Adult Education,Stratification,and Regime Change: Upgrading and Sidestepping in Russia,1965

Abstract

Adult education influences how labor market opportunities are structured in the later life course. We propose a theoretical framework for understanding the stratifying role of adult education resting on the distinction between two forms of adult education—upgrading and sidestepping: Resources, incentives, and selection processes systematically structure rates of participation. Using educational history data from Russia, we test hypotheses derived from our framework and examine the impact of the Soviet collapse and the ensuing economic recovery. Upgrading exacerbates patterns of socioeconomic stratification by delivering better credentials to individuals with higher levels of initial resources. Sidestepping is less common than upgrading and less related to socioeconomic origins and previous attainment. The Soviet collapse produced short-term declines in the rates of both upgrading and sidestepping. However, once growth resumed, market institutions proved durable, and the political regime stabilized, rates of upgrading soared to levels exceeding those of the Soviet era.

Keywords

adult education upgrading sidestepping regime change GGS-EES Russia social origin

Inequality in adult schooling plays an important, growing, and underappreciated role in the stratification systems of modern societies. Growing returns to education and surging earnings inequality increase incentives for adults who lack higher credentials to return to school to obtain them. Accelerating technological change places an ever greater premium on adult education as a means of retooling and refurbishing skills during adulthood to keep up with rapidly shifting labor markets. Population aging, rising life expectancy, and reduced fertility have shifted the balance from contributors to beneficiaries of pension systems in many countries, bolstering the need for adult education to meet evolving skill demands and ensure that all potential workers remain in the labor market. Under these conditions, formal education is transforming into a lifelong process whereby individuals acquire and improve skills, knowledge, and competencies throughout the life course.

As adult education comes to play a more prominent role in many nations and constitutes a growing share of the total volume of education (Blossfeld et al. 2014), it takes on greater importance in understanding inequality in educational attainment. However, sociologists of education have only recently begun to explore the consequences of adult education for stratification. The classics of the adult education literature (e.g., Jarvis 2004; Merriam and Caffarella 1998; Richardson and King 1988) tend to offer normative accounts of how adult education emerged out of social movements for self-improvement rather than sociological analysis of how adult education might influence educational stratification processes (for exceptions, see Elman and O’Rand 2004; Hällsten 2011). Its potential stratifying role is intuitive: Adult education may help mitigate social distances and inequalities that emerged during schooling earlier in the life course by allowing educationally disadvantaged groups to catch up to their advantaged peers and thereby improve their career prospects. However, it could also exacerbate prior inequalities if individuals who are already advantaged are disproportionately likely to undertake adult education. The relationship between adult education and inequality is thus theoretically ambiguous and requires empirical study in concrete cases. Only by taking into account educational reentries during adulthood can we fully understand how inequality in educational opportunity unfolds and develops over the entire life course.

To advance sociological understanding of the stratifying role of adult education, we ask: Who is more (and who less) likely to undertake formal schooling during adulthood, and how is participation shaped by national context? Adult education takes different forms, which in turn exhibit varying levels of reward and patterns of association with socioeconomic status. Thus, to help conceptualize its role in stratification systems, we first propose a distinction between two broad types of adult education: upgrading and sidestepping. We then develop a theoretical framework that predicts how participation in each form relates to socioeconomic origins, prior attainments, and national institutions and policies. Finally, we test hypotheses we derive from our framework using retrospective survey data from Russia on participation in adult education from 1965 to 2005. The Russian case is particularly well suited for assessing the impact of institutions on adult education participation rates due to its wholesale regime change following the collapse of the USSR in 1991.

Prior research has linked participation in adult education to positive employment outcomes: higher earnings (Jacobson, LaLonde, and Sullivan 2005; Jenkins et al. 2003), increased employment opportunities (Jenkins et al. 2003; Kilpi-Jakonen et al. 2012), better occupational standing (Kosyakova 2018; Li et al. 2000), and exit from precarious jobs (Vono de Vilhena et al. 2016).¹ However, to understand how adult education relates to stratification, we must analyze how factors such as social origins, initial educational attainment, and current occupation relate to the propensity to undertake adult education.

The core issue is whether adult learners are drawn disproportionately from the already advantaged or the initially disadvantaged. Given its potential benefits, adult education can offer a “second chance” to adults who fell short in their initial schooling—or who “never had a first chance”—to secure better labor market opportunities (Jarvis 2007:191). Participation in adult education might therefore counteract inequality in “initial education” obtained during childhood and young adulthood. However, adult education could also serve as a mechanism of cumulative advantage (DiPrete and Eirich 2006), used disproportionately by elites to consolidate, extend, and reinforce their own and their children’s privileged positions in the labor market. Indeed, there is evidence that individuals who have already achieved relatively favorable educational or labor market positions are more likely than their less advantaged peers to undertake adult education in the United States (Elman and O’Rand 2004) and, to a lesser extent, Sweden (Hällsten 2011).

We go beyond these studies by proposing a broader theoretical framework for understanding the stratifying role of different types of adult schooling, which includes individual and institutional characteristics. We start with a crucial distinction between two forms of adult education: upgrading, which involves pursuing higher educational credentials than one obtained in one’s initial education, and sidestepping, which involves attaining additional training and qualifications at the same or (less commonly) lower level. Upgrading intuitively bears greater potential labor market returns than sidestepping because it stands to yield higher educational credentials than one obtained in one’s initial schooling. Sidestepping can also yield benefits, especially when it involves obtaining a similar level of education but in a different (presumably better rewarded) field of study. Here we show that the two forms of adult education exhibit different patterns of association with initial education, social origin, and career trajectories.

According to the life course approach, individual life trajectories result from the choices individuals make and the actions they take in response to the opportunities and constraints posed by social structures and institutional environments (Elder, Kirkpatrick Johnson, and Crosnoe 2003; Kerckhoff 1995). Following this premise, we argue that national educational and labor market institutions, as well as policies, can influence overall rates of participation in the two forms of adult schooling. Empirical evidence from cross-national studies tends to support this claim (Blossfeld et al. 2014; Kilpi-Jakonen et al. 2012). An alternative approach to studying the impact of institutions is to use a cross-temporal lens, testing for changes in participation patterns within a single country after it undergoes a major institutional transformation. Our empirical case, contemporary Russia, is ideal for this purpose due to the massive upheaval associated with the collapse of the Soviet system. Given that “institutional context changed suddenly while the persons involved remained the same” (Mayer 2006:14), this important “natural experiment” allows us to disentangle the impact of (post-socialist) institutional change from cross-national variation in individual traits (e.g., norms and culture).

Our retrospective longitudinal data set, the linked data of the Russian Generations and Gender Survey (GGS) and the Education and Employment Survey for Russia (EES), provides detailed information on parents’ characteristics and respondents’ education and job histories, which we use to construct time-varying measures for dynamic analyses of how social origins, initial educational attainment, and current occupation affect enrollment in schooling during adulthood. The observation period covered by the data, 1965 to 2005, allows us to assess how the changes in institutions, economic conditions, and policies associated with the collapse of Communism at the end of 1991 shaped overall rates of participation in the two forms of adult education.

Educational Stratification in Soviet and Post-Soviet Russia

The Soviet education system was highly centralized, standardized, and geared toward ensuring a sufficient supply of graduates with qualifications matching the skill requirements of the planned economy. Following the Second World War, most students completed the equivalent of middle school (eighth grade) and then had four choices: They could end their schooling and enter employment, continue to “general secondary” school (which provided the equivalent of a high school diploma after two or three years of additional study and was by far the most common pathway to university education), enroll in a program of lower vocational training designed to impart skills in manual trades, or enroll in a “specialized secondary” institution, which provided training in low-level professional or high-level clerical and technical occupations while providing a secondary diploma. Over time, larger proportions of lower vocational programs provided a secondary diploma along with a certificate of vocational training, but this remained a low-status form of education, and the distinction between vocational training with and without a secondary degree was immaterial in terms of the labor market (Connor 1991). After obtaining a general secondary degree, students could apply for enrollment in a university or in another form of specialized secondary (also called a tekhnikum) that provided two to three years of training in low-level professional or technical occupations. The communist system purportedly sought to eliminate origin-based inequalities in access to education, yet such inequalities persisted in Russia throughout the Soviet period despite a massive educational expansion in the post-war era (Gerber 2000; Gerber and Hout 1995).

The transition from school to work was coordinated by state institutions, which assigned graduates of many specialized secondary institutions and all university programs to specific jobs, ensuring high occupational specificity and tight links between educational certificates and occupations (Gerber 2003). Unemployment, formally forbidden, remained at minuscule levels (Gregory and Collier 1988). Strong occupational boundaries, state control over hiring and firing of personnel, standardized promotion plans linking career advancement to qualifications and seniority, and policies tying tenure with the same employer to benefits and privileges restricted occupational mobility over the life course (Gerber 2002).

