi-professional in the Baltics (B=1.30). Semi-professionals seem to be the only stratum that is influenced by a parent decision making position.Therefore, having a parent in a decision-making position did not guarantee a high position, but it did ensure a non-manual job. Semi-professionals also have low odds to work in the service industry in Estonia, Latvia (B= − 0.73) and Russia, Ukraine (B= − 1.05).

Routine non-manual workers are set as referent category in this analysis. It seems that the routine non-manual stratum is quite a particular one. As it is the most feminine category, men had higher probabilities to be in every other position in 1998. Routine non-manual workers had high odds to stay in their category during the observed period in all regions.

Skilled workers have extremely good odds of gaining their position over the seven years of transition (in the Baltics B=4.03 in Model I, 2.66 in Model II; in Belarus 3.72 in Model I, 4.00 in Model II; in Russia, Ukraine 2.51 in Model I, 2.86 in Model II). Again, routine non-manual workers have the least chances of becoming skilled workers. Men have more possibility of becoming skilled workers in the Baltics (B=1.53) and in Russia and Ukraine (B=1.36). There is a phenomenon that the higher the father's education, the lower the odds that respondents have to find themselves in the position of skilled worker only in Estonia and Latvia (B= − 0.07). However, higher income in 1991 increases the odds of being a skilled worker in 1998 (B=0.05) in the Baltics. Logically, the position of skilled worker is also influenced by the economic sector in 1991. Those working in agriculture or industry and construction have a significant probability to be skilled workers in 1998.

In Estonia, Latvia and also in Russia and Ukraine, unskilled workers have high odds to stay in their position (in the Baltics B=2.97 in Model I, 2.02 in Model II; in Russia and Ukraine 6.23 in Model I, 5.79 in Model II). Those who were peasants in 1991 have a significant probability (B=1.17) of finding themselves in the stratum of unskilled workers in the Baltics. Our data show that skilled workers have certain ways of upward and downward mobility in the Baltics. As mentioned before, they had significant mobility to the stratum of semi-professionals and, in terms of downward mobility, they had significant probability (B=1.55) to find themselves in the stratum of unskilled workers in 1998. Routine non-manual workers have lower odds to find themselves in the category of unskilled workers in Estonia, Latvia and Belarus. Gender is also a significant predictor of unskilled workers’ stratum in the Baltic Men have higher odds to be unskilled workers in 1998. The higher the income, the higher the probability for unskilled workers to retain their position (B = 0.04) in Estonia and Latvia. Industry has also an important role of predicting position in the category of unskilled workers in 1998. In the Baltics, working in agriculture in 1991 meant higher odds of being an unskilled worker in 1998 (B=0.89). In Estonia, Latvia, Russia and Ukraine, being employed in industry and construction in 1991, gave a significant probability to be in the stratum of unskilled workers in 1998.

In the Baltics, peasants have high odds to retain their position (B=3.04). Those who were routine non-manual workers in 1991 have lower odds to be in the peasants category in the Baltics and in Belarus. Respondent's father's position as a peasant gives the respondent higher odds (B=2.39) to be a peasant in 1998 in Russia and Ukraine. Higher income raises the odds to stay in the peasant position during the seven-year period of transition in the Baltics (B=0.06). The position of peasant is also influenced by economic sector in 1991. Logically, workers in agriculture have a significant probability to be peasants in 1998 in the Baltics (B=1.57) but workers in the service industry also have higher odds to be peasants in 1998 (B=1.10).

In conclusion it is possible to say that in the ‘bottom up’ transition the chances of mobility and one's position can be predicted by individual and structural factors. In a ‘top down’ transition, only a few of these factors are significant (gender and industry, for example). In shrinking command economy the process of mobility is not as widespread as in ‘top town’ and ‘bottom up’ mobility, and cannot be predicted by the factors which shape mobility in the two former societies.

