Unequal Returns: How Regional Origin Shapes Intergenerational Mobility in Kazakhstan

Introduction

Intergenerational mobility refers to the relationship between the socioeconomic status of parents and that of their children, often serving as an indicator of societal openness and equality of opportunity. Within this context, education is frequently called the “great equalizer” (Blanden & Macmillan, 2014; Hout, 2015; Torche, 2019), while internal migration is seen as a strategic means for social advancement (Fatimah & Kofol, 2023; Fielding, 1992; Michelangeli & Türk, 2021). Nevertheless, increasing evidence indicates that these mechanisms are not fully effective. Structural factors such as social background and geographical origin (Bernardi & Ballarino, 2016; Chetty et al., 2014) continue to influence life chances, even after migration and educational attainment. This issue is particularly pronounced in developing and post-Soviet societies, where the transition to a market economy has exacerbated regional inequalities, such as the rural-urban divide, and restricted access to opportunities (Reimer & Pollak, 2010; Zhussupova, 2016). In Kazakhstan, despite educational expansion and increased internal mobility, these processes may actually reinforce rather than mitigate existing social hierarchies.

The regional disparities in the rural-urban divide (Zhussupova, 2016) have triggered a significant flow of internal migrants to Almaty, making it a key destination for educational and economic advancement (Aldashev & Dietz, 2014; Makhmutova, 2012). However, internal migration alone is insufficient to eliminate the (dis)advantages tied to social origin. Migrants carry with them their entire socioeconomic backgrounds, whether advantageous or not, which can influence their outcomes in the city. In this context, a rural origin can be seen as a structural disadvantage. Rural areas often lack developed infrastructure, offer lower-quality basic education, and typically house working-class families. In contrast, urban-origin migrants are more likely to access better schools and a wider range of educational choices, often coming from intermediate-class households. These origin-based differences result in unequal starting points, suggesting that while migration may provide access to opportunities, it does not guarantee an equal ability to benefit from them. ¹

Since gaining independence, Kazakhstan has expanded access to higher education through state scholarships, the growth of private universities, and the introduction of Kazakh-language instruction (Mynbayeva & Pogosian, 2014). Although this expansion has increased formal access, it has also intensified existing inequalities. Students from urban, affluent families are more likely to attend prestigious universities and utilize family resources, while many students from rural areas remain concentrated in lower-tier universities with limited opportunities for upward mobility (Roberts et al., 2009; Shnarbekova, 2021). Thus, although education is often viewed as a great equalizer, it does not entirely fulfill that promise and continues to reproduce social hierarchies based on region and class. While this study does not directly examine institutional hierarchy, these contextual factors are essential for interpreting patterns of intergenerational mobility.

While higher education is often believed to enhance intergenerational mobility, research increasingly indicates that the returns to education differ based on an individual’s social and spatial origin. In other words, obtaining a degree does not ensure equal outcomes for all. Migrants from rural and working-class backgrounds may encounter limited occupational mobility despite obtaining higher education, due to factors such as institutional bias, labor market segmentation, or a lack of social capital (Bernardi & Ballarino, 2016; Shnarbekova, 2021). This study specifically examines these unequal returns to education as a primary mechanism through which initial disadvantages persist, even among upwardly mobile individuals. It thereby questions the assumption that education alone can serve as a complete equalizer of opportunity in the context of internal migration.

This study specifically focuses on cumulative disadvantage rather than the broader concept of cumulative (dis)advantage, while recognizing that the theoretical framework encompasses bidirectional processes (DiPrete & Eirich, 2006). Although urban migrants possess relative advantages compared to their rural counterparts, including better educational preparation, stronger parental human capital, and familiarity with urban institutional environments, still both migrant groups can experience disadvantages relative to non-migrant urban residents from established middle-class families. Urban migrants from working-class backgrounds, despite their comparative privilege, may still face barriers such as weaker social networks, limited access to elite educational institutions, and residential concentration in less advantaged neighborhoods compared to the non-migrant middle class. However, this study’s analytical focus on cumulative disadvantage is justified on two grounds. First, empirically, the most pronounced and policy-relevant inequalities emerge among rural migrants, who face compounding barriers from poor-quality rural schooling, predominantly working-class origins, and constrained family resources (Makhmutova, 2012; Zhussupova, 2016). These initial deficits persist and intensify through residential segregation in peripheral districts, concentration in lower-tier universities, and differential labor market returns to education. Second, theoretically, examining cumulative disadvantage allows to trace how structural inequalities become entrenched across life course transitions, which is a central concern for understanding social reproduction in transitional societies. While urban migrants’ trajectories may demonstrate how moderate advantages facilitate successful integration, the cumulative disadvantage lens more precisely captures the mechanisms that perpetuate inequality for the most structurally vulnerable groups, which is the primary analytical and policy concern of this study.

Understanding how social origin influences intergenerational mobility is crucial for assessing whether internal migration and education provide equal opportunities. In Kazakhstan, rural areas face structural constraints, including weaker schools, limited access to extracurricular activities, and fewer family resources. These disadvantages persist even after migration, affecting migrants’ ability to access higher education and convert educational credentials into occupational success. Despite the expansion of education and increased internal mobility in Kazakhstan, empirical research has seldom explored how these processes impact intergenerational mobility across different regional backgrounds. Most studies isolate either migration trends or educational outcomes, neglecting the persistence of origin-based disparities across generations.

