Luke Dorman
Political Social Sciences
Michaelmas Term, 2024
Cambridge Journal of Political Affairs, 5(2), pp. 173-186

https://doi.org/10.5281/zenodo.14290479

1. Introduction

The man in front of us used to live in Wuhan but has chosen to return to his rural hometown of Jiangxia, a 30-minute drive from the city. He smiles warmly as he pours us tea and then tells us how he saw his hometown slowly empty of young people until there was only one man left. He now runs a successful retreat centre where schoolchildren can stay for a few days to do activities and agricultural work. He talks passionately about the holistic education the countryside provides and his hope that some of the schoolchildren who stay may choose to return to the countryside. If we are to believe the website ‘China’s Rural Revitalisation’, there are many more like him. We hear stories of Ma Qinyan, the thirty-five-year-old architectural designer-turned chilli sauce seller (China Internet Information Centre, 2024a), and Gan Wenyong, who returned home to build a library after graduating (China Internet Information Centre, 2024b). These are individuals, and their reasons for migration are unique and multitudinous. However, they also form part of both a broader trend and a key government initiative. China has been witness to unprecedented urbanisation in the last 50 years. Yet, at the same time, there has been a notable backflow of migration out of cities (Hao and Tang, 2019), particularly after the COVID-19 pandemic (Ma et al., 2024). At the same time, in the government’s ‘rural revitalisation’ strategy (proposed in 2017 and progressively implemented since 2021), the Chinese Communist Party (CCP) sets its sights on turning China into an agricultural superpower and holistically developing its rural areas (Hillman, 2023). Rural revitalisation, of which urban-rural migration is an explicit part, is sold as a unique vision of development in a world where extreme urbanisation and rural deprivation are accepted prices to be paid for growth.

This paper, then, seeks to explain why individuals in Chinese cities move to rural areas and whether the reasons that they move are consistent with and conducive to rural revitalisation in China. There is currently no literature directly evaluating the feasibility of the Chinese government’s recent attempts to promote urban-rural migration, and the current literature on return migration is often not based in China or relies upon in-depth individual interviews. An analysis of the feasibility of rural revitalisation requires a study in a specifically Chinese context, with a broader scope than individual interviews and a focus on the demographics the Chinese government wishes to migrate.

To this end, I draw upon the existing return migration literature to propose a wide range of hypotheses illustrating potential motives for migrating. I first draw on theories from neo-classical economics that see migration as an attempt to maximise income. From these theories I delineate two hypotheses, that people migrate due to (1) short-term economic incentives and (2) long-term economic incentives. I secondly draw on theories predicting that individuals will migrate to maximise social or cultural ties, such as friendships, family relationships, or affinity with local customs. Again, I delineate two hypotheses from these theories: that people migrate due to (3) pressing short-term social or cultural incentives or (4) long-term social or cultural incentives. Finally, I directly assess the efficacy of government media campaigns by proposing that individuals migrate due to (5) altruistic desires to improve the lives of rural residents and (6) media exposure and inspiration. These come to six hypotheses in total. My work is unique in clearly delineating these incentive theories’ short- and long-term dimensions.

I test these hypotheses with a survey of 110 students and highly educated individuals in work, as these are the demographics particularly encouraged by the government to relocate. I allowed participants to self-report their likelihood of relocating and any reasons for this hypothetical move, treating the latter as my independent variable. Using self-reported motivation is unprecedented in the literature, and it usually involves regressing demographic variables on an individual’s migration status. This allows me to more clearly determine migration decision mechanisms and focus on individual motives, which will be more susceptible to government influence than structural factors.

My analysis indicates that immediate economic threats and personal ties are the defining reasons why people will be prepared to relocate as opposed to less pressing cultural or economic reasons, which are reported as hypothetical motives for relocation even among those who report no possibility of migrating. Similarly, motivating ideals and media influence are shown to have no effect. These results pose a challenge to the Chinese government, which has, up until now, relied upon subtle rhetoric and economic pull factors, such as financial support programs, to motivate migration. They tell a story about migration motivated more by need than desire for those who are highly educated and employed. The story has its exceptions, as proven by Gan Wenyong. However, this survey marks him as an exception to the rule.

Numerous policy lessons can be drawn from this. On the one hand, we are warned not to overestimate the potential of government incentives to encourage people to uproot and migrate. On the other hand, we see that a specific emphasis on the employment opportunities offered in rural areas and the potential for fulfilling part-time work in conjunction with family care are effective ways that the Chinese government should encourage rural relocation.

2. Drivers Of Migration

2.1 China’s Historical Migration Policy

It is important to establish the historical context of the Chinese government’s actions to evaluate current government incentives for relocation. Migration policy in China is a far cry from migration policy in the West. Whilst European countries’ migration policies focus on regulating the flow of international migrants and refugees into the country, international immigration is the least of the Chinese government’s concerns. China’s 2020 census identified about one million international immigrants to China, 10.5 million emigrants from China, and 376 million domestic migrants (those living somewhere other than their household registration area) (Haugen and Speelman, 2022). If we understand the relative scales of these types of migration, we can begin to understand why domestic return migration is a salient topic to study and a policy priority for the Chinese government. For these reasons, this section will be devoted solely to explaining China’s historical domestic migration policy.