The collapse of the Soviet Union at the end of 1991 threw Russia’s labor market and education institutions into turmoil. Liberalization of prices, exchange rates, and trade; privatization of enterprises; foreign competition; and the dismantling of Soviet-era economic and political institutions produced widespread economic dislocations, hyperinflation, recession, and a highly uncertain, crisis-like environment for most of the 1990s. These problems culminated in a financial crisis in August 1998, precipitated by a government bond default. The economy stabilized in 1999, then grew impressively during the 2000s, fueled by advantageous global prices for oil and natural gas (Russia’s main exports) and facilitated by growing political stability as Vladimir Putin consolidated and centralized power during his first two terms as Russia’s president.

The sphere of education also experienced a cycle of turmoil and uncertainty in the 1990s followed by recovery and stabilization in the 2000s. These trends were particularly evident in higher education (Konstantinovskiy 2017; Smolentseva 2017). In the 1990s, demand for university education declined. Russians had to scramble to make ends meet, raising the opportunity cost of full-time university enrollment. Professions of long-standing prestige, such as scientists and engineers, saw severe losses in relative earnings and job stability due to the contraction of state-based employment. Government investment fell sharply as state budgets shrank, and the devaluation of Soviet-era credentials raised questions as to the value of obtaining a postsecondary degree. As the economy recovered in the 2000s, so did the system of higher education: Institutions became more differentiated based on their quality, specialization in different fields, and levels of government investment. Tuition payments were introduced in the public sector, and private, fee-charging institutions emerged. Employers once again placed value on university degrees as a measure of applicants’ ability to master new skills and knowledge, and they increasingly distinguished between institutions of different quality. Enrollments reversed their initial decline by the end of the 1990s, and they soared during the 2000s as institutions increasingly competed for fee-paying students. At the same time as the supply of seats in universities grew, demographic trends created demand-side pressure because the size of cohorts entering the standard age of university matriculation declined steadily starting in the mid-2000s due to fertility declines that started in the late 1980s (Smolentseva 2017). Other levels of education (lower vocational schools, general secondary schools, and specialized secondary schools) also experienced turmoil, reform, and contraction in the 1990s, with some recovery (uneven across sectors and regions) in the 2000s.

Adult Education

The Soviet government promoted lifelong learning through a formal system of adult education (Kljucharev 1997), with a large network of participating educational institutions and numerous programs, sometimes compulsory, of “qualification upgrading” (povysheniia kvalifikatsii) and retraining (Zajda 2003). From 1940 to 1980, the number of Soviet adults participating in qualification upgrading grew from 1.7 to 32.7 million workers annually, and those obtaining new qualifications (i.e., changing specialties or professions) grew from 1.9 to 5.9 million (GOSKOMSTAT USSR 1981). However, no studies examine how participation in different forms and types of adult education varied by social background and socioeconomic status in Soviet-era Russia. With our focus here on enrollment in formal degree programs as an adult, we do not consider short-term qualification upgrading because such programs generally did not result in degree-level qualifications. If anything, they provided certificates of completion.

Following the collapse of the Soviet Union, the Russian government issued various pronouncements and official acts regarding adult education (for details, see Zajda 2003). Evening classes for adults were bolstered, with the stated goals of preventing structural unemployment through retraining, preparing qualified workers for the new market economy, and compensating for previous inequalities in access to secondary education. However, Russia’s institutions have no separate branch of adult education or special legislation or a legal framework governing it (Khokhlova, Kozlovskiy, and Veits 2013). Official rates of participation in all forms of adult education (including short-term qualification upgrading, which we do not consider here) fell sharply compared to the Soviet era (e.g., reaching only 4.5 percent in 2006; GU VShE 2010).

Upgrading Versus Sidestepping

We follow the convention in the growing literature on adult education of defining the concept as undertaking a formal degree program of schooling at some point subsequent to the completion of “initial” education, that is, a relatively uninterrupted sequence of education enrollment begun during childhood (see e.g., Blossfeld et al. 2014; Kosyakova 2018). Our definition allows for multiple spells of adult education: Any undertaking of a formal degree program after the completion of initial education counts.

We introduce a distinction between two types of formal adult education: (1) enrollment in a level of education higher than that previously attained (upgrading) and (2) enrollment in education at the same level or below (sidestepping). The potential incentive for upgrading is evident: It provides a higher formal qualification than one’s initial (or current) credential. The motive for sidestepping is most likely to obtain training in a different field of study or at a more prestigious institution—that is, a “horizontal” dimension of educational stratification—either of which might pay off in the labor market. The depreciation of previously accumulated skills due to technological change could provide incentive to obtain the same type of credential as one had initially but in a different field (Li et al. 2000). As such, sidestepping might be a strategy to keep up with rapidly changing skill demands of labor markets in rapidly restructuring economies, such as that of post-Soviet Russia.

Based on theoretical considerations regarding the different nature of upgrading and sidestepping, we propose hypotheses about how enrollment rates in the two types of adult education might be structured—in some cases similarly, others differently—by social origin, initial educational attainment, occupational trajectory, and institutional context (Soviet vs. post-Soviet).² Our distinction between upgrading and sidestepping could explain contradictory empirical findings regarding the relationship between social position and the propensity of enrolling in undifferentiated adult education (i.e., ignoring the difference between upgrading and sidestepping). For example, some studies link adult education to higher initial education and initial labor market disadvantage (Felmlee 1988; Hällsten 2011), whereas others find the opposite relationships for these variables (Kilpi-Jakonen et al. 2012; Kosyakova 2014). Our hypotheses associate the former combination (overqualification) with sidestepping, the latter (underqualification) with upgrading.

Theory and Hypotheses

Although adult education has potentially important implications for a society’s ability to adapt to population aging and technological change, our primary interest here lies in its role in social stratification and its relationship to institutional and political change. Given that both upgrading and sidestepping enhance the educational resources available to those who complete adult education programs, the critical issue for understanding their role in stratification is whether adult learners are drawn disproportionately from the already advantaged or the initially disadvantaged. In the first case, the overall impact of adult education will exacerbate inequalities; in the second case, adult education will diminish inequality by providing more opportunities to the disadvantaged than advantaged groups. We address the potential influence of upgrading and sidestepping on stratification by specifying a series of hypotheses about how the propensity to enroll in them relates to measures of social position.

Our underlying theoretical assumption is that individuals make rational choices about whether to enroll in adult education based on its costs (direct and indirect), their probability of success, and the potential payoff of the educational outcomes. The rational action framework has been explicitly developed in analyses of educational choices in childhood (Breen and Goldthorpe 1997), but it can also be used to derive expectations regarding educational choices later in the life course. Our expectations are also rooted in a strand of life course theory that argues initial advantages in educational careers are likely to be reproduced and amplified at later educational and labor market junctures (see e.g., the “cumulative advantage effect” discussed by DiPrete and Eirich [2006] or Merton’s [1968]“Matthew effect”), resulting in a steadily growing gap between initially advantaged and disadvantaged groups. In other words, earlier selection processes in children’s education may enhance or constrain later opportunities for adult education. Accordingly, the following discussion draws on a rational action framework and life course theory to develop hypotheses regarding how various push-and-pull factors, resources (as reflected in background characteristics), and incentives interact to generate unequal rates of participation in upgrading and sidestepping.

Measures of Social Position: Resources, Incentives, and Selection

We expect three measures of social position—social origins, initial educational attainment, and current occupation—to affect the odds of undertaking adult education (either upgrading or sidestepping) because these measures differentially shape the resources available for adult education and the incentives for doing so. Both resources and incentives should positively affect an individual’s propensity to enroll in adult education. The same variables may also serve as proxies for earlier education selection processes that are relevant for the prospects of adult education; as we will explain, earlier selection on the basis of social origins may also influence rates of adult enrollment. We first discuss how and why these measures might exhibit net additive associations with enrollment rates; we then consider potential interactions among them.

Social origins

Advantageous family background, broadly understood as having parents with relatively high educational and occupational status, facilitates educational attainment (see Breen and Jonsson 2005). Resources are one potential mechanism through which social origin influences educational attainment: Offspring can use parents’ educational and occupational resources to cover the costs (including direct costs and opportunity costs) of education. Although parents may be less inclined to invest in adult children’s education, on average, adults with higher status parents should have more resources available to finance additional education. Moreover, higher status parents are more likely to have socialized their children to value education as a marker of status or an intrinsic good. Thus, both origin-based resources and origin-based normative incentives point to the following hypothesis:

Hypothesis 1a: Adults with high-status parents upgrade and sidestep at higher rates.

However, social origin–based educational selection during childhood might create a situation where, net of initial (childhood) education, social origins are negatively correlated with academic ability and motivation during adulthood. This argument assumes that due to origin-based advantages during childhood, higher origin individuals who fall short in educational attainment tend to have lower levels of academic ability, motivation, or other unobservable traits associated with educational attainment compared to lower origin individuals with the same (lower) levels of initial attainment because the social disadvantages experienced by lower origin individuals may drive them from school prematurely despite high levels of ability and motivation. Origin-based educational selection (“social selection”) during childhood would thus work in favor of lower origin students in adulthood because in contrast to higher origin students, their earlier failure to attain their desired education level is less likely to be due to lower academic ability. This selection mechanism should mainly be relevant for upgrading because only upgrading involves advancing to a higher level of education than that attained in childhood.