Comparing three different transition strategies in seven years produced not only economically unequal results but also individual patterns of intragenerational mobility. Our first hypothesis claimed that new market opportunities created by a ‘bottom up’ transition increase upward intragenerational mobility among young adults more than ‘top down’ transition or a society which moves slowly to a market economy. To draw conclusions about our first hypothesis it is necessary to take into account the main differences between the observed regions, which include a prolongation of a command economy (Belarus), a rapid transition to the market economy opening numerous opportunities to many people (Estonia and Latvia), and privatisation from top to down (Russia and Ukraine). When we look at intragenerational mobility, we see that more than three-quarters of young urban adults in Belarus (agricultural workers constituted only 3% of the respondents) were not mobile. In practically every stratum the percentage of stable respondents was the highest in Belarus. We must note that Belarus promoted very few young adults to managerial positions before the collapse of the Soviet Union and in the early years of transition.

The oligarchic pattern of the market (with only 5% of peasants among the respondents) was rather close to the Baltic's intragenerational mobility. If we look at the mobility table (Appendix 1) we see that the ‘bottom up’ market pattern in the Baltic countries has two strata with significant odds of becoming a manager (including entrepreneurs). Firstly, being a manager in 1991 very strongly raises odds to remain in the same stratum in 1998. Professionals have also high odds to move to managerial positions. In other types of transition, there is no significant mobility into the ranks of manager from other strata. In the Baltic states, education is a very strong predictor of mobility into the ranks of managers. In the other two types of transition, education is not a significant predictor of mobility to manager. It is interesting that years of education are significant in increasing chances of becoming a manager or a professional in ‘bottom up’ and from ‘top down’ marketisation, but not in the command economy of Belarus. The first hypothesis is supported as repetition of the historical path of a development market society through broad engagement of the people which produces higher mobility. Wide openings of entrepreneurial opportunities simultanously allowed for young adults to compete for those positions equally with older generations. It was an unique historical opportunity that was widely used by young adults in the context of bottom up transition.

Our second hypothesis claimed that the change in the industry structure of economy is an important factor of intragenerational mobility in the early stage of transition. In many surveys, industry was classified in the following way: state owned, private and budget financed. It was predictable that the whole goal of transition is to create privately owned sectors of economy to allow market forces to boost economic development. In our opinion it made no sense to study trivia. We used four categories: agriculture, industry, construction and service industry. Mobility among managers had significantly higher odds in construction linked to heavy and light industry in the Baltic countries. ‘Bottom up’ transition also reduced odds for semi-professionals to enter into service industry. Heavy and light industry and agriculture raised odds to be mobile into the ranks of skilled and unskilled workers in a ‘bottom up’ strategy of transition. Our hypothesis was only partially supported, as in ‘top down’ shock therapy no industry has significance in prediction of intragenerational mobility. Estonia and Latvia are very specific because of their ‘bottom up’ marketisation, and Belarus brcause of its surviving command economy and significance of economic sectors as predictors of intragenerational mobility. We were not surprised that the structure of industry did not have any impact on the mobility of young adults in the command economy of the Belarus. Possibly, in the industrial centres of Russia and Ukraine, the first years of ‘top down’ shock therapy created a lot of officially and unofficially unemployed people and little traceable mobility. But there may be a deeper explanation from ‘bottom up’ transition as the labour market emerged more quickly and industries changed very rapidly. This explains why in every stratum there are industries that raise odds to be mobile into other strata. In conclusion, we can say that the development of a labour market increases the predictability of intragenerational mobility through the industry structure.

Osborn and Slomczynski (1997) and Dievald et al. (2002) have also found that father's position becomes less important and education more important to be a manager in a post-socialist society.

Our third hypothesis was raised as a typical one for Western market-based democracies: ‘parents’ social strata predicts intragenerational upward mobility’. How strong is the impact of parents’ social position on the intragenerational mobility? Intergenerational mobility analyses in the Soviet era (Titma et al. 2003) show that parental social status had little influence, as state-sponsored mobility tried to eliminate parents from the educational process. As we see from Appendix 1, father's social stratum has little influence on any type of intragenerational mobility (only in Russia and Ukraine does the social reproduction of peasants occur). We can conclude that the cohort of young adults entered certain social strata before the transition started, and the mobility from those strata was not influenced by parental social status. Intragenerational mobility was surprisingly intensive, especially in the Baltic countries, but we cannot trace any direct impact from parents’ social status. The first suggestion is that, under the conditions of heavy competition, parents were not able to use their social capital. Otherwise, particularly in the movement to the position of managers it could have been a predictor of mobility. It has been found (Domanski 1995; Mayer et al. 1999; Diewald et al. 2002) that there are no impacts of political capital and parental position in nomenclature on social mobility. Our analyses confirmed this, as we did not show any trace of impact of parents in a decision-making position. The second suggestion is based on studies of Western countries; they have revealed that parents’ education and occupation influence most respondents’ education and first job, but their effect diminishes over time during the working career (Blau and Duncan 1967; Sewell et al. 1980; Warren et al. 2002; Sieben and Graaf 2003). The radical change from a socialist party–state to a market-oriented one forces young adults to change their occupations independently of the parental background.