China’s domestic migration policy has been defined for over 70 years by the Hukou household registration system (Gustafsson et al., 2020). This was introduced in the 1950s and provided each citizen with a registration identifying their place of birth. Individuals were not permitted to live outside of their Hukou registration without approval to change their registration from the government (Gustafsson et al., 2020). This approval was rarely given, so the Hukou system effectively halted freedom of residence within China, providing a valuable means of public administration and political control for the CCP (Gustafsson et al., 2020). After the death of Mao Zedong in 1976, these regulations gradually loosened, allowing individuals to migrate in the short term for work, though still barring permanent relocation. From the 1990s onwards, this loosening of restrictions encouraged urbanisation, as the Chinese government prioritised economic growth (Gilley, 2019). Whilst migrants to cities have previously faced discriminatory policies, such as unequal access to housing, education, and other public services under the Hukou system (Xu et al., 2011), the National Plan for New Urbanization issued in 2014 (CPC Central Committee and State Council, 2014), grants full citizen status to migrants in cities and towns, effectively removing systemic barriers to integration (Chen and Liu, 2016). Though now ten years old, this legislation is often missed by much of the English-speaking literature on migration in China, which still significantly overestimates the role of the Hukou registration system in current Chinese migration politics. Though migrants still face discrimination from the urban population (Wong et al., 2007), the policy barriers to their urban relocation continue to be lowered, with the government currently focussing on skills and employment support for rural migrants and aiming to integrate 100 million more rural residents into cities by 2025 as part of the fourteenth five-year plan (The People’s Government of Fujian Province, 2021).

The flip side of this focus on effective and integrated urbanisation is an increasing number of policies incentivising migration back from cities to the countryside. These are intricately linked with China’s rural revitalisation strategy proposed at the Nineteenth National Congress in October 2017. The strategy is wide-ranging, aiming to modernise agriculture, develop rural industries and e-commerce, promote rural tourism, and ensure high-quality rural governance (Chen et al., 2023); however, a central requirement of this strategy is the movement of skilled labour from the cities to the countryside. The government has been implementing schemes to incentivise this since 2015 and oversaw 341 pilot projects between 2015 and 2021 (NDRC Employment, 2021). It should be noted that the nature of these pilot projects means that the precise implementation of rural revitalisation policies across China has been irregular between regions. Rural revitalisation is by no means a uniform national policy.

Measures implemented include economic incentives, such as designated funds and infrastructure for new businesses and relocation subsidies, general infrastructure and social service investment in rural areas, and promoting rural cultural events and community centres (NDRC Employment, 2021). In addition to this, outside rural areas, national media campaigns have highlighted the mental and physical health benefits of rural life. An indication of just how pervasive this messaging is can be found in the huge success of the 2022 TV series The Daughter of the Mountains, which documented the life of a communist party official who moved to rural Leye County to engage in poverty alleviation (Beijing Youth Daily, 2022).

So, why this two-sided approach? If the Chinese government has been engaged in a long-term concerted effort to drive urbanisation, why is there a parallel emphasis on rural revitalisation and urban-rural migration? This can be seen as a response to the limitations of China’s urbanisation strategy, outlined below.

Firstly, an excessive emphasis on urbanisation neglects the importance of China’s agricultural industries. As Lester Brown (1995) explained in his seminal book, Who Will Feed China?, it is a matter of national and international trade security that China can feed its massive population. If China cannot guarantee domestic food security, Brown argues that Chinese demand for food imports could lead to skyrocketing global food prices. For the Chinese government, in the wake of US-China tensions, the ability to move away from dependence on food imports is invaluable (Hillman, 2023). It is for these reasons that China cannot ever fully transition to a service economy, and this is recognised by the government. Innovations in agricultural practices are essential to develop one of China’s most essential industries.

Secondly, solely emphasising urbanisation misses the work still to be done in rural poverty alleviation. Though Xi Jinping declared a ‘complete victory’ in the struggle against extreme poverty in 2021, the livelihoods of many farmers remain modest (Hillman, 2023). The egress of rural youth to cities has left many rural elderly without the family care provision upon which they would have previously relied (He and Ye, 2013), placing rural residents in precarious situations and stunting local economies. The combination of improved service delivery and the return of more extensive family support networks that rural revitalisation aims to provide will be essential in reducing rural poverty.

Finally, China’s urban graduates face a highly competitive job market, with many opting to return home of their own volition (Eckhardt, 2023). The unemployment rate among 16 to 24-year-olds recently hit a record high of 21.3 per cent due to a lack of high-paying, high-skilled jobs and an overabundance of highly qualified youth (Eckhardt, 2023). Encouraging urban-rural migration and new rural enterprises can be seen as an exercise in job creation for Chinese graduates.

These factors explain why the Chinese government is now encouraging return migration and allow us to understand what the government is looking for in these migrants. Urban-rural migrants are expected to be highly skilled, able to integrate well and engage in entrepreneurial and innovation-centred activity upon relocation, ideally with a social goal or benefit. For these reasons, I restrict my sample to university students and highly educated individuals in work.

