Since economic reform in 1978, migration and the expansion of higher education have become important factors in China's modernization process and are closely linked with social mobility (Cebolla-Boado & Soysal, 2018): Rapid economic growth has stimulated massive migration both at home and abroad for educational and/or career opportunities. When market forces continue to permeate every sector, including education and the labor market, the special institutional arrangement (i.e. hukou system) and the structure of the labor market are still organizing the social structure of the entire society (Bian & Logan, 1996; Xiao & Bian, 2018).This background provides a context for studying the relationship among socioeconomic status attainment, higher education and migration, namely, how migration produces and maintains social inequality under the joint action of centralist state-control and market forces.

At present, the research on “how mobility affects college graduates’ performance in China” mainly has the following shortcomings. First, in 2019, China’s Gini Coefficient was reported at 0.465 NA. The main investigation of previous research on individual social and economic status is ascribed factors (such as family background, residence, gender) (He & Wu, 2018), achieved status, such as level of education (Xiao & Bian, 2018), social capital (Bian, 1997), or the structure of the labor market (Li et al., 2015), while the geographical mobility and its interaction with other factors impact on individual socioeconomic status is ignored. Second, in terms of higher education, similar to many western developed countries in the 1950s, China's higher education began to expand in the late 1990s. Since 1999, the Chinese government has adopted the higher education expansion policy, resulting in the rapid growth of higher education opportunities. A striking example is that in 1998, the gross enrollment rate of higher education was 5.86% and by 2018, the ratio had reached 50.60 percent (World Bank, 2020). However, research on higher education focuses on the influence of family background and factors in the education process on the access to educational opportunities and academic performance (Liu,2015; Tam & Jiang, 2015), while there is a lack of research on the relationship between higher education and labor market. Third, in terms of migration, similar to many other developing countries, China has been experiencing the massive volume of internal migration. In 2017, 244 million people was estimated to move across regions within the country (Report on China’s Migrant Population Development, 2017), which is close to the total global international migration (United Nations, 2017). Because the rural-to-urban is quantitatively so important, many studies have been conducted to explore the labor market performance of low-skilled migrants in China (Fan, 2002; Wu & Treiman, 2004; Zhang & Wu, 2017), whereas research on higher-education and highly skilled migration within China is scarce.

Research question

This dissertation is composed of three empirical studies focusing on the effects of migration on the socioeconomic status of college graduates and whether such effects are limited by other structural factors. Specifically, Chapter 4 studies the interaction of migration and gender on college graduates’ starting salary and work organization entry. Chapter 5 explores the influence of migration and family background on college graduates' employment outcomes; examines the role of locality (cities) in differentiating socio-economic outcomes of migrants. The last empirical chapter, Chapter 6, examined whether Beijing, Shanghai, Guangzhou and Shenzhen, the first-tier cities, have emerged as an “upward social class escalator region” (Fielding, 1992) for the young adults in the Chinese graduate labor market. The research questions were as follows:

1.Is there an economic premium attached to graduate migration? If yes, do all young people with different characteristics benefit equally from the migration premium?

2.Is there a double negative effect between gender and migrant status on college graduates’ initial salaries and entry into the state sector for employment?

3.Is there migrant selectivity among graduate return migrants in terms of human capital characteristics? What is the impact of family background on graduate return migrants’ labor market outcomes?

4.Is there a positive association between moving to first-tier cities (Beijing, Shanghai, Guangzhou, and Shenzhen) and college graduates’ monthly starting salaries in China’s labor market? Due to the stringent local hukou barrier in Beijing, Shanghai and Guangzhou and Shenzhen, do graduate migrants have equal access to employment opportunities in government organizations and public enterprises compared to their local peers?

Data and methods

Data

Findings of the three empirical studies in this dissertation are based on second-hand data. The data came from the “China College Student Survey” (CCSS), which was conducted by China Data Center at Tsinghua University in 2010, 2013 and 2015. The project used stratified probability-proportional-to-size random sampling and 60 universities from 23 provinces participated in the survey. It is a nationally representative survey. This survey aimed to understand the academic performance of Chinese college students during school, as well as their performance in the labor market and personal career choices. The survey also included information on the students’ socioeconomic status, school education during the middle school, and college entrance examinations. Since undergraduate study in China is usually a four-year program, students in the last semester of undergraduate study (four-year program) were defined as a sample population. The selected respondents need to fill in their placement after graduation, and students who chose to seeking a job after graduation need to give some information about employment (such as number of offers, place of employment, income, etc.).