Our fourth hypothesis stated that gender could be a strong predictor of intragenerational mobility. Higher coefficients are the result of the reference category, which has a more feminine composition. As we can see from Appendix 1, it is indeed the strongest predictor of intragenerational mobility. In all three strategies of transformation, men have higher odds to be managers, especially entrepreneurs (Pals and Tuma 2004). Among professionals who are mostly women, in the ‘bottom up’ strategy men have also higher chances to be professionals after the seven years of the transition. In this case, however, we can say that men are catching up with women. The same characterizes the semi-professionals. The composition of professionals and semi-professionals was mostly feminine (Mayer et al. 1999). Men have better opportunities to be semi-professionals in both types of the transition to a market economy. Only a shrinking command economy in the Belarus doesn't provide opportunities for men to be semi-professionals. In the Baltic countries, men have a higher probability to be employed in the sectors of manual work. In Russia and Ukraine, this is true only in the case of skilled workers. In Belarus there is no gender effect. In other words there are no changes in the gender segregation compared to the Soviet era. The strongest difference has occurred in the ‘bottom up’ transition. There appears a trend of gender segregation that characterized Western market economies a generation ago. First of all, manual labour, especially a skilled labour force, is heavily male dominated. Secondly, managerial, especially the entrepreneurial strata, is also dominated by men. At the same time, the professionals and semi-professionals strata are dominated by women. A more specific interpretation of the emerging gender segregation in transitional societies needs a more detailed analysis of this process.

Our fifth hypothesis was based on the quite widely traced pattern in the early years of transition: job changes were governed by income. Our analysis shows that income turned out to be a significant factor only in ‘bottom up’ transition in intragenerational m obility. Income in 1992 was highly significant and strongly raised odds to be managers and farmers in 1998 in Estonia and Latvia. It was a significant predictor of being skilled workers, but a weak predictor of mobility to the professionals and unskilled workers strata. Part of the prediction is due to the general rise and revaluation of comparative wages to the different social groups in labour market by markets. The market did not increase openings in lower white collar jobs as widely as for occupational groups. The fact that income in 1998 started to differentiate most occupational groups in the Baltic region indicates that, under ‘bottom up’ transition, the time lag needed for labour markets to evaluate properly the cost of the labour was short compared with ‘top down’ transition. ‘Top down’ and surviving command economy did not have income as a predictor of any occupational group in 1998.

Our analysis of intragenerational mobility in the transitional societies emerging after the collapse of the Soviet Union shows that differences in the strategy of transition are an important factor in intragenerational mobility. Mobility rates in countries where shock therapy, either from ‘bottom up’ or ‘up down’, were used, showed that at least a third of young adults were mobile (an extraordinarily high number in only seven years). Estonia and Latvia underwent the first version of transition by choosing a ‘bottom up’ pattern already in the early 1990s. ‘Bottom up’ transition provided opportunities for many and rapidly shaped a new labour market. Young adults were mostly moving up, and more than 12% reached a manager's position, doubling the size of this category for their age group.

Most important results of the ‘bottom up’ transition were those after six years, as our data show signs of a functioning labour market emerging in the Baltic countries. First of all, intragenerational mobility is predicted by the usual factors which contribute to mobility.