Hypothesis 1b: Adults with high-status parents upgrade at lower rates.

Educational level

Individuals with higher levels of educational attainment should typically have more resources available (e.g., more wealth from accumulated savings) to undertake adult education. Also, the fact that they attained a relatively high level initially implies that—controlling for social origins—they have the relatively higher levels of academic ability necessary for completing formal courses of education as adults, more positive attitudes toward learning, or greater appreciation for education (Bills 2000). In contrast, lower educated individuals are likely to be deterred by negative earlier life course experiences with schooling or a lack of self-efficacy or because they place lower value on education (Valentine and Darkenwald 1990).

Hypothesis 2a: Adults with higher levels of education upgrade and sidestep at higher rates.

Individuals with higher levels of education may also have lower incentives to enroll in further education: the higher the initial investment in education, the higher the potential earnings possibly forgone while enrolled in adult education (Schömann and Becker 1995). This can be controlled for to some extent, but not completely, with a measure of current occupation. Moreover, individuals who already invested more time (and forgone earnings) in initial education may be reluctant to risk investing more in adult education. There is also a potential ceiling effect (Blossfeld et al. 2014) on upgrading for individuals with higher levels of education: They simply have fewer options available for upgrading compared to people with lower levels of education. This does not apply to sidestepping: Higher educated individuals may be especially prone to return to formal schooling in the later life course due to their positive attitudes toward learning and their greater appreciation of education.

Hypothesis 2b: Adults with higher levels of education upgrade at lower rates.

Occupational class

Empirical findings from other countries indicate that adult learners tend to be employed (Kilpi-Jakonen et al. 2012), have higher earnings (Hällsten 2011), and have higher occupational status or positions (Elman and O’Rand 1998). Higher status occupations presumably offer more resources for smoothing consumption while spending time in formal adult education. Individuals in such occupations have a stronger need to retool to keep their skills up to date (Carr and Sheridan 1999; Elman and O’Rand 2007), which is especially likely via sidestepping to maintain current occupational status (Hällsten 2011). Finally, controlling for initial education and origin-based resources, selection into higher status occupations may be based, in part, on ambition and motivation, which should increase the probability of enrolling in adult education.

Hypothesis 3a: Higher occupational status increases rates of upgrading and (to a greater extent) sidestepping.

At the same time, the higher earnings associated with higher status occupations may produce the opposite incentive as taking time off work for adult education would generally be associated with more lost income (Carr and Sheridan 1999; Schömann and Becker 1995). Other things being equal, individuals in less favorable occupations should generally have greater incentives to pursue additional education to improve their labor market position. Consistent with this argument, research shows lower wages (Jepsen and Montgomery 2012; Stenberg 2011) and (previous) unemployment (Hällsten 2011; Stenberg 2011) “push” individuals toward adult education. Both upgrading and sidestepping may facilitate exit from disadvantaged labor market positions, but upgrading would seem to offer better prospects for doing so; hence, it should appeal particularly to people in disadvantageous labor market positions.

Hypothesis 3b: Lower occupational status increases rates of upgrading and (to a lesser extent) sidestepping.

Once again, we have competing hypotheses regarding the direction of the additive effects of measures of social position, this time for upgrading and sidestepping, although the effects on sidestepping should tilt more toward the positive direction. Depending on the validity of our hypotheses and the relative weight of the positive and negative influences, we may find either null effects or curvilinear effects in models containing only additive effects.

Measures of Social Position: Interactions

Apart from the additive effects of measures of social position hypothesized so far, we anticipate that individuals’ current occupation interacts with their social origins and initial education in intuitive ways.

Social origin and occupational class

Adults from advantaged social origins who attain lower occupational positions have both origin-based resources and occupation-based incentives that would support upgrading in particular but perhaps also sidestepping. Such individuals represent one side of a process that Gerber and Hout (2004) call “regression toward origins,” whereby previously downwardly mobile individuals (relative to their origins) improve their occupational position over the life course. In turn, individuals who at first experience upward mobility relative to their parents have greater risk of downward occupational mobility as they age. Adult education, particularly upgrading, can help downwardly mobile individuals “catch up” up to their origin status.

Hypothesis 4a: Rates of upgrading are elevated for individuals who are downwardly mobile with respect to their parents.

At the same time, the fact that downwardly mobile individuals were not able to convert their origin-based resources into occupational attainment initially may indicate they lack the ambition, ability, or interest for labor market success. Thus, selection on such unobservable variables into low-status occupations among people with advantaged social origins may yield the opposite effect.

Hypothesis 4b: Rates of upgrading are particularly low for individuals who are downwardly mobile with respect to their parents.

Educational level and occupational class

We expect that two forms of mismatch between educational and occupational attainment have opposite effects on upgrading and sidestepping. Individuals who are educationally underqualified for their current occupations have greater incentives to upgrade their education level in order to reinforce their occupational position. Consider, for example, a woman who earns an executive position in a real estate company due to outstanding sales and strong lower level managerial performance despite having only a general secondary diploma or a man whose on-the-job experience yields a promotion to an engineering job that usually requires a university degree despite having only specialized secondary training normally appropriate for a technician position. Although both individuals experience career advancement via means other than formal credentials, they would do well to solidify their standing in their high-level positions by attaining the educational credentials normally associated with such positions. However, they would have no incentive to sidestep because obtaining a credential at the same level or below their current one would do nothing to solidify their occupational status.

Hypothesis 5: Underqualified individuals (higher status occupation with low education credentials) upgrade at higher rates and sidestep at lower rates.

In contrast, individuals who are overqualified for their current job (they have high educational credentials but low-status jobs) presumably obtained their credentials in specific specializations or at specific institutions that have proven unhelpful in securing a labor market position appropriate to their education credential. Consider the university graduate with a degree in humanities who cannot find professional employment and thus drives a taxi or one who obtains a secondary professional degree to become a nurse in a specialty where demand falls due to technological advances that place a higher premium on the skills of medical technicians. In such cases, sidestepping—obtaining a credential of the same level but in a more lucrative specialization or at a more prestigious institution—is an especially appropriate strategy for improving one’s labor market position. Upgrading is less appealing because one is already overqualified. Moreover, the initial schooling process may have provided individuals a strong signal for what their appropriate level of education should be.

Hypothesis 6: Overqualified individuals (low-status occupation with high education credential) sidestep at higher rates and upgrade at lower rates.

Note that our proposed interactions are independent of the additive effects proposed in Hypotheses 1a through 3b. This means they may offset or outweigh the hypothesized additive effects. Likewise, not controlling for the interactions may result in weaker or misleading estimates of the main effects. For example, if it is only the combination of high origins with a low current occupation that encourages upgrading, then excluding the interaction could yield additive effects of both high origin and low current occupation, which would be tantamount to misspecifying the model if the two variables only exert effects in interaction with each other.

Regime Change

Technological change, globalization, and innovations accelerate the depreciation of previously accumulated human capital, which may increase firms’ and workers’ incentives to invest in adult education to meet the new requirements of changing labor markets (Bartel and Sicherman 1998). An extreme version of this process is the “transition shock” caused by radical liberalizing reforms in Russia after the Soviet collapse, which stimulated considerable restructuring of the labor market (Gerber 2002). Many qualifications and skills obtained under the Soviet system became obsolete and inadequate, resulting in a massive devaluation of previously accumulated human capital (Sabirianova 2002). This unusually rapid labor market restructuring should, in principle, foster a surge in adult education as workers seek to obtain new skills suitable to a competitive market economy and employers reward such skills.

Hypothesis 7a: The collapse of the Soviet regime in 1991 increased rates of upgrading and sidestepping.

Russia’s shock therapy also produced a massive and extended economic contraction featuring political instability, surging inflation, declining incomes in professional state sector jobs, and new opportunities in trade, sales, services, and other activities where social network connections often count for more than educational credentials. The ultimate outcome of market reforms—and thus their longevity—was highly uncertain. Moreover, the educational system was demonopolized, mandatory (and thus guaranteed) job assignment for graduates of universities and specialized secondary schools was abandoned, and educational institutions and teachers received greater autonomy (Gerber 2003). These components of the economic turbulence of the post-Soviet reform era may have created tremendous uncertainty about the value of educational credentials, thereby diminishing incentives while increasing opportunity costs (e.g., income in non–credential based activities) of adult education.

Hypothesis 7b: The collapse of the Soviet regime in 1991 decreased rates of upgrading and sidestepping.

Although many studies of Russia’s post-Soviet transition treat the post-Soviet era as a uniform period, the turbulent period in Russia’s labor market and education system came to an end in 1999 (Gerber and Radl 2014). The 2000s witnessed political stability, rationalization of economic institutions such as the labor and tax codes, and impressive and persistent economic growth. Chaos in the higher education system subsided as new hierarchies and a new institutional framework for public and private institutions emerged (Smolentseva 2017). The chaos of the 1990s reduced the value and signaling function of educational credentials, but the 2000s saw a resurgence of their importance, and shrinking cohorts entering the standard ages for university enrollment created more opportunities for older students to enroll (Konstantinovskiy 2017; Shishkin 2004). Thus, Hypothesis 7a applies more clearly to the turbulent years of the post-Soviet era, whereas Hypothesis 7b quite likely applies to the recovery years of the 2000s.