Further, the data used in this dissertation was limited to a sample with the following characteristics: Students who had no experience of migration before college and chose to find a job after graduation. In addition, only those who had received (at least) one offer and provided income information in the survey were included. The final sample consisted of 5,906 eligible individual students.

Data analyses

Considering the sample selection problems caused by these abovementioned conditions, the dissertation used Heckman’s (1979) two-step sample selection models to test the effect of sample selection bias on the results. Specifically, in the first step, a probit model was fitted to estimate the selection into our analytical sample, based on which the inverse Mill’s ratio, λ, was calculated (Flippen, 2013). In the second step, the λ parameter was included in the model predicting starting salary. In migration studies, researchers also pointed out the migrant self-selection issues (Borjas, 1987; Chiswick, 1999). Chapter 6 also used the two-step selection model to identify the return migration selectivity.

Besides, this dissertation adopted propensity score matching and weighting methods. The matching method was used to compare the aggregate impact of education migration and work migration on the salary outcome in the whole sample as well as in disaggregated samples according to hukou origin (rural or urban); while the latter was to reduce selectivity bias in regression models, following the argument from Ridgeway, Kovalchik, Griffin and Kabeto (2015), I include the product of a propensity score weight (correcting for graduate migration selection bias) and the sampling weight as the final weight in estimations of regression models.

Overview of the findings

Chapter 4 examines how internal graduate migration interacts with gender and play a key role in producing inequality among recent college graduates in China. Results show that, on one hand, female graduate migrants had less chance to enter governmental organizations which affords institutional protection against gender discrimination. On the other hand, the effect of geographical mobility (migration) varied by work organizations and female graduate migrants who ended up in the non-state sector were more likely to experience an income penalty in earnings attainment. The results suggest that although female graduate migrants are a highly selective group in terms of human capital characteristics, they are disadvantaged twice in the labor market because of the existing barriers based on gender and hukou locality.

In Chapter 5, the impact of graduate return migration on initial salaries of college graduates in Chinese graduate labor market is investigated. Results reveal that graduates moving back to their pre-college hukou located cities had more opportunities to get a placement in government organizations and public enterprises, relative to onward migrants, and the advantage was enhanced by the introduction of family political capital. Besides, despite of the effect of migrant selectivity indicated that they were not much different than the onward migrants with respect to the characteristics that determine starting salaries, graduate return migrants had slightly higher earnings than onward migrants. The findings suggest that family background plays an important role in shaping differences in labor market outcomes among graduate migrating population and graduate return migration seems to be a strategy for upward social mobility taken by college graduates from privileged families.

Chapter 6, the last empirical chapter of the dissertation, examined whether Beijing, Shanghai, Guangzhou and Shenzhen have emerged as an “upward social class escalator region” (Fielding, 1992) for young people in China. After accounting for observed demographic and human capital characteristics and migrant selectivity, migrating into Beijing, Shanghai, Guangzhou and Shenzhen has been found to be positively associated with earnings attainment, and the economic benefit from relocation was greater than that experienced by migrants elsewhere in the system. However, in-migrants to Beijing, Shanghai, Guangzhou and Shenzhen have fewer opportunities to work in the state sector (government organizations in particular). These results suggest that migration to Beijing, Shanghai, Guangzhou and Shenzhen can only bring short-term economic benefits, but it cannot guarantee the upward social mobility of migrants. It remains to be further investigated as to whether such a region has the potential to become an escalator region in China.

Implication of the findings

Findings above demonstrate the impact of internal migration on the socioeconomic status of graduates in China’s labor market. Based on these findings, there are some policy recommendations and theoretical directions that can be pointed out.

Theoretical implications

Overall, through nuanced investigation of the influence of migration on labor market outcomes, my dissertation makes an important contribution to the field of social mobility literature on labor market performance of highly-educated migrants.