This study addresses this gap by examining educational and occupational mobility among first-generation internal migrants in Almaty, comparing those from rural areas with their counterparts from other urban regions. The analysis investigates whether these groups follow distinct mobility trajectories and how educational attainment mediates the relationship between origin and mobility outcomes.

Rural-origin migrants often have lower intergenerational educational and occupational mobility compared to their urban-origin peers. This disparity is mainly due to starting with lower parental education levels, weaker economic resources, and limited access to high-quality schools. Although many rural-origin individuals do attain higher education, they are frequently funneled into lower-tier or less prestigious programs. Consequently, the potential of their degrees to enhance their status is diminished. Even when rural and urban migrants possess similar qualifications, those from rural backgrounds encounter difficulties entering high-status occupations. This suggests that entrenched structural disadvantages, rather than individual effort, continue to influence their educational and occupational outcomes.

In this framework, education is viewed not as a universal equalizer but as a conditional one. Its capacity to reduce disparities based on origin is influenced by institutional context, regional opportunity structures, and family background. The analysis concentrates on identifying both observable structural disparities and the persistent disadvantages that persist even after these factors are considered. This approach enables us to assess the extent to which regional origin continues to shape life chances in a transforming post-Soviet society.

Literature Review

The literature review offers a systematic overview of existing research on mechanisms influencing the intergenerational mobility of internal migrants. Cumulative (dis)advantage theory, initially introduced by Merton (1968) and further developed by DiPrete and Eirich (2006), serves as the theoretical framework for analyzing inequalities in intergenerational mobility processes. This theory posits that socioeconomic (dis)advantages tend to accumulate over time, a phenomenon Merton termed the “Matthew effect,” implying that the rich become richer while the poor become poorer. Rural migrants often begin their urban integration with disadvantages such as lower education levels and origins in predominantly working-class families with limited economic resources (Makhmutova, 2012; Zhussupova, 2016). These disadvantages create reinforcing cycles that hinder advancement, even after moving to cities with significant opportunities. Conversely, urban migrants may possess more advantages at the onset of their migration due to access to extracurricular activities, better-equipped schools, and middle-class family resources. According to DiPrete & Eirich (2006), systematic disparities in access to opportunity structures lead to the accumulation of (dis)advantages. For migrants, this means that their initial resources and social origins influence their access to quality housing, schools, professional networks, and job prospects in new locations.

While cumulative (dis)advantage theory captures the temporal dimension of how inequalities compound over time, intersectionality theory (Crenshaw, 1989) complements this framework by highlighting how multiple social categories such as rural origin, class background, gender and ethnicity intersect to create distinct mobility experiences. Intersectionality posits that social identities do not operate independently but interact to produce qualitatively unique forms of advantage and disadvantage (Collins, 2015; McCall, 2005). For example, rural origin women from working class families may face compounded barriers that differ fundamentally from those experienced by urban origin men or even rural-origin individuals from intermediate classes. In the context of internal migration, this means that the disadvantages associated with rural origin are not uniform but vary depending on how they intersect with other axes of stratification. Although a full intersectional analysis is beyond the scope of this study, recognizing these intersections helps explain variations in mobility outcomes within origin groups and suggests that origin-based disadvantages operate in conjunction with (and are potentially amplified by) other dimensions of social inequality.

Expanding on the cumulative (dis)advantage theory, research on regional inequality suggests how geographic location influences life chances and mobility opportunities. Fielding (1992) introduced the concept of an escalator region, positing that certain cities can facilitate advancements in mobility. Almaty exemplifies such a city, with its 42 higher education institutions, a diverse labor market, and a GDP accounting for 21% of the national total (Bureau of National Statistics of the Republic of Kazakhstan, 2024). Zhussupova (2016) discusses regional disparities, particularly the rural-urban divide, highlighting significant differences in socioeconomic development. For instance, the gross regional product per capita varies by a factor of 7.5 between the most and least developed regions, and average monthly wages differ by 2.5 times. Michelangeli & Türk (2021) underscore the critical role of educational migration in intergenerational mobility, noting that students who relocate for education have greater opportunities for upward mobility. The characteristics of the destination city are more influential in determining mobility than those of the origin city.

Research from other developing countries provides additional insights. Fatimah & Kofol (2023) demonstrate that parental migration positively impacts children’s futures, particularly in terms of educational attainment and adult income. Nonetheless, intergenerational mobility improves only under specific conditions, such as when migrant children live in urban areas, originate from the poorest families, and migrate during childhood. Conversely, Chetty et al. (2014) provide evidence for the idea of a “land of opportunity,” showing that childhood neighborhood characteristics significantly affect adult outcomes. This regional dimension is crucial in the context of internal migration, where migrants bring forward the disadvantages and advantages of their places of origin.

Moreover, Galster and Sharkey (2017) introduce the concept of spatial foundations to explain how residential characteristics influence individual outcomes through institutional resources, social networks, the physical environment, and cultural norms. Although these factors fall outside the primary focus of this research, they offer a broader perspective on how spatial characteristics affect life trajectories. For internal migrants, these spatial foundations operate both in the places of origin, shaping pre-migration human capital development, and in the destination places, influencing post-migration integration and access to opportunities.