Hypothesis 7c: Rates of enrollment in both types of adult education declined during the 1990s and increased during the 2000s.

Data and Method

Data and Sample

We test our hypotheses using retrospective data from the merged Generations and Gender Survey (GGS) and the Education and Employment Survey for Russia (EES),³ which cover the years 1965 to 2005. The first wave of the Russian GGS was carried out in 2004 based on a nationally representative sample of 11,261 respondents between 18 and 79 years old, with each respondent representing one household. The response rate was around 15 percent in St. Petersburg and Moscow and around 57 percent in all other areas (IISP 2014). The EES is a follow-up study of 18- to 54-year-old GGS respondents performed in 2005 (the response rate was around 85 percent, resulting in a sample of 6,455 respondents).⁴

These linked data contain information on the timing and nature of all respondents’ educational enrollments, including those undertaken after the completion of “initial education.” We define initial education as that completed prior to leaving the education system for at least 24 months or for a year during which it is normative to be enrolled in secondary or presecondary levels (i.e., a dropout from secondary or pre-secondary school).⁵ We define the risk set as person-months during which respondents are adults (at least 18 years old), have completed initial education, have entered their first job, and are not currently enrolled in adult education.⁶ Respondents are censored when they turn 45 years old (after which adult education is rare)⁷ or if they were under 45 when surveyed, at the interview date. After these restrictions, the risk sample includes 5,696 EES respondents, about 88 percent of the original sample. We use episode-splitting to create time-varying measures of education (which we update after additional degrees are obtained during adulthood), occupational class, employment status, family status, childbearing, and place of residence (Blossfeld, Golsch, and Rohwer 2007). For instance, episodes were split into subepisodes when partnership status changed. The resulting sample includes 47,447 person-spell observations, with average spell duration of 25 months and median duration 14 months.⁸

Variables

Dependent variables

We define adult education events as upgrading if the level in which the respondent enrolled is higher than the level previously attained and sidestepping if the level is equal to or lower than that previously attained. Table 1 shows the classification and number of occurrences in the sample of each combination of prior and adult-enrolled levels. Overall, there are 718 enrollments in upgrading and 170 in sidestepping. Note that nearly half of the examples of upgrading (334) and nearly two-thirds of the examples of sidestepping (107) involve adult education in institutions below tertiary level. This demonstrates it is insufficient to only analyze university attendance by adults, as some studies have done (e.g., Carr and Sheridan 1999; Hällsten 2011, 2012), to understand the full scope of upgrading, sidestepping, or adult education in general.

Table 1.

Incidences of Adult Education, Distinction between Upgrading (UP) and Sidestepping (SST).

Educational Attainment at the Beginning of Adult Education Spell	Type of Adult Education Activity
	Lower Vocational	General Secondary	Secondary Professional	Higher	Postgraduate
Incomplete secondary	UP 7	UP 26	UP 44	UP 9	UP —
Lower vocational	SST 14	UP 13	UP 55	UP 24	UP —
Complete secondary	SST 32	SST —	UP 189	UP 128	UP 1
Secondary professional	SST 5	SST 3	SST 48	UP 206	UP —
Higher	SST —	SST —	SST 5	SST 54	UP 16
Postgraduate	SST —	SST —	SST —	SST 3	SST 6

Source: Linked Generations and Gender Survey (GGS; 2004) and the Education and Employment Survey for Russia (EES; 2005) data; authors’ calculations.

Independent variables

To address inequality in adult education enrollment rates, we estimate the effects of social origin (Hypotheses 1a and 1b), initial education (Hypotheses 2a and 2b), occupational class (Hypotheses 3a and 3b), origin-occupation mismatch (Hypotheses 4a and 4b), and education-occupation mismatch (Hypotheses 5 and 6).

Social origin is measured by parents’ highest occupational class. We first convert the original International Standard Classification of Occupations (ISCO) occupation codes into the modified version of the Erikson-Goldthorpe-Portocarero (EGP) class schema outlined by Gerber and Hout (2004). To enhance the power of our tests, we then collapse the EGP schema into three classes: salariat (managers and professionals), intermediate (upper and lower routine nonmanual, skilled manual), and unskilled (unskilled and semi-skilled).

Education is a time-varying categorical measure of the highest level attained by the start of each spell. Following Gerber (2003), we define five educational attainment levels: (1) incomplete secondary, (2) lower vocational, (3) general secondary, (4) specialized, and (5) higher.

Occupational class is a time-varying variable for current—or, if the respondent is not employed, previous—occupational class.⁹ In contrast to the ISCO-coded parental occupation variables, current occupations are coded in the survey using categories similar to one-digit ISCO codes, which we sort into (1) unskilled and semi-skilled manual, (2) skilled manual, (3) lower-level routine nonmanual, (4) upper-level routine nonmanual, and (5) salariat. Detailed information on variable construction is available in Appendix Table A1.

Origin-occupation mismatch is derived by cross-classifying parental and current occupation; it is potentially time-varying given change over time in the latter. We classified respondents as upwardly or downwardly mobile during spells when their current (or most recent) occupation was, respectively, higher or lower than their parents’ occupation. Spells of immobility, when parental and respondent occupational class are equal, serve as the baseline (omitted) category.

Education-occupation mismatch is another time-dependent variable with three categories based on cross-classification of the two component variables (education and occupational class). We determined a priori the normative level of education typically required for jobs in each class.¹⁰ We then coded respondents as overqualified or underqualified during spells when their education exceeded or fell short, respectively, of the normative education for their current (or, for spells of nonemployment, most recent) occupational class. Spells when respondents’ education matches the norm for their class are treated as matched, the baseline category.

To address the impact of the regime change on adult educational transitions (Hypotheses 7a, 7b, and 7c), we defined a time-varying variable for period with three categories: (1) the Soviet period (if the spell occurred 1965–1991), (2) the turmoil period of the transition (1992–1998), and (3) the recovery period (1999–2004). To test Hypotheses 7a and 7b, we combined the latter two into a single post-Soviet measure.

Controls

We control for several variables that may confound relationships between our variables of interest and adult education enrollment hazards: (1) gender, being married, and having children, because of women’s and men’s distinct life course patterns and the potential impact of family roles in educational and employment trajectories; (2) type of locality, because structural opportunities and the supply of educational institutions likely differ in cities of different sizes; (3) previous adult education enrollment, a proxy for unmeasured variables that positively predict enrollment; (4) previous number of jobs, which measures propensity to change jobs, a potential motive for adult education; (5) current employment, because the opportunity costs of adult education are lower for people who are not employed; and (6) sector of (current or most recent) employment, to account for possible sectoral variation in formal education requirements for advancement.

Method

Our data are potentially right censored; individuals may enroll in adult education at any point in time when they are at risk, and our hypotheses involve time-varying covariates. Therefore, we need dynamic models rather than simple logistic regressions predicting any enrollment during the observation window (Blossfeld et al. 2007; Light 1995). We estimate separate continuous-time parametric event-history models for upgrading and sidestepping.

The dependent variables are the logged transition rates (hazard rates), specified as the logistic transformations of the probabilities of enrolling in upgrading, sidestepping, or any adult education in month t conditional on being at risk. Respondents leave the risk set when they enroll in adult education and reenter upon completing it: Because individuals can and do enroll multiple times in adult education, censoring observations after first enrollment would entail a loss of information.¹¹

Our time metric is the number of months since respondents’ first job entry after attaining initial education. We specify duration dependence of the baseline hazard as a Gompertz function, which assumes a monotonically decreasing transition rate to adult education. This is in line with our expectations: Experience (time since an individual entered the labor force following initial education) leads to a monotonic decline in the enrollment rate because as time passes, the horizon for realizing returns to adult education declines. The Gompertz function proved optimal in preliminary tests of functional forms.¹²

We exclude spells when respondents have already attained higher education from the models for upgrading (because entry to postgraduate programs is rare) and spells when respondents had only attained incomplete secondary from the sidestepping models (because no respondents enrolled as adults at that level or lower; see Table 1).

Results

Enrollment Rates in Adult Education in Russia

Kaplan-Meier estimates for enrollment in upgrading and sidestepping in Russia during the observation window show that upgrading is much more common than sidestepping, and it tends to happen sooner after entering the labor force (Figure 1). This difference is not surprising because there are fewer labor market incentives to sidestep. Ten years after labor force entry, 12 percent of individuals upgrade, whereas only 3 percent sidestep. Overall, 16 percent participate in upgrading by age 45, 4 percent in sidestepping. These numbers show that a nontrivial proportion of Russian adults have undertaken adult education since the 1960s.

Figure 1.