1. The impact of internal migration on the socioeconomic status of college graduates

In the field of higher education, the relationship between education and employment is a topic that scholars and policy makers have long been concerned about (Schomburg & Teichler, 2007). Research in the field of higher education is focused on the impact of education-related mechanisms (national education system, school type or major, etc.) on employment outcomes. In this study, the influence of higher education related factors has been verified（academic performance at college such as undergraduate GPA, awards, CCP membership, English language certification; fields of study and college type (whether “Project 211” or not) .

More importantly, findings indicate that graduate migration has a non-negligible impact on the employment results of college graduates. The results of this dissertation also show that in China, as economic reforms continue to advance, students have more freedom to make use of geographical mobility to find the most suitable job for themselves, so as to maximize the economic return brought by work.

The labor market performance of migrants also reflects the socio-economic integration of migrants. The findings challenge the idea of self-selection of migrants. Results from the empirical chapters show that in general, there is a positive association between geographical mobility and the initial salaries of college graduates, even after controlling for their migration-related characteristics. Migrant experience itself, rather than migrant selectivity, helps graduate migrants to achieve higher earnings.

My dissertation suggests that although there is a positive association between migration and salary outcome, migration has an inhibiting effect on entering the primary sector in the labor market. The existing institutional and structural factors in the society have an important impact on the consequences of migration. Different from the segmented mechanism in developed countries, China’s labor market is divided by state-owned and non-state-owned sectors and the hukou system is the main channel for sending workers into different sectors. Hukou hinders the free movement of college graduates, making it impossible for migrants to achieve upward mobility through migration.

Besides, the results of this dissertation prove that cities play an important role in creating unequal consequences of migration. High-skilled migrants have often been associated with global elite, as free-floating mobiles that are disembodied from localities and moving outside of the constraints of nation states (Hannerz 1996, 129; Sklair 2005). In China, these highly educated migrants are not “free-floating mobiles”, instead, their mobility and its consequences are constrained by institutions and structures. Results show that, despite of a substantial income premium from migration that in-migrants to Chinese megacities (Beijing, Shanghai, Guangzhou, and Shenzhen) experience, they face more restrictions from the hukou system (comparing migrants elsewhere in the system) when seeking employment in government organizations and public enterprises.

Last, in China, citizenship has been found to be stratified in terms of hukou locality. In the Chinese context, citizenship has been interlocked with the hukou (household registration) system for more than 50 years. With the reform of the household registration system，the hukou location，rather than the hukou classification, has become more important in determining access to resources and defining one’s life chances. The larger the city is, the more valuable is its hukou because there are more government-provided benefits. Beijing, Shanghai, Guangzhou and Shenzhen, where the barriers or social boundaries associated with hukou are stronger than any other cities (Wang et al., 2017; Zhang & Tao, 2012), as a result, migrants encounter more difficulties in labor market if they want to settle down in these cities.

2.Maintenance of inequality: the interaction between migration and other social and demographic characteristics

The findings of this dissertation indicate that although graduate migration is generally considered to be “a process wherein higher-education students …for the purpose of career advancement or upward social mobility” (Li, et al. 2020, p.4); however, mobility did not play a role in reducing existing inequalities. In the Chinese context, The reason why migration is associated with inequality lies in its mechanism: Exclusion (Immigration Policy and Citizenship at the societal level) and opportunity hoarding (Opportunity structures in employment markets at the societal level). The hukou system in China, is a typical example of exclusion based on legal boundaries enforced by the state. This dissertation also wants to emphasize the socio-economic differentiation brought about by the classification of hukou system (especially local-nonlocal). Just as immigration policy measures and citizenship which is closely related to legal affiliation in international migration, the nature of legal exclusion provided by hukou restricts their access to the same job opportunities as local residents, although it does not limit the migration behavior of graduates. The linked mechanism is opportunity structure in the labor market. For graduate migrants, the hukou system does shape the opportunity structure in the labor market, that is, compared with local people, migrants have fewer job opportunities. But more importantly, it is important to note that the hukou system works in conjunction with the segmented labor market: what migrants lack is not jobs that are generally available, but jobs in the primary sector of the segmented labor market. Therefore, migration produced inequality through exclusion (in the sense of legal aspect) and opportunity structure.