‘Top down’ shock therapy created a volatile situation where most people were deprived of the opportunities of the emerging labour market. In Russia and Ukraine, the building up of a market economy took place from ‘top to down’, leaving good chances for upward social mobility only for very capable young adults. As a result, intragenerational mobility for young adults was extensive in both directions. This can also be predicted by a limited number of factors. From the beginning, state bureaucracy was in command of the whole transition process; it administratively blocked young adults from the privatisation process as they lacked political and social capital. Compared to the ‘bottom up’ strategy, young adults had twice fewer opportunities to be mobile into the manager strata. If we look at the entrepreneurs, then the ratio is even lower. In general, upward mobility was also lower in ‘top down’ shock therapy. The problem was structural, as privatisation of large factories did not create new opportunities, but instead left people formally employed without salaries, as all bureaucratic structures and new owners were interested in hiding unemployment. It was very important that in both previous defence industry regions the factory towns were separated from each other, and in local areas with only one employer the only option was to migrate far away into cities like Moscow or Saint-Petersburg where jobs had been created. The opposite happened in Belarus, where the command economy was gradually shrinking and intragenerational mobility was low in all strata in comparison to both the ‘top down’ transition and the gradually shrinking command economy. Especially notable is that young adults have few options to be mobile into the ranks of managers. Compared to the Baltics, the proportion of managers among young adults is almost four times lower. Entrepreneurial positions are almost non-existent for them. mobility in Belarus was predictable by very few factors, as the command economy worked in the same way as during the Soviet era. The impact of intergenerational mobility on intragenerational mobility under radical change of society in early transition is insignificant. This is the major difference between transitional societies and the Western countries where linkage between social origin and destination is usual. Titma et al. (2003) found that, in the Soviet era, intergenerational mobility was predicted overwhelmingly by education, and the direct influence of parents’ social strata was weak and noticeable only in the Baltic countries. This was not a case of meritocracy forming the bases of intergenerational mobility. The main reason was that the rest of parental impact was eliminated by the state, which ran the educational process; this ended through tracking of very concrete work assignments after graduation. Our findings suggest that in the conditions of transition society parents did not emerge as a significant factor in intragenerational mobility. It is a predictable result, as the changes in intergenerational mobility can emerge through changing generations. Interestingly, father's education was the first sign of parental cultural capital to emerge as a predictor of intragenerational mobility.

Looking at labour market demand, we found that employers classified through the broad groups of industries were the predictors of mobility in some strata in the Baltic states. Surprisingly, this did not happen in Russia and Ukraine. Not surprisingly, this was not true for Belarus where the command economy was gradually shrinking. Our idea that previous command economy, structured by industry, has still some impact turns out not to be the case. Contrary to this, manufacturing/building industry was significant in intragenerational mobility. As far as the supply side was concerned, market valued education as a factor of mobility to the professional and managerial strata in shock therapy transition toward the market economy, with the exception of Belarus. But the market still did not evaluate education enough as a factor in mobility for the majority of strata. The most interesting finding was linked to gender. The Soviet Union and most Central European countries had specific gender segregation, where women prevailed in professional, semi-professional and routine non-manual strata. Marketisation made gender a very important predictor of mobility into certain strata. To be a man increased the odds to be a manager in all three types of transition. In Belarus it was the only stratum where gender had impact. Men in Estonia and Latvia, as well as in Russia and Ukraine, had good chances of moving into the ranks of professionals or semi-professionals. In other words, the market increased the trends of men moving into two previously very feminine strata. In Estonia and Latvia, men had a higher probability to be mobile to all other strata, but it is partially the effect of the reference category, routine non-manual, which is heavily feminine.

Six years of a transitional process is too short a period to make conclusions about long-term trends. more time is needed to make comparative analyses about the lasting effects of different strategies of transition. Social mobility is a fundamental social process and reflects transfer of a multiplicity of resources and, based on them, individual choices in a market society. More than a generation is needed to shape a new mobility pattern.

We gratefully acknowledge the support of the US National Science Foundation grants SBR-9710399 and SBR-0115028 and of the Estonian Ministry of Education, grant 018178s01. We would like to thank Rein Murakas, Liina-Mai Tooding, Nancy Brandon Tuma, Andu Rämmer, Villu Talv, Marit Veerus for help and advice, and anonymous reviewers for useful comments on an earlier draft.

Mikk Titma is Professor of Sociology at the University of Tartu and a senior researcher at Stanford University. He is interested in stratification and social mobility issues. Mikk Titma is the initiator and head of the ‘Paths of a Generation’ project.

Ave Roots is a PhD student of Sociology at the University of Tartu. Her research topics are intragenerational mobility in transitional society, especially in Estonia, and marriage as a channel of social mobility.