Plot of survivor functions (product-limit estimation) for upgrading and sidestepping.

Transition to Upgrading and Sidestepping: Multivariate Results

We estimated models for upgrading, sidestepping, and entry into any adult education. For each outcome, we started with additive specifications of the effects of social origin (Hypotheses 1a and 1b), initial education (Hypotheses 2a and 2b), occupational class (Hypotheses 3a and 3b), and period (Hypothesis 7a), net of the control variables described earlier. We then incorporated the appropriate interaction terms to test for the effects of origin-occupation mismatch (Hypotheses 4a and 4b) and education-occupation mismatch (Hypotheses 5 and 6) as well as the finer-grain specification of the period effect to test the hypothesis of variation within the post-Soviet era (Hypotheses 7b and 4c).¹³ We fit an additional model for sidestepping (2.3) with an alternative specification of the overqualified term, which we discuss in the following. To conserve space, we present all our estimation results in Table 2. We first discuss the additive and interactive effects of social position and the period effects for upgrading and sidestepping, then we compare those results to what we would find if we were to ignore the distinction between the two forms of adult education.

Table 2.

Gompertz Regression Models Predicting Transition to Upgrading and Sidestepping (competing risk approach).

	Upgrading		Sidestepping			Any Adult Education
	Model 1.1	Model 1.2	Model 2.1	Model 2.2	Model 2.3	Model 3.1	Model 3.2
Social origin (reference = unskilled)
Intermediate	.19*	−.05	.41*	.43*	.45*	.23**	.05
Salariat	.38**	−.25	.30	.38	.43	.36**	−.09
Educational level (reference = secondary completed)
Incomplete secondary	−.43*	−.77**				−.55**	−.91***
Lower vocational	−.69***	−.82***	−.31	−.16	−.31	−.66***	−.77***
Secondary professional	−.72***	−.63***	.75**	.70**	.71**	−.49***	−.38***
Higher			1.50***	1.30***	1.23***	−1.59***	−1.31***
Occupational class (reference = unskilled and semi-skilled)
Skilled manual	.13	.34*	−.70*	−.72*	−.75*	.04	.18
Lower level routine nonmanual	.25+	.41*	−.37	−.40	−.93*	.16	.24
Upper level routine nonmanual	.79***	.99***	−.58*	−.47	−.63+	.57***	.67***
Salariat	1.63***	1.64***	−.57+	−.26	−.51	1.06***	1.03***
Own-parents’ occupation match (reference = immobile)
Upwardly mobile		−.32*		.15	.15		−.24*
Downwardly mobile		.36**		.03	−.01		.27*
Education-occupation match (reference = matched qualifications)
Overqualified		−.07		.28			−.02
Underqualified		.35**		−.09			.45***
Educational level, higher × lower level routine nomanual					1.25*
Period (reference = Soviet period)
Post-Soviet	−.07		−.22			−.09
Post-Soviet, turmoil		−.54***		−.48*	−.47*		−.48***
Post-Soviet, recovery		.38***		.04	.05		.29**
Controls	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Number of spells	40,814	40,814	41,260	41,260	41,260	47,447	47,447
Number of individuals	4,759	4,759	5,054	5,054	5,054	5,694	5,694
Number of failures	701	701	170	170	170	887	887
Model fit
Log likelihood	−2,665	−2,627	−851	−847	−845	−3,319	−3,280
χ²	262	339	146	153	158	316	394
Degrees of freedom	25	31	25	31	30	26	32
Akaike Information Criterion	5,384	5,319	1,755	1,760	1,753	6,695	6,628
Bayesian Information Criterial	5,617	5,604	1,988	2,045	2,029	6,940	6,926

Source: Linked Generations and Gender Survey (GGS; 2004) and the Education and Employment Survey for Russia (EES; 2005) data; authors’ calculations.

Note: Controls include being female, married, having children, rural area of residence, experience of adult education, previous number of jobs, currently working, sector of economy in the current or previous job, approximate education, GGS residence area in Moscow or St. Petersburg, and missings for occupational class (including armed forces) and social origin (including armed forces or nonworking for both parents).

p < .01. *p < .05. **p < .01. ***p < .001.

Additive Effects of Social Position

Russians from advantaged social origins upgrade at higher rates, net of their own educational and occupational attainment and other controls (Model 1.1). The offspring of salariat parents have a 46 percent (=exp[.38] – 1) higher hazard than the offspring of semi- and unskilled families; offspring of parents with intermediate class positions have a 21 percent higher hazard. For sidestepping, the effect of social origin is nonlinear: Rates of sidestepping are highest for respondents from intermediate social origins (Model 2.1). Hypothesis 1a, regarding higher rates of adult education among offspring of high-status parents, is supported only for upgrading, not sidestepping. Once we introduce the interaction effects into the upgrading models, the main effects of higher origins reverse signs and lose statistical significance (see the following). Hypothesis 1b, predicting lower propensity of upgrading for adults with high-status parents, received no empirical support, indicating that low academic ability among high-origin adults who fell short in their initial education does not deter them from attempting to catch up in terms of formal education.

The effect of education on the rate of upgrading is nonlinear, and it is positive for sidestepping. Respondents with complete general secondary education undertake upgrading at the highest rates (Model 1.1): 1.5 times (=1/exp[–.43]) higher than the rate for respondents with incomplete secondary and about twice as high as the rates for respondents with lower vocational and secondary professional education.¹⁴ In contrast, the hazard of sidestepping is 4.5 times higher for university graduates than for respondents with general secondary degrees (Model 2.1). Hence, Hypothesis 2a, that higher educated adults are more likely to upgrade and sidestep, is confirmed only for sidestepping. Hypothesis 2b, predicting a lower propensity of upgrading for the higher educated on the basis of social selection in initial educational attainment is not supported. Graduates of specialized secondary institutions (which provide lower level professional degrees or technician certification, an intermediate form of qualification between general secondary and university) have lower rates of upgrading but higher rates of sidestepping compared to adults with general secondary education.

Current occupational class also has different effects on the two types of adult education. Conforming to Hypothesis 3a but in contrast to Hypothesis 3b, the hazard of upgrading is higher for respondents in more advantageous occupations: For example, compared to semi- and unskilled workers, managers and professionals have a 5 times higher rate of upgrading, and upper routine nonmanual workers have a 2.2 times higher rate (Model 1.1). In contrast, the results for sidestepping conform to Hypothesis 3b (contrary to our expectation that Hypothesis 3b would apply more clearly to upgrading) and indicate that the lowest occupation class, semi- and unskilled manual workers, sidestep at the highest rates. This additive effect is offset, however, via the interactions with education (see the next section).

Interaction Effects

The results for the interaction effects between own and parents’ occupation indicate that it is not social origins per se but rather a mismatch between individuals’ occupational attainment and their parents’ occupation that drives rates of upgrading (Model 1.2). Consistent with Hypothesis 4a (and hence, contradicting Hypothesis 4b), downwardly mobile individuals (with respect to origins) are particularly prone to upgrade. Upwardly mobile Russian adults have the lowest hazard of upgrading, presumably because, having already surpassed their origin class, they have less incentive to do so. These findings clearly indicate that upgrading functions more as a means for “catch-up” mobility on the part of higher origin adults whose initial occupations represent downward intergenerational mobility rather than as an opportunity for the disadvantaged to gain ground later in the life course. As noted earlier, net of the origin-occupation interaction, the additive effects of class origins reverse signs and lose statistical significance: It is not higher origins that increase rates of upgrading but the combination of higher origins (with associated greater resources) and low occupational attainment (with associated greater incentives).¹⁵ For sidestepping, the interactions between origins and occupation are not statistically significant (Models 2.2 and 2.3).

Compared to adults whose educational credentials match their occupational class, underqualified adults clearly upgrade at higher rates; overqualified adults exhibit no statistically significant difference in the hazard of upgrading. These findings are in line with Hypothesis 5 (the underqualified are more prone to upgrade, less prone to sidestep) but not Hypothesis 6 (the overqualified are less prone to upgrade, more prone to sidestep). The corresponding effects for sidestepping (Model 2.2) are in line with Hypothesis 6, but the effect is not significant. We speculated that only specific forms of overqualification increase the hazard of sidestepping, and we tested a series of specific combinations.¹⁶ We found that the hypothesized effect is statistically significant only for university-educated adults working in lower routine manual occupations (Model 2.3). Although we should interpret these results with caution, because our initial global measure was not statistically significant and we resorted to testing finer grained specifications, the effect is nonetheless consistent with our intuition. For example, a taxi driver with a university degree in sociology or a sales clerk with a degree in Russian literature may decide to return to university to get a degree more likely to result in professional employment.¹⁷

Regime Change

Using a single dummy variable to distinguish the post-Soviet from the Soviet period yields no evidence that regime change affected rates of upgrading (Model 1.1) or sidestepping (Model 1.2). Thus, neither Hypothesis 7a nor 7b is supported. However, when we divide the post-Soviet period into two distinct periods, we find support for Hypothesis 7c, which predicted decreased rates during the turmoil of 1992 to 1998 (which we find for upgrading and sidestepping) and increased rates from 1999 onward (for upgrading), when the economy recovered and political stability set in. In fact, after plunging during the turmoil period, compared to the Soviet period, the upgrading hazard was 1.4 times higher in the recovery period (Model 1.2). We cannot say which aspect of the recovery period led to the surge in upgrading—relative not only to the turmoil period but also to the Soviet era: It could have been the restoration of economic growth, greater certainty of the durability of market reforms, expanded economic inequality, the general consolidation of post-Soviet institutional changes, some combination of these, or additional factors. But our evidence clearly shows dramatic swings, for upgrading in particular.