The role of education in intergenerational mobility has been extensively studied, with competing perspectives on its equalizing mechanisms. Human capital theory, as posited by Becker and Tomes (1986), suggests that education acts as the primary mechanism for intergenerational mobility by providing credentials that translate into economic returns. However, this perspective has been challenged by research demonstrating the persistent effects of social origin, even after controlling for educational attainment. Bernardi and Ballarino (2016) introduce the concept of direct effects of social origin (DESO), showing that family background continues to influence outcomes independent of educational achievements. These theories are primarily applicable to developed Western nations, where the educational system has remained stable for a significant period. In contrast, Kazakhstan has experienced changes in its educational system, which may alter its role as an equalizing force.

Kazakhstan’s educational reforms following independence led to an expansion of higher education, both vertically and horizontally. This expansion was facilitated by initiatives such as rural voucher systems (kvota), scholarship programs, and Kazakh-language tutoring, which improved access to higher education for various groups in Kazakhstan. However, (Roberts et al., 2009) and (Shnarbekova, 2021) note that this expansion resulted in new stratification within the higher education system. (Shnarbekova, 2021) discovered that while access was broadening, there was an increasing differentiation in the quality and prestige of universities. By the 2015-2016 academic year, 68.9% of students were enrolled on a fee-paying basis. This statistic underscores the reliance on family resources, rather than state support, to access educational opportunities. Consequently, family capital continues to heavily influence educational trajectories. Students from privileged backgrounds tend to pursue high-quality education, whereas those from less advantaged backgrounds prioritize social factors such as free tuition, proximity to home, or dormitory availability. Zhanbyrbayeva et al. (2023) add that basic educational quality in rural areas lags significantly behind that in urban areas, creating competition for rural-origin students seeking admission to prestigious universities. These circumstances may contribute to what Zhang (2017) describes as a new educational poverty trap, characterized by inequality between urban and rural school quality, which could hinder the successful integration of migrants into urban educational systems.

The notion of education as the great equalizer (Hout, 2015) is questioned when considering the disparities in educational quality between rural and urban areas. While education does enhance intergenerational mobility, Bernardi and Ballarino (2016) demonstrate that simply expanding educational opportunities does not automatically increase mobility. Instead, persistent effects of social origin indicate that education acts as a partial rather than complete equalizer, with its impact varying according to family background and structural context. These patterns of partial equalization highlight differential returns to education, which are crucial for understanding mobility outcomes. Even when rural and urban migrants attain similar educational credentials, their labor market returns may differ. Factors such as educational quality, field of study, employer perceptions, and network effects contribute to these differences. This mechanism of differential returns explains why educational attainment alone cannot fully offset origin-based advantages or disadvantages. Rural migrants often experience lower returns on equivalent educational investments compared to their urban-origin counterparts, reflecting enduring structural inequalities that persist beyond migration and educational achievement.

The literature on cumulative (dis)advantage, social-origin effects, education as a “great equalizer,” and internal migration as a path to social advancement suggests that three key elements in the mobility chain require simultaneous examination: the resources migrants inherit from their parents, the education they achieve, and the status they ultimately secure in the urban labor market. In this study, each of these elements receives a specific indicator. Parental resources are measured through the highest level of education attained, occupational class categorized using the Eriksson, Goldthorpe, and Portero (EGP) schema (professional-managerial, intermediate, working-class) (Goldthorpe, 2013), and a three-step self-reported living-standard scale distinguishing “survival,” “basic needs,” and “affluent” households. Intergenerational educational mobility is assessed by comparing migrants’ years of schooling and highest diploma to those of their parents, leading to both absolute gaps and an “upward–stable–downward” mobility trichotomy. Returns to education are evaluated through attained occupational class, again utilizing the EGP categories, allowing for the examination of whether the same diploma propels rural- and urban-origin migrants into different tiers of the occupational hierarchy.

This study, grounded in the theories of cumulative disadvantage and the direct effects of the social origin framework, hypothesizes that initial resource disparities significantly influence educational and occupational trajectories among internal migrants in Kazakhstan. We anticipate, first, that the intergenerational transmission of education and occupational status will be more pronounced among rural-origin migrants compared to their urban counterparts, reflecting a heavier reliance on parental resources in environments with limited opportunities. Second, we propose that differences in access to high-status occupational classes will primarily be mediated by migrants’ educational attainment and their parents’ occupational class. Consequently, the direct effect of rural origin is expected to become negligible once these factors are controlled. Lastly, while higher education is predicted to reduce the risk of downward class mobility for all migrants, rural-origin individuals are expected to retain a modestly elevated risk. This suggests that, although education offers protection, it cannot entirely eradicate the structural disadvantages rooted in origin. Collectively, these hypotheses highlight the enduring influence of social origin and the partial buffering effect of education on mobility outcomes.

The following section outlines the stratified sample comprising 455 first-generation migrants, explains the construction of each variable, and describes the regression models employed to test these claims.

Methodology

Analytical Strategy

This study investigates the influence of internal migrants’ places of origin—rural or urban—on intergenerational educational and occupational mobility. The analysis employs the theory of cumulative (dis)advantage (DiPrete & Eirich, 2006). This theory proposes that initial disparities in access to resources and opportunities tend to compound over time, leading to diverse life trajectories. By adopting the cumulative (dis)advantage framework, this study seeks to reveal how early-life conditions, including social origin and parental status, interact with migration experiences. It also aims to demonstrate how these factors perpetuate or alter social inequality across generations.