Additionally, this dissertation suggests that, migrant status interacts with pre-existing inequalities such as hukou status(rural/urban) and gender and shapes outcomes among college graduates in China’s urban labor market. Rather than being natural, categories, such as gender and race, are socially constructed, and they not only influence individual identities but also provide principles of organization in the social system (Browne & Misra, 2003). Thus, these results imply that migration may serve to entrench pre-existing socioeconomic inequalities between rural and urban hukou holders, as well as between men and women.

3. Family background, education and migration: The reproduction of social inequality

In the field of stratification research and higher education, one of the focal points of debate is the extent to which higher education reduces the influence of family background on an individual's economic success (Bloome, Dyer, & Zhou, 2018; Hout, 1988; Torche, 2011; Witteveen & Attewell, 2017, 2020). Findings of the study support the reproduction hypothesis, that is, family background still has a very important influence on the employment outcomes of college graduates: Family income is significantly positively correlated with graduate starting salaries.

In addition, although the political capital of parents has no significant impact on the salary of a graduate, it has a significant impact on whether college migrants can enter into the employment of primary sectors. This shows that power maintenance still plays an important role in the graduate labor market. Political elites are more likely to pass on their dominant position to their children, and the formation of elites is exclusive.

Another important finding of this dissertation is that the influence of family background on one's economic achievement can also be realized by influencing one's migration behavior. The return migration behavior in particular, is a resource and power of dominant families: graduate return migrants are more likely to be singletons from one-child urban families with higher family income, cadre parents, and more familial political capital. Furthermore, those with more political capital are more likely to enter government organizations and public enterprises. By influencing their children's migration choice, the children from the advantaged families obtain higher socioeconomic status by means of mobility, thus realizing the intergenerational transmission of class. Therefore, geographical mobility becomes a tool for the reproduction of social class.

Practical implications

In terms of the labor market, the findings of this dissertation suggest that the government needs to establish policy evaluation and supervision mechanisms for gender equality. At the same time, laws and regulations related to the labor market should be established to enhance the awareness of labor equality and the ability to protect rights, so that female workers can learn to use legal means to fight for and protect their rights. Moreover, such local protectionism should be corrected in policy design, given the unequal employment opportunities for locals and outsiders brought about by the hukou system. In the field of higher education, students should have more freedom of choice in their educational mobility. For rural students, attention should be paid to the improvement of rural educational resources, including teachers and schools. In the college application stage, teachers or organizations should fully explain the situation of each higher education institution (including education and employment), so that students and parents can have a deeper understanding of the school. What’s more, the establishment of various types of colleges and universities, such as vocational skills schools, and the further improvement of the higher education system will enable students to have more choices when applying for colleges and universities. In terms of the household registration system, the welfare and rights attached to the household registration system should be removed to break the institutional bottleneck of social mobility of the floating population. Given the important role of first-tier cities in China's economic system, how to solve the problem of floating population in first-tier cities has become a key issue for policy makers to solve in the future.

Limitations and directions for future research

Finally, several important limitations need to be considered. First, the dissertation was limited by the use of a cross-sectional design conducted, and it was not able to investigate the long-term impact of migration on college graduates’ labor market outcomes. Future research should attempt to collect longitudinal data that trace a representative sample of higher education leavers after graduation in order to have a better understanding of the long-term consequences of migration. Second, although this dissertation tried best to include as many variables as possible to correct for migrant self-selection bias, it should be acknowledged that there is still a possibility that unobserved traits, such as personality, risk-taking, and confidence, are also related to the labor market outcomes of graduates. It is suggested that the association of these factors is investigated in future studies. Third, this dissertation was exclusively focused on the effect of graduate migration on the labor market outcomes among bachelor’s degree holders. However, migration is associated with education, and individuals with the different educational background may benefit differently from geographic mobility. Future work from China to include individuals at different education levels (such as B.A., M.A., and PhD.) would be of great help in knowing the same issue in the Chinese context. Fourth, this dissertation merely examined the economic outcomes of migration. While both earnings attainment and work organization entry are essential indicators of the social and economic status in China (Li et al., 2015; Xiao & Bian, 2018), the consequence of migration is related to other aspects of life, including housing ownership, marriage, and family support (Cui et al., 2016; Qian & Qian, 2017). One possible area of future research would be to conduct follow-up surveys of the college-educated cohorts of the current CCSS sample. This could help researchers gain a more complete picture of the graduate migration process and its importance for determining individuals’ life opportunities.