Differences between Upgrading and Sidestepping

Finally, our results highlight the importance of differentiating upgrading and sidestepping. Models 3.1 and 3.2 show parameter estimates for models of the hazard to any adult education. These results are broadly consistent with those for upgrading, which is not surprising given that most incidents of adult education enrollment in our data involve upgrading. However, there is a systematic tendency to underestimate the parameters for effects on upgrading if we pool upgrading and sidestepping. Sidestepping is less structured by measures of socioeconomic status, and thus it plays a lesser role in the educational stratification regime. Therefore, combining the two forms of adult education leads to systematic underestimation of the extent to which upgrading is related to resources and incentives. We even observe cases where effects on upgrading and sidestepping are opposite in sign or only pertain to one of the forms—such as the main effect of a specialized secondary degree relative to a general secondary degree (opposite signs) and the effects of underqualification and overqualification.

Conclusions

Upgrading of educational credentials or attaining additional credentials during adulthood will likely become more common, not only in Russia but in other advanced economies. Growing credential-based inequality in labor markets increases the incentives to attain better credentials. Rapid technological change makes some credentials obsolete and others in high demand (Bartel and Sicherman 1998). Population aging creates pressure on societies to retrain workers in adulthood rather than rely on new cohorts of workers to meet demands for new skills. Although it affects most developed societies, population aging is particularly acute in Russia due to demographic trends (Gerber and Radl 2014); this led the Russian government to announce in June 2018 that it will raise the retirement age. Doing so expands the time horizon during which adults can reap returns from improving their education credentials, another impetus increasing the political and economic importance of adult education in the coming decades.

As adult education becomes more common, its significance as a component of educational stratification will only increase, in Russia and elsewhere. Our theoretical framework lays out mechanisms whereby measures of social origin and current socioeconomic standing influence rates of participation in two forms of adult education because of the associated differences in resources and incentives for adult schooling. Our empirical findings from Russia provide evidence that participation in adult education is conditioned by class origins, previous education, and current occupation. These stratification variables generally have stronger effects on rates of upgrading, the more lucrative form of adult education, than sidestepping. In particular, upgrading is most common for groups that have both greater resources and higher incentives to upgrade, such as downwardly mobile children of professionals and managers (whose social origins provide resources and whose lower occupational standing provide incentive) and underqualified workers (whose upper tier occupations offer the resources and whose need for improved credentials to reinforce their position provides incentive). Apart from these groups, Russian adults with an intermediate level of education (general secondary schooling) and those with salariat occupations also upgrade at high rates. We do not find empirical support for the theoretical argument that social selection during initial education favors adults in less privileged situations in upgrading: The data did not support Hypotheses 1b, 2b, and 4b. Overall, upgrading tends to exacerbate patterns of socioeconomic stratification in initial educational attainment by delivering better credentials to individuals who already have higher levels of resources, particularly if they also have strong incentives to upgrade.

Our analysis shows that sidestepping is less common than upgrading and less related to measures of social origin and current socioeconomic standing. The weaker role of sidestepping in the stratification process means analyses that collapse the two forms of adult education into a single undifferentiated phenomenon tend to underestimate the extent to which upgrading reinforces existing sources of socioeconomic advantage. We do find evidence that sidestepping plays such a role independently: Sidestepping is particularly common for Russian adults who attained specialized secondary or university degrees in their initial education. Overqualified Russians are also more likely to sidestep, albeit one particular group of them (university graduates working in lower routine nonmanual occupations). Such individuals likely pursue second university degrees in different fields of study or at more prestigious institutions in the hope of gaining professional or managerial work. One promising avenue for future research would be to investigate whether sidestepping does tend to involve matriculation into more lucrative specializations or more prestigious institutions. This would require detailed data on characteristics of adult education spells that our data set lacks.

Our study also demonstrates that institutional and historical context matters. The dramatic economic shocks, institutional disruptions, and political uncertainty that Russia experienced in the immediate wake of the Soviet collapse led to sharp declines in rates of upgrading and sidestepping relative to their respective baseline rates during the Soviet era. However, once the Russian economy recovered, growth resumed, market institutions proved durable, and the political regime stabilized under Putin, rates of upgrading soared to levels that well exceeded those of the Soviet era. The surge of upgrading during Russia’s post-Soviet recovery suggests adult education helped facilitate the “great human capital re-allocation,” as Sabirianova (2002) described structural changes in the Russian labor market during the 1990s (see also Gerber 2002). An important caveat is that adult education did not begin playing its role as a source of new skills and credentials responding to emerging labor market demands until the turmoil associated with the immediate aftermath of regime change receded and yielded a (relatively) more stable institutional and economic environment.

Cross-national comparative research indicates that adult education rates are shaped by the organization of educational and employment systems (e.g., Blossfeld et al. 2014; Kilpi-Jakonen et al. 2012; Saar, Ure, and Desjardins 2013; Wolbers 2005). These studies find that adult education rates are lower in countries with more stratified, vocationally oriented, and standardized educational systems (e.g., Germany) due to a lower necessity for skills upgrading and compensation for lack of specific skills. Adult education rates tend to be higher in systems oriented toward producing more general skills (e.g., the United States), with a stronger common core in the curriculum that is less time-consuming and demanding for adult learners. The degree of labor market institutionalization and level of employment regulation may also shape participation rates in adult education (Blossfeld et al. 2014; Dieckhoff, Jungblut, and O’Connell 2007; Vono de Vilhena et al. 2016). In general, stronger institutional boundaries, occupational rigidity, and less dynamic labor markets limit opportunities for reaping returns from adult education, making it less attractive. In more liberalized employment systems, labor market success is strongly determined by individual resources, such as job performance, recent education, and employment history, and initial educational attainment recedes into the background (DiPrete et al. 1997).

It is not straightforward to classify Russia into this institutional framework. But generally speaking, after the Soviet collapse, Russia shifted from a rigid, centralized, closed system geared toward production of specific skills toward a destandardized and destratified educational system loosely linked to labor market needs and growing turnover rates (Gerber 2003; Gimpelson, Kapeliushnikov, and Lukiyanova 2010; Knell and Srholec 2007; Kosyakova 2014, 2018). In this sense, our results are in line with predictions from the literature: After an initial shrinkage in the turmoil period, adult education rates began proliferating in post-Socialist Russia. Thus, our findings from a temporal research design are consistent, in broad terms, with findings from cross-national comparisons. Comparative studies examining trends in upgrading and sidestepping in multiple countries may help establish which particular aspects of national context drive surges or declines in upgrading and, possibly, sidestepping.

Our conceptual approach and empirical results provide new insights into the mechanisms linking the two forms of adult education to stratification in education and the labor market in advanced economies. Our study points to the importance of taking adult education into account to understand fully how educational and labor market opportunities are structured at later phases of the life course, which the typical focus on childhood educational attainment overlooks. On top of that, our findings illustrate how adult education can contribute to a life course process of “cumulative advantage” (DiPrete and Eirich 2006): Initial social advantages are reproduced, maintained, and amplified rather than offset by patterns of participation in adult education in Russia. Social origins, initial educational attainment, and occupational status shape access to upgrading and sidestepping in ways that promote broader social inequality.

Adult education, far from a rare experience in most developed economies, will undoubtedly become more prevalent for the reasons discussed at the outset. Accordingly, its impact on education stratification regimes will grow apace. If, as we advocate, social scientists who study educational stratification recognize this growing role, the advantages of distinguishing upgrading from sidestepping and grounding analyses in a comprehensive theoretical framework, rather than narrow analyses of particular types of adult education (e.g., attending university) or ad hoc empirical studies of particular correlates of participation in adult education, should be apparent.

Footnotes

APPENDIX

Table A2.

Comparison of LR Improvement Test.