To address potential confounding and mediating factors, often termed social origins, we included a comprehensive set of covariates. These encompassed sociodemographic characteristics (age, gender, ethnicity), cultural factors (language use and participation in extracurricular activities during childhood), parental background (educational attainment and occupational class), and migration-specific variables (length of residence in Almaty and district of settlement). Additionally, consumption capacity—both before and after relocation—served as a proxy for maternal well-being and was included in the analysis.

The analysis proceeded in four stages, each aligned with the logic of cumulative (dis)advantage, aiming to identify critical points where disparities begin to diverge. In the first stage, the characteristics of the two populations were compared using univariate logistic regression. Here, place of origin (rural vs. urban) served as the dependent variable, while all covariates functioned as predictors. This step highlighted structural and cultural differences between rural- and urban-origin migrants, identifying early sources of (dis)advantage that could accumulate over time. ¹¹ In all models, the urban population was used as the reference group.

To investigate the differences in educational outcomes between rural and urban migrants, we analyzed the respondents’ total years of schooling. We conducted independent t-tests to compare the average years of education between these two origin groups. To assess the strength of the association between parental and respondent education, we utilized Pearson correlation coefficients. General linear models were employed to estimate educational elasticity and examine how children’s education levels vary in response to parental education (see Table 1). Furthermore, we calculated Gini coefficients to capture the overall inequality in years of schooling within each group (Figure 1).

OPEN IN VIEWER

Table 1.

T-Test, Pearson Correlations, and Elasticity Estimates by Migrant Origin

Difference in average years of schooling (T-test results)
Variable	F value	p-value
Migrant	2.01	<.0001
Father	1.5	0.0049
Mother	1.21	0.1533
Parents	1.3	0.0515

Pearson correlation between migrant’s and parents’ years of schooling
Origin	Parent	Coefficient	p-value
Rural	Father	0.41364	<.0001
Urban	​	0.18326	0.0123
Rural	Mother	0.28338	<.0001
Urban	​	0.35281	<.0001
Rural	Parents	0.35587	<.0001
Urban	​	0.31582	<.0001

Education elasticity estimates based on parental education
Origin	Parents	Estimate	Standard error	p-value
Rural	Father	0.5497	0.09	<.0001
Urban	​	0.1428	0.06	0.0123
Rural	Mother	0.3367	0.08	<.0001
Urban	​	0.2689	0.05	<.0001
Rural	Parents	0.4881	0.09	<.0001
Urban	​	0.2699	0.05	<.0001

Note. Average differences in years of schooling were assessed using independent sample t-tests. Pearson correlation coefficients measure the linear association between migrants’ and parents’ education. Elasticity estimates are derived from generalized linear regressions of migrant education on parental education, calculated separately for rural and urban samples.

OPEN IN VIEWER

Figure 1.

(a) and (b). Educational inequality among rural and urban migrants (Lorenz curve). Note. Educational attainment is measured in completed years of schooling. Gini coefficients are calculated based on cumulative education shares relative to population shares. A lower Gini coefficient indicates a more equal distribution

Third, we employed the chi-square test to determine if the distribution of educational and occupational class mobility significantly differed between rural and urban migrants (Tables 2 and 3). Subsequently, we conducted univariable logistic regression models, stratified by origin, to identify which covariates significantly influenced the mobility outcomes. ¹²

OPEN IN VIEWER

Table 2.

Distribution of Educational Attainment by Origin (Chi-Square Test)

Educational mobility	Chi-square value	p-value
Downward, retained disadvantage or stable vs. Upward or retained advantage	14.81	0.0001

OPEN IN VIEWER

Table 3.

Distribution of Occupational Class Stratified by Origin (Chi-Square Test)

Occupational class	Chi-square value	Adjusted p-value*
Downward or retained disadvantage	13.29	0.0009
Stable intermediate	1.74	0.5631
Upward or retained advantage	18.49	0.0003

*p-values adjusted using Bonferroni correction for three comparators (α = 0.005/3 = 0.017).

Finally, we simultaneously conducted multivariable logistic regression analyses to investigate the association between multiple independent variables and mobility outcomes, adjusting for potential confounders (Tables 4 and 5).

OPEN IN VIEWER

Table 4.

Multivariable Model Addressing Factors Associated with Educational Mobility Stratified by Origin

​	Rural	Urban
	OR (95% CI)	OR (95% CI)
Consumption capacity after relocation (vs high capacity)
Basic	3.8 (1.1, 13.0)*	4.5 (1.0, 19.8)*
Intermediate	3.3 (1.0, 10.7)*	0.5 (0.1, 2.1)
Migrant’s occupational class (vs working class)
Intermediate	0.2 (0.1, 0.6)*	0.1 (0.0, 0.4)*
Salariat	0.0 (0.0, 0.1)*	0.0 (0.0, 0.1)*
Parents’ highest occupational class (vs working class)
Intermediate	1.0 (0.4, 2.5)	0.1 (0.0, 0.3)*
Salariat	1.1 (0.4, 2.9)	0.1 (0.0, 0.4)*
Migrants’ distribution by Almaty districts (vs Prestigious areas)
Peripheral areas	5.3 (1.3, 21.1)*	1.3 (0.3, 6.4)
Professional middle-class areas	3.1 (0.6, 14.5)	2.5 (0.7, 9.7)

*Statistically significant, p-value ≤ 0.005.

OPEN IN VIEWER

Table 5.

A Multivariable Model Investigating the Association of Social Class and Origin was Adjusted for Educational Variables Compared to Stable Intermediate Mobility

Factors	Downward or retained disadvantage	Upward or retained advantage
	OR (95% CI)	OR (95% CI)
Origin (Rural vs Urban)
Rural	1.8 (0.9, 3.4)	0.9 (0.5, 1.6)
Migrant’s educational level (vs high school or less)
Associate degree	0.4 (0.1, 1.1)	6.0 (0.6, 55.7)
University degree or higher	0.2 (0.1, 0.6)*	21.3 (2.4, 191.2)*
Parents’ highest educational level (vs high school or less)
Associate degree	0.3 (0.1, 0.9)*	0.4 (0.1, 1.0)
University degree or higher	3.2 (1.0, 9.6)*	2.4 (0.8, 7.1)

*Statistically significant, p-value ≤ 0.005.

Data analysis was conducted using SAS version 9.4 (SAS Institute Inc., Cary, NC, USA).

Results

Educational Attainment

As Figure 1(a) and (b) illustrates, the Gini coefficients for both populations were near zero, signifying a low level of educational inequality. Nonetheless, educational inequality was slightly more pronounced in the rural population, with a coefficient of 0.091, compared to 0.061 for the urban population.

The practical significance of this gap is evident in average educational attainment levels (Figure 2). Urban migrants typically complete 14.6 years of education, whereas rural migrants complete 13.6 years. A similar pattern emerges among parents: urban parents have an average of 13.6 years of education, compared to 13.4 years for rural parents.

OPEN IN VIEWER

Figure 2.

Average years of schooling among rural and urban migrants

The t-test results indicated a significant difference in the years of schooling among migrants (p < 0.0001), although parental education did not exhibit a statistically significant difference (p = 0.0515). Among rural migrants, Pearson correlations were notably stronger, particularly with fathers (r = 0.41 compared to r = 0.18). Educational elasticity was more pronounced among rural migrants, especially regarding fathers’ education (β = 0.5497 for rural versus β = 0.1428 for urban). This was followed by the combined influence of both parents (β = 0.4881 for rural versus β = 0.2699 for urban) and mothers (β = 0.3467 for rural versus β = 0.2689 for urban).

Educational Mobility

These patterns persist across generations, as demonstrated by the educational mobility matrices in Figure 3. The heat maps indicate a pronounced diagonal concentration in both groups, highlighting educational persistence. However, rural migrants exhibit greater “stickiness” at lower educational levels, with 44.2% of individuals whose parents are less educated remaining at that level, compared to only 33.3% for urban migrants. Conversely, urban migrants display greater potential for upward mobility; 66.7% of those with the lowest parental education achieve university degrees, in contrast to 55.8% of rural migrants. These matrices illustrate that an urban origin provides better opportunities for educational advancement, irrespective of parental background.

OPEN IN VIEWER

Figure 3.

Educational mobility matrices

The chi-square test results (Table 2) indicated that both downward and upward educational mobilities varied significantly by origin (p = 0.0001).

Appendix Table 3A illustrates several factors associated with higher odds of downward (as opposed to upward) educational mobility based on univariable models. Among rural migrants, factors such as speaking Kazakh (OR = 2.9, 95% CI: 1.4–5.8), low consumption capacity after relocation (OR = 7.6, 95% CI: 2.8–21.0), and living in peripheral districts (OR = 10.1, 95% CI: 3.0–34.4) significantly increased the risk of downward mobility. Conversely, being in the salariat class (OR = 0.0, 95% CI: 0.0–0.1) or having parents in the salariat class (OR = 0.3, 95% CI: 0.1–0.6) offered strong protection against such mobility. Similarly, among urban migrants, low consumption capacity (OR = 7.8, 95% CI: 2.8–21.7), Russian ethnicity (OR = 3.8, 95% CI: 1.1–12.8), and peripheral residence (OR = 3.6, 95% CI: 1.2–10.3) were linked to downward mobility. Current and parental salariat status continued to provide protective effects.

In the multivariable model (Table 4) comparing downward to upward educational mobility, lower consumption capacity after relocation was significantly associated with higher odds of downward mobility in both rural (OR = 3.8, 95% CI: 1.1–13.0) and urban groups (OR = 4.5, 95% CI: 1.0–19.8). The current occupational class of migrants strongly predicted mobility: those in the salariat class had near-zero odds of downward mobility in both rural (OR = 0.0, 95% CI: 0.0–0.1) and urban settings (OR = 0.0, 95% CI: 0.0–0.1). Similarly, having parents in the salariat class was protective among urban migrants (OR = 0.1, 95% CI: 0.0–0.4). Among rural migrants, living in peripheral districts was also associated with higher odds of downward mobility (OR = 5.3, 95% CI: 1.3–21.1), suggesting a continued influence of spatial disadvantage.