Model Specification		LL	df	AIC	BIC	$Δ$	LR $χ^{2}$	Pr > $χ^{2}$
Upgrading
M1.1	S, E, O, P, C	−2,665	25	5,384	5,617
M1.2.1	S, E, O, P, S-O-M, C	−2,658	28	5,375	5,634	M1.1	15.17	.002
M1.2.2	S, E, O, P, E-O-M, C	−2,662	27	5,382	5,631	M1.1	6.67	.036
M1.2.4	S, E, O, PD, C	−2,638	26	5,331	5,573	M1.1	54.77	.000
M1.2	S, E, O, S-O-M, E-O-M, PD, C	−2,627	31	5,319	5,604	M1.1	76.79	.000
Sidestepping
M2.1	S, E, O, P, C	−851	25	1,755	1,988
M1.2.1	S, E, O, P, S-O-M, C	−850	28	1,760	2,019	M2.1	.62	.892
M1.2.2	S, E, O, P, E-O-M, C	−850	27	1,758	2,008	M2.1	.95	.621
M1.2.3	S, E, O, P, HE × LRNM, C	−848	26	1,751	1,993	M2.1	5.86	.016
M1.2.4	S, E, O, PD, C	−848	26	1,752	1,993	M2.1	5.25	.022
M2.2	S, E, O, S-O-M, E-O-M, PD, C	−847	31	1,760	2,045	M2.1	6.86	.334
M2.3	S, E, O, S-O-M, HE ×LRNM, PD, C	−845	30	1,753	2,029	M2.1	11.89	.036

Note: Model specification: S = social origin; E = educational level; O = occupational class; S-O-M = own parents’ occupation match; E-O-M = education-occupation match; HE×LRNM = higher educated × lower routine nonmanual; P = period; PD = detailed period; C = controls; interaction indicated by ×; $Δ$ = comparison model; LR = likelihood ratio; AIC = Akaiki Information Criterion; BIC = Bayesian Information Criterion; Pr = probability.

Source: Linked Generations and Gender Survey (GGS; 2004) and the Education and Employment Survey for Russia (EES; 2005) data; authors’ calculations.

Acknowledgements

We thank David Bills and Hans-Peter Blossfeld for providing feedback on the earlier versions of this paper, our anonymous reviewers for their insightful comments, and all those who commented on the earlier presentations of this work at the following events: the RC28 Meeting in Columbia University New York (August 2017) and the Wisconsin Russia Project Young Scholars Workshop at the University of Wisconsin-Madison (July 2018). We also acknowledge the financial support of the European Research Council (ERC) through the Advanced Grant awarded to Hans-Peter Blossfeld and the Wisconsin Russia Project, funded by a grant from the Carnegie Corporation of New York.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by an Advanced Grant from the European Research Council (ERC) to Hans-Peter Blossfeld (Call details ERC-2010-AdG, SH2, Project-ID 269568). Writing was also supported by the Wisconsin Russia Project, funded by a grant from the Carnegie Corporation of New York.

ORCID ID

Y. Kosyakova

Notes

Author Biographies

Yuliya Kosyakova is a postdoctoral senior researcher at the Migration and International Labour Studies Department at the Institute for Employment Research (IAB), Nuremberg, Germany. She received her PhD in sociology at the European University Institute (EUI), Florence, Italy. She conducts research on social stratification, (refugee) migration and integration, institutional change in Russia and other former socialist countries, international comparisons, gender, and the life course.

Theodore P. Gerber is a professor of sociology and director of the Center for Russia, East, Europe and Central Asia and the Wisconsin Russia Project at the University of Wisconsin-Madison. He conducts research on social stratification, demographic change, housing, and public opinion in Russia and other former Soviet countries.

References

AEC. 2018. “Mutual Recognition of Qualifications: The Russian Federation and the Other European Countries.” Retrieved February 1, 2018 (http://www.aic.lv/ace/tools/leg_aca/guid_rus.htm).

Bartel

Ann P.

Sicherman

Nachum

. 1998. “Technological Change and the Skill Acquisition of Young Workers.” Journal of Labor Economics 16(4):718–55.

Bills

David B.

2000. Social, Demographic, Economic, and Technological Trends Affecting Lifelong Learning. Lifelong Learning NCES Task Force: Final Report, Vol. II. Washington, DC: U.S. Department of Education.

Blossfeld

Hans-Peter

Golsch

Katrin

Rohwer

Götz

. 2007. Event History Analysis with Stata. New York: Taylor & Francis.

Hans-Peter

Blossfeld

Kilpi-Jakonen

Elina

Vono de Vilhena

Daniela

Buchholz

Sandra

, eds. 2014. Adult Learning in Modern Societies: An International Comparison from a Life-course Perspective. EduLIFE Lifelong Learning Series. Cheltenham, UK, and Northampton, MA: Edward Elgar.

Breen

Richard

Goldthorpe

John H.

1997. “Explaining Educational Differentials: Towards a Formal Rational Action Theory.” Rationality and Society 9(3):275–305.

Breen

Richard

Jonsson

Jan O.

2005. “Inequality of Opportunity in Comparative Perspective: Recent Research on Educational Attainment and Social Mobility.” Annual Review of Sociology 31(1):223–43.

Bühler

Christoph

Magun

Vladimir

Kozyreva

Polina M.

Kosolapov

Mikhail S.

Sinyavskaya

Oksana

Shkolnikov

Vladimir

Kulu

Hill

Vikat

Andres

Houle

Rene

. 2007. The Education and Employment Survey for Russia: Survey Instruments. Rostock: Max Planck Institute for Demographic Research.

Carr

Deborah

Sheridan

Jennifer T.

1999. Back to School: A Life Course Approach to Understanding Men’s and Women’s College Attendance in Adulthood. University of Michigan Population Studies Center, Report No. 99-437.

10.

Connor

Walter D.

1991. The Accidental Proletariat: Workers, Politics, and Crisis in Gorbachev’s Russia. Princeton, NJ: Princeton University Press.

11.

Dieckhoff

Martina

Jungblut

Jean-Marie

O’Connell

Philip J.

2007. “Job-related Training in Europe: Do Institutions Matter?” Pp. 77–103 in Employment Regimes and the Quality of Work, edited by Gallie

Oxford: University Press.

12.

DiPrete

Thomas A.

Eirich

Gregory M.

2006. “Cumulative Advantage as a Mechanism for Inequality: A Review of Theoretical and Empirical Developments.” Annual Review of Sociology 32(1):271–97.

13.

DiPrete

Thomas A.

De Graaf

Paul M.

Luijkx

Ruud

Tahlin

Michael

Blossfeld

Hans-Peter

. 1997. “Collectivist Versus Individualist Mobility Regimes? Structural Change and Job Mobility in Four Countries.” American Journal of Sociology 103(2):318–358.

14.

Elder

Glen H.

Jr. Johnson

Monica Kirkpatrick

Crosnoe

Robert

. 2003. “The Emergence and Development of Life Course Theory.” Pp. 3–19 in Handbook of the Life Course, edited by Mortimer

J. T.

Shanahan

M. J.

New York: Kluwer Academic Publishers.

15.

Elman

Cheryl

O’Rand

Angela M.

1998. “Midlife Entry into Vocational Training: A Mobility Model.” Social Science Research 27(2):128–58.

16.

Elman

Cheryl

O’Rand

Angela M.

2004. “The Race Is to the Swift: Socioeconomic Origins, Adult Education, and Wage Attainment.” American Journal of Sociology 110(1):123–60.

17.

Elman

Cheryl

O’Rand

Angela M.

2007. “The Effects of Social Origins, Life Events, and Institutional Sorting on Adults’ School Transitions.” Social Science Research 36(3):1276–99.

18.

Felmlee

Diane H.

1988. “Returning to School and Women’s Occupational Attainment.” Sociology of Education 61(1):29–41.

19.

Gerber

Theodore P.

2000. “Educational Stratification in Contemporary Russia: Stability and Change in the Face of Economic and Institutional Crisis.” Sociology of Education 73(4):219–46.

20.

Gerber

Theodore P.

2002. “Structural Change and Post-socialist Stratification: Labor Market Transitions in Contemporary Russia.” American Sociological Review 67(5):629–59.

21.

Gerber

Theodore P.

2003. “Loosening Links? School- to-work Transitions and Institutional Change in Russia since 1970.” Social Forces 82(1):241–76.

22.

Gerber

Theodore P.

Hout

Michael

. 1995. “Educational Stratification in Russia during the Soviet Period.” American Journal of Sociology 101(3):611–60.

23.

Gerber

Theodore P.

Hout

Michael

. 2004. “Tightening up: Declining Class Mobility during Russia’s Market Transition.” American Sociological Review 69(5):677–703.

24.

Gerber

Theodore P.

Radl

Jonas

. 2014. “Pushed, Pulled, or Blocked? The Elderly and the Labor Market in Post-Soviet Russia.” Social Science Research 45:152–69.

25.

Gimpelson

Vladimir

Kapeliushnikov

Rostislav

Lukiyanova

Anna

. 2010. “Employment Protection Legislation in Russia: Regional Enforcement and Labor Market Outcomes.” Comparative Economic Studies 52(4):611–36.

26.

GOSKOMSTAT USSR. 1981. НародноеХозяйствоСССРв 1980 г. СтатистическийЕжегодник [National Economy of the USSR in 1980. Statistical Yearbook]. Moscow: Finansi is statistika.

27.

Gregory

Paul R.

Collier

Irwin L.

1988. “Unemployment in the Soviet Union: Evidence from the Soviet Interview Project.” The American Economic Review 78(4):613–32.

28.