Class Mobility

Figure 4 illustrates the translation of educational advantages into occupational outcomes, depicted through the social class mobility matrices. Patterns of class mobility differed based on both origin and parental class. Rural migrants exhibited high rates of downward mobility across all class backgrounds, with significant rates among the working class (45.5%) and the salariat (44.9%). Conversely, urban migrants experienced more upward mobility or retained their advantages, particularly from the salariat (72.3%) and intermediate origins (47.8%). These results underscore the greater intergenerational stability and upward mobility observed among urban migrants.

OPEN IN VIEWER

Figure 4.

Social class mobility matrices

We employed a chi-square test to determine whether the distribution of class mobility varies between origins. The results indicate that both downward (p = 0.0009) and upward (p = 0.0003) class mobilities exhibit significantly different distributions between rural and urban populations.

Several factors demonstrated significant associations in univariable models comparing downward and upward mobility to stable mobility (Appendix Table 3A). Among urban migrants, factors associated with higher odds of downward mobility included being female (OR = 0.3, 95% CI: 0.1–0.7), lower post-relocation consumption capacity (OR = 0.1, 0.0–0.5), and lower parental education (OR = 0.2, 0.1–0.5). Conversely, Kazakh ethnicity (OR = 5.2, 1.8–15.1) and having parents in the salariat class (OR = 8.7, 1.7–43.9) were linked to upward mobility. Among rural migrants, the strongest predictor was migrant education: individuals with university degrees had lower odds of downward mobility (OR = 0.2, 0.1–0.7) and higher odds of upward mobility (OR = 15.4, 6.1–38.6). Additionally, residing in peripheral districts was associated with increased odds of downward mobility (OR = 5.1, 1.6–16.4).

In the multivariable model presented in Table 5, which was adjusted for origin and educational background, higher education among migrants remained a strong predictor of upward mobility (OR = 21.3, 95% CI: 2.4–191.2). Conversely, lower education levels were associated with an increased likelihood of downward mobility (OR = 0.2, 95% CI: 0.1–0.6). Additionally, parental education exhibited protective effects; specifically, having a parent with a university degree decreased the odds of downward mobility (OR = 3.2, 95% CI: 1.0–9.6). Origin from a rural area was not statistically significant following adjustment.

Discussion

This study examined intergenerational educational and occupational mobility among first-generation internal migrants in Almaty, comparing those from rural areas with their counterparts from other urban regions. The findings provide substantial evidence for all three hypotheses while revealing important nuances about how regional origin continues to shape mobility outcomes in post-Soviet Kazakhstan.

The first hypothesis proposed that intergenerational transmission of education and occupational class would be stronger among rural-origin migrants than among their urban counterparts. Multiple findings support this hypothesis. Educational elasticity was significantly higher for rural migrants, while Pearson correlations between parental and child education being stronger, particularly for fathers. Educational mobility matrices further revealed greater “stickiness” at lower education levels among rural migrants, with fewer advancing to university degrees when starting from low parental education. Class mobility patterns reinforced this trend: rural-origin individuals exhibited substantially higher rates of downward occupational mobility, even when they originated from salariat or intermediate backgrounds. For example, 44.9% of rural migrants from salariat families moved down the class ladder, compared to only 27.7% of urban migrants. Chi-square analyses confirmed that both upward and downward class mobility distributions differed significantly by origin. Together, these findings suggest that rural migrants experience stronger intergenerational reproduction of both education and class, reflecting deeper reliance on inherited resources and more constrained trajectories despite migration.

The particularly strong influence of fathers’ education on rural migrant’ outcomes (β = 0.5497, r = 0.41) warrants deeper interpretation within the contexts of restricted opportunity structures and intersecting disadvantages. In rural Kazakhstan, where formal institutional support is weak and market-based educational resources are scarce, family capital becomes a critical compensatory mechanism (Shnarbekova, 2018). Fathers, traditionally positioned as primary earners and authority figures in Kazakhstani households, often serve as the main channels of both material resources and strategic information about educational pathways. For rural families, a father with higher education may possess crucial knowledge about university application processes, urban labor markets, and the strategic value of different credentials information otherwise inaccessible in under-resourced rural schools. This heightened dependence on paternal education among rural migrants reflects what Bernardi and Ballarino (2016) term “compensation effects,” which is when institutional channels fail to provide equal opportunities, families with educated fathers can partially buffer their children against structural disadvantages through direct intervention such as financing preparatory courses, facilitating migration to cities for better schooling, or leveraging professional networks. Conversely, rural-origin children whose fathers lack higher education face compounded disadvantages: not only do they attend weaker schools, but they also lack the family-mediated capital necessary to navigate the stratified educational system effectively.

This pattern contrasts with urban migrants, where the father’s education shows weaker associations (β = 0.1428, r = 0.18). In urban contexts, a more developed institutional infrastructure may reduce dependence on individual family members’ knowledge. From an intersectional perspective, these patterns become even more nuanced when considering gender. Rural-origin women, who comprise a disproportionate share of rural migrant sample (OR = 1.8), may be particularly dependent on paternal educational capital, as traditional gender norms in rural areas often position fathers as gatekeepers of daughters’ educational and migration decisions. The intersection of rural origin, working-class background, and female gender may thus create a configuration of constraints in which paternal education becomes the primary pathway to mobility. Thus, the stronger father effect among rural migrants signals not simply cultural tradition, but rather a structural reality, meaning that in contexts of institutional scarcity and intersecting disadvantages, family becomes the default source of educational strategy and support.

The second hypothesis proposed that the effects of rural versus urban origin would essentially vanish when adjusting for educational attainment and parental occupational class. The multivariable regression models partially confirmed this hypothesis. The results indicated that origin had limited independent predictive power in multivariate analyses; however, residual effects persisted through indirect mechanisms, such as residential segregation and differential returns eliminated by formal education alone, pointing to additional pathways through which origin-based disadvantages operate.

The third hypothesis proposed that education reduces the risk of downward mobility but is less protective for rural migrants. The data fully supported this hypothesis. Although university education was associated with significantly lower odds of downward mobility for both groups, rural migrants remained disproportionately vulnerable. The multivariable model showed that having a university degree strongly increased the likelihood of upward mobility. However, despite similar qualifications, rural migrants were more likely to fall into lower-status occupations. Class mobility matrices revealed that even rural individuals from salariat families experienced high rates of downward mobility (44.9%), indicating that their educational credentials translated into weaker labor market outcomes. These findings confirm that education does offer protection. However, its equalizing effect is partial: rural-origin individuals face diminished returns to the same credentials, reflecting structural barriers such as lower institutional prestige, weaker networks, and residential segregation. As a result, rural migrants continue to face cumulative disadvantage even after achieving educational mobility.

The results indicate that rural migrants achieve fewer years of schooling compared to urban migrants and exhibit greater dependence on their parental background. This disparity may occur due to rural-urban differences, which arise from under-resourced schools and limited preparatory support (Nurbayev, 2021; Zhanbyrbayeva et al., 2023). Similar findings have been documented in other transitional societies, such as China, where regional disparities in educational quality undermine intergenerational mobility and equal opportunities (Zhang, 2017).

Limited access to high-quality education, both in areas of origin and in Almaty, affects the occupational mobility or rural migrants. The results show that they are less likely to convert their academic credentials into high occupational classes. Even when controlling for educational level, the returns on education vary by origin: urban migrants are more likely to enter the salariat class, whereas rural migrants are often restricted to working-class jobs. This pattern of differential returns is also observed in international mobility research (Torche, 2019), where educational attainment yields unequal occupational outcomes based on an individual’s social origin and institutional background. This, in turn, challenges the notion of education as the great equalizer, despite increased access to higher education in Kazakhstan. The underlying cause may be new forms of educational stratification (Roberts et al., 2009). Rural migrants likely attend lower-tier universities, where their degrees hold limited value in the labor market. Furthermore, research from Central Asian countries has revealed that graduates from a select few top universities frequently secure elite positions in the labor market, which are typically accessible to students with greater family resources (Roberts et al., 2009). Consequently, even after earning a university degree and relocating to Almaty, structural disadvantages continue to influence life opportunities. Although this study did not directly measure institutional prestige, the unequal returns to education observed here may partly reflect stratification within Kazakhstan’s higher education system. Post-independence expansion created a hierarchy in which a few nationally ranked universities with competitive admission and strong employer recognition coexist alongside numerous less selective regional and private institutions (Shnarbekova, 2018, 2021). Roberts et al. (2009) found that the public-private university distinction per se mattered less for occupational outcomes than complementary investments such as English-language training and private coaching. Given that rural-origin migrants in this study had lower parental education and fewer economic resources, they may have been more likely to attend less selective institutions, which could partly account for their lower returns to formally equivalent credentials. Future research capturing institutional tier, for example, through admission selectivity, ranking status, or employer recognition, would help disentangle this mechanism from broader origin effects.

These findings align with the theories of cumulative (dis)advantage and the direct effect of social origin. Rural migrants with lower parental education, limited economic resources, and restricted access to basic education represent these early disadvantages. This scenario reflects a failure of educational mobility to rectify origin-based disadvantages, as individuals struggle to overcome their initial conditions. Furthermore, rural migrants face occupational ceiling effects, where despite similar educational attainments, they are less likely to advance into salariat or intermediate classes. This indicates that disadvantages accumulate across life stages, limiting the returns on even successful educational investments.

These educational and occupational dynamics challenge the notion of Almaty as a region facilitating upward mobility. Although Almaty provides a more diverse range of labor market positions and greater access to institutions, the findings reveal that these opportunities are not distributed equally. Rural migrants, who predominantly settle in the city’s peripheral areas, face barriers in the labor market. Consequently, the city’s escalator system may selectively filter rural migrants into lower occupational classes, while more effectively elevating urban migrants. This unequal distribution indicates that migration alone cannot eliminate accumulated regional disadvantages.

Conclusion

This study demonstrates that while internal migration and educational attainment provide avenues for upward mobility, social origins continue to significantly influence outcomes. Migration to a major city such as Almaty does not ensure equal opportunities; rather, it reveals how background conditions shape mobility gains. Rural-origin migrants face stronger intergenerational transmissions of educational and occupational status, achieve lower returns to their educational credentials, and experience higher rates of downward mobility compared to their urban-origin counterparts. Education, while offering some protection against downward mobility, functions as a conditional rather than universal equalizer—the capacity to reduce origin-based disparities is constrained by institutional context, regional opportunity structures, and family background.