GU VShE. 2010. ОбразованиевРоссийскойФедерации: СтатистическийСборник [Education in the Russian Federation: Statistical Collection]. Moscow: GU-HSE.

29.

Hällsten

Martin

. 2011. “Late Entry in Swedish Tertiary Education: Can the Opportunity of Lifelong Learning Promote Equality Over the Life Course?” British Journal of Industrial Relations 49(3):537–59.

30.

Hällsten

Martin

. 2012. “Is It Ever Too Late to Study? The Economic Returns on Late Tertiary Degrees in Sweden.” Economics of Education Review 31(1):179–94.

31.

Houle

Rene

Shkolnikov

Vladimir

. 2005. Low Response Rates in the Cities of Moscow and Sankt- Petersburg and GGS-Census Comparisons of Basic Distributions. Rostock: Max Planck Institute for Demographic Research.

32.

IISP. 2014. “Parents and Children, Men and Women in Family and Society. Project within the Pan-European ‘Generations and Gender Program/Survey.’” Retrieved (http://www.socpol.ru/eng/research_projects/proj12.shtml).

33.

IQAS. 2009. International Education Guide. For the Assessment of Education from the Former USSR and the Russian Federation. Alberta: International Qualifications Assessment Service.

34.

Jacobson

Louis

LaLonde

Robert J.

Sullivan

Daniel

. 2005. “The Impact of Community College Retraining on Older Displaced Workers: Should We Teach Old Dogs New Tricks?” Industrial and Labor Relations Review 58(3):398–415.

35.

Jarvis

Peter

. 2004. Adult Education and Lifelong Learning. Theory and Practice. 3rd ed. London: RoutledgeFalmer.

36.

Jarvis

Peter

. 2007. Globalisation, Lifelong Learning and the Learning Society. Sociological Perspectives. New York: Routledge.

37.

Jenkins

Andrew

Vignoles

Anna

Wolf

Alison

Galindo-Rueda

Fernando

. 2003. “The Determinants and Labour Market Effects of Lifelong Learning.” Applied Economics 35(16):1711–21.

38.

Jepsen

Christopher

Montgomery

Mark

. 2012. “Back to School: An Application of Human Capital Theory for Mature Workers.” Economics of Education Review 31(1):168–78.

39.

Kerckhoff

Alan C.

1995. “Institutional Arrangements and Stratification Processes in Industrial Societies.” Annual Review of Sociology 21:323–47.

40.

Khokhlova

Anisya

Kozlovskiy

Vladimir

Veits

Maria

. 2013. “Formal Adult Education in the Context of the Transforming Labour Market in Russia.” Pp. 327–349 in Lifelong Learning in Europe; National Patterns and Challenges, edited by Saar

Ure

O. B.

Holford

Cheltenham, UK, Northampton, MA: Edward Elgar.

41.

Kilpi-Jakonen

Elina

Vono de Vilhena

Daniela

Kosyakova

Yuliya

Stenberg

Anders

Blossfeld

Hans-Peter

. 2012. “The Impact of Formal Adult Education on the Likelihood of Being Employed: A Comparative Overview.” Studies of Transition States and Societies 4(1):48–68.

42.

Kljucharev

Grigorij

. 1997. “Observing Adult Education in Contemporary Russia.” Community Development Journal 32(3):280–86.

43.

Knell

Mark

Srholec

Martin

. 2007. “Diverging Pathways in Central and Eastern Europe.” Pp. 40–64 in Varieties of Capitalism in Post-communist Countries, edited by Lane

Myant

London: Palgrave Macmillan.

44.

Konstantinovskiy

David L.

2017. “Expansion of Higher Education and Consequences for Social Inequality (the Case of Russia).” Higher Education 74(2):201–20.

45.

Kosyakova

Yuliya

. 2014. “Job-related Adult Learning in the Russian Federation: More Educational Opportunities without an Equalization Effect.” Pp. 140–61 in Adult Learning in Modern Societies: Patterns and Consequences of Participation from a Life-course Perspective, edited by Blossfeld

H.-P.

Kilpi-Jakonen

Vono de Vilhena

Buchholz

Cheltenham, UK, Northampton, MA: Edward Elgar.

46.

Kosyakova

Yuliya

. 2018. “Cumulation or Compensation? Returns to Adult Education and Social Inequalities in Soviet and Post-Soviet Russia.” European Societies 20(1):65–88.

47.

Jiang Hong

König

Markus

Buchmann

Marlis

Sacchi

Stefan

. 2000. “The Influence of Further Education on Occupational Mobility in Switzerland.” European Sociological Review 16(1):43–65.

48.

Light

Audrey

. 1995. “Hazard Model Estimates of the Decision to Reenroll in School.” Labour Economics 2:381–406.

49.

Mayer

Karl Ulrich

. 2006. “After the Fall of the Wall: Living through the Post-socialist Transformation in East Germany.” Pp. 1–28 in After the Fall of the Wall: Life Courses in the Transformation of East Germany, edited by Diewald

Goedicke

Mayer

K. U.

Stanford: Stanford University Press.

50.

Merriam

Sharan B.

Caffarella

Rosemary S.

1998. Learning in Adulthood: A Comprehensive Guide. 2nd ed. San Francisco, CA: Jossey-Bass Publishers.

51.

Merton

Robert K.

1968. “The Matthew Effect in Science: The Reward and Communication Systems of Science Are Reconsidered.” Science 159(3810):56–63.

52.

Purcell

Kate

Wilton

Nick

Elias

Peter

. 2007. “Hard Lessons for Lifelong Learners? Age and Experience in the Graduate Labour Market.” Higher Education Quarterly 61(1):57–82.

53.

Richardson

John T. E.

King

Estelle

. 1988. “Adult Students in Higher Education.” The Journal of Higher Education 69(1):65–88.

54.

Saar

Ellu

Ure

Odd Bjørn

Desjardins

Richard

. 2013. “The Role of Diverse Institutions in Framing Adult Learning Systems.” European Journal of Education 48(2):213–32.

55.

Sabirianova

Klara Z.

2002. “The Great Human Capital Reallocation: An Empirical Analysis of Occupational Mobility in Transitional Russia.” Journal of Comparative Economics 30(1):191–217.

56.

Schömann

Klaus

Becker

Rolf

. 1995. “Participation in Further Education over the Life Course: A Longitudinal Study of Three Birth Cohorts in the Federal Republic of Germany.” European Sociological Review 11(2):187–208.

57.

S. V.

Shishkin

, ed. 2004. ДоступностьВысшегоОбразованиявРоссии [Access to Higher Education in Russia]. Moscow: Nezavisimyj institut social’noj politiki.

58.

Sinyavskaya

Oxana.

2012. Managing Non-response in the Russian and in East European Context. Presentation made at the SCOHOST’s Seminar Managing Problems with Missing Data in Survey Research, Sigtuna, Sweden, 10.09.2012.

59.

Smolentseva

Anna

. 2017. “Universal Higher Education and Positional Advantage: Soviet Legacies and Neoliberal Transformations in Russia.” Higher Education 73(2):209–26.

60.

Stenberg

Anders

. 2011. “Using Longitudinal Data to Evaluate Publicly Provided Formal Education for Low Skilled.” Economics of Education Review 30(6):1262–80.

61.

United Nations. 2005. Generations & Gender Programme: Survey Instruments. New York and Geneva: United Nations.

62.

Valentine

Thomas

Darkenwald

Gordon G.

1990. “Deterrents to Participation in Adult Education: Profiles of Potential Learners.” Adult Education Quarterly 41(1):29–42.

63.

Vono de Vilhena

Daniela

Kosyakova

Yuliya

Kilpi-Jakonen

Elina

McMullin

Patricia

. 2016. “Does Adult Education Contribute to Securing Non-precarious Employment? A Cross-national Comparison.” Work, Employment & Society 30(1):97–117.

64.

Wolbers

Maarten H. J.

2005. “Initial and Further Education: Substitutes or Complements? Differences in Continuing Education and Training over the Life- course of European Workers.” Review of Education 51:459–78.

65.

Zajda

Joseph

. 2003. “Lifelong Learning and Adult Education: Russia Meets the West.” International Review of Education 49(1–2):111–32.

Adult Education,Stratification,and Regime Change: Upgrading and Sidestepping in Russia,1965–2005

Abstract

Keywords

Educational Stratification in Soviet and Post-Soviet Russia

Adult Education

Upgrading Versus Sidestepping

Theory and Hypotheses

Measures of Social Position: Resources, Incentives, and Selection

Social origins

Educational level

Occupational class

Measures of Social Position: Interactions

Social origin and occupational class

Educational level and occupational class

Regime Change

Data and Method

Data and Sample

Variables

Dependent variables

Independent variables

Controls

Method

Results

Enrollment Rates in Adult Education in Russia

Transition to Upgrading and Sidestepping: Multivariate Results

Additive Effects of Social Position

Interaction Effects

Regime Change

Differences between Upgrading and Sidestepping

Conclusions

Footnotes

APPENDIX

Acknowledgements

Funding

ORCID ID

Notes

Author Biographies

References