This study has several limitations. First, it lacks a non-migrant comparison group, which complicates the assessment of how internal migration alters mobility relative to individuals who remain in place. Second, its cross-sectional design limits the ability to make causal claims regarding long-term mobility trajectories. Third, while capturing education levels, the study did not measure the prestige or quality of educational institutions, potentially obscuring variations in the returns to credentials. Finally, factors such as social capital, informal networks, and employer bias were not directly assessed. Future studies should employ longitudinal designs to better trace life-course mobility and evaluate cumulative effects over time. Comparative studies involving non-migrants and second-generation migrants would further deepen understanding of mobility mechanisms and the long-term consequences of internal migration. Research examining institutional prestige and its interaction with regional origin would also illuminate additional mechanisms of stratification.

Addressing these inequalities requires moving beyond access-focused policies toward structural interventions that directly target the mechanisms perpetuating unequal returns to education. The findings suggest several evidence-based policy directions.

The quality gap in educational preparation demands targeted equalization programs that extend beyond infrastructure investment. Japan’s mandatory teacher rotation system (“jinji idou”) offers a model, where public school teachers are systematically transferred every three to 6 years, ensuring that each teacher works at a small rural school at least once during their career. Research demonstrates that this centralized system produces a more equal distribution of teacher quality compared to decentralized labor markets (Seebruck, 2021). For Kazakhstan, implementing comparable rotation requirements with substantial salary premiums (30–50% above urban rates) could help equalize educational quality across regions, while subsidized digital learning platforms would provide rural students access to high-quality Unified National Testing (UNT) preparation currently available only in urban centers.

The transition from rural education to urban higher education represents a critical juncture where accumulated disadvantages intensify. Germany’s BAfoG system exemplifies comprehensive support, providing up to 992 euros monthly for students from low-income families, structured as 50% grant and 50% interest-free loan (BMBF 2024). Australia’s Tertiary Access Payment specifically targets students from regional and remote areas who must relocate more that 90 minutes from their family home (Australian Government, 2024). The United Kingdom’s Widening Participation framework employs contextual admissions with reduced grade requirements for first-generation students from underrepresented areas (Office for Students, 2019). For rural-origin students in Kazakhstan, such comprehensive transitional support should include mandatory pre-enrollment summer programs addressing gaps in academic preparation, extended peer and faculty mentorship during the first 2 years, and enhanced financial support covering living costs with higher stipends for students choosing housing in non-peripheral districts.

Labor market discrimination against rural-origin graduates requires direct intervention in hiring practices. Field experiments demonstrate that resumes with names associated with minority groups receive substantially fewer callbacks than identical resumes with majority-group names (Bertrand & Mullainathan, 2004). Anonymous hiring practices, including name-blind resume screening and standardized skill-based assessments, have been successfully implemented to reduce such bias. For Kazakhstan, mandatory anonymous resume screening in public sector hiring (removing names, photographs, and place of origin indicators) combined with incentivized adoption in large private firms would address employer bias. Structured internship programs specifically targeting rural-origin university students, with government subsidies to participating employers, would further help overcome network disadvantages.

The strong association between peripheral residence and mobility constraints documented in this study (OR = 11.3 for rural migrants residing in peripheral districts) demands explicit spatial integration policies. France’s Solidarity and Urban Renewals (SRU) law provides an instructive model, requiring most urban municipalities to ensure at least 20–25% of their housing stock is social housing (Maaoui, 2023). Research demonstrates that this law successfully stimulated public housing construction in municipalities with low proportions of social housing and reduced spatial segregation (Chapelle et al., 2022). For Almaty and other Kazakhstani cities, mixed-income housing requirements mandating that 20–30% of units in new residential developments be allocated at subsidized rates for recent migrants would prevent the concentration of disadvantaged populations that perpetuates mobility constraints.

Finally, sustainable progress requires institutional transparency and systematic monitoring. Universities should be required to publicly report admission rates, graduation rates, and employment outcomes disaggregated by migration status, origin, and gender, with consequence for institutions showing persistent disparities without remedial action. Annual labor market survey tracking wages, employment rates, and occupational class by field of study, origin, and gender would provide the data necessary to evaluate policy effectiveness and guide adjustments. An independent monitoring body with authority to recommend policy changes would ensure accountability.

These interventions recognize that equality of access does not automatically produce equality of outcomes. Access alone cannot overcome the structural filters that shape life chances such as educational quality differentials, employer perceptions, residential segregation, network effects, and intersecting disadvantages based on gender and class. In Kazakhstan and similar transitional societies, reducing inequality requires not only expanding educational access or supporting mobility but also addressing the deeper institutional and structural barriers that perpetuate disadvantage.

Uneven school systems, especially the rural-urban divide in educational quality, perpetuate early disadvantages that accumulate over a lifetime. Even when individuals from rural backgrounds attain higher education, hierarchically structured labor markets often diminish the value of their credentials. This segmentation within urban labor markets—characterized by informality, networks, and status considerations—tends to favor those with stronger social capital, typically individuals originating from urban areas or those with better connections. These processes contribute to the covert reproduction of class advantage, even in systems that appear meritocratic.

Ultimately, addressing inequality in transitional contexts requires a shift in focus. It is not enough to merely expand access; attention must be given to dismantling the systematic filters that dictate who benefit from opportunities and who remains excluded, even after taking all the prescribed actions.

Supplemental Material