2.2 Theories Of Return Migration

So, why do people migrate? The main theoretical approaches to these questions can be understood broadly as (sometimes implicitly) expecting migrants to maximise certain things. I divide these approaches into theories maximising (a) income or human capital and (b) social or cultural ties.

Firstly, and most straightforwardly and explicitly, theories informed by neo-classical economics expect migration to occur due to a cost-benefit evaluation maximising income (Sjaastad, 1962; Todaro, 1976). The new economics of labour migration can be seen as a variation, emphasising migration as a response to market failures at the immigrant’s place of origin (Stark, 1991). However, both theories fundamentally see migration as an individual or household’s attempt to maximise their income. These theories predict that urban-rural migrants will be more likely to be unemployed and have lower incomes (Stark, 1991).

A variation of neo-classical theories is human capital theories, which view migrants as maximising the value of human capital (Constant and Massey, 2003; Massey and Akresh, 2006). Under this framework, settlement decisions are viewed in light of an individual’s current and expected human capital so, for example, it is predicted that those who are more educated and speak the local language will find it easier to grow their human capital and so are more likely to settle in a particular area. Similarly, a more secure or stable occupation with a higher income indicates a more productive use of one’s human capital and should ensure that an individual does not migrate. This theory is similar to the first but is more comfortable explaining situations where individuals are rejected from their place of migration due to cultural differences, poor education, or language barriers.

Secondly, a wide range of other theories explain migration decisions by implicitly suggesting that individuals maximise particular social or cultural ties when they decide to migrate (Massey and España, 1987; Constant and Massey, 2003; Haug, 2008; De Vroome and Van Tuberg, 2014). These theories acknowledge that people may have attachments to places, whether due to people, landscape, or culture, that cannot be explained economically. They vary by which ties they emphasise: family ties (Dustmann, 2003) or general ties to a community. Still, they are unified in their emphasis on the ‘dynamic social process’ (Massey and España, 1987, p. 739) behind migrants’ return decisions. Theories emphasising social or cultural ties predict that urban-rural migrants are more likely to have relatives in rural areas (Dustmann, 2003). They also indicate that older people are more likely to relocate from cities as, due to the history of generational migration in China, elderly urban residents are highly likely to have been born in the countryside (Constant and Massey, 2003).

These theories all explain particular empirical phenomena, and it is important not to misconstrue them as mutually exclusive. Each theory purports to explain micro-level behaviour, so it is perfectly possible that they can each explain the decision processes of a portion of the population or even a portion of an individual’s decision processes. Indeed, Constant and Massey (2002), when testing the neoclassical theory and new economics of labour migration, concluded that both processes were at work in their German sample.

2.3 Hypotheses

The above theories yield extensive hypotheses predicting the demographic makeup of urban-rural migrants, generally predicting that they will be less educated, older, and earn a lower income (Chen and Liu, 2016). While these hypotheses are interesting, the purpose of my research is policy evaluation, so its focus is necessarily confined to the opposite demographic: those who are highly educated and employed. This affects the scope of my hypotheses significantly. Instead of focussing on demographic predictors across society, as in the literature, all my hypotheses apply only to highly educated and employed individuals. Since my sample is limited, and I cannot rely on demographic characteristics to predict migration, I focus on self-reported migration motivation. By asking survey participants why they would consider migrating, I can directly test the micro-mechanisms proposed in the theories above, albeit within a narrower demographic. In light of these methodological considerations, I set out six hypotheses below, each drawing upon the logic of the abovementioned theories.

Firstly, within the neo-classical approaches, I distinguish between two types of motivation: immediate and long-term economic threats. If an individual has difficulties finding work and believes that they will find work easier in rural areas, then it is obviously in their income-maximising interest to migrate. However, this cost-benefit calculation could also be performed over the long term, considering the cost of living in urban and rural areas. For these reasons, I propose that individuals struggling to find a job and struggling with the cost of living will find these particularly motivating reasons to migrate. These mechanisms constitute my first two hypotheses.

H 1.1 (short-term economic threat): Individuals who report struggling to find a job as a reason to migrate will themselves be more likely to migrate.
H 1.2 (long-term economic threat): Individuals who report the cost of living as a reason to migrate will themselves be more likely to migrate.

Secondly, within theories maximising social or cultural ties, I distinguish between an immediate need for social ties and long-term social or cultural ties with places. An individual might feel a pressing need to move home if they were homesick or felt an obligation to their family. However, they might also have less pressing but relevant long-term considerations about city or rural lifestyles. These might be equally effective at motivating them to move to the countryside. I expect that these will both be motivating factors. So those who are homesick, feel obligations toward their families or report missing rural lifestyles will be more likely to migrate to a rural area. These are my third and fourth hypotheses.

H 2.1 (short-term ties): Individuals who report homesickness or family responsibilities as a reason to migrate will be more likely to migrate.
H 2.2 (long-term ties): Individuals who report rural or city lifestyles as a reason to migrate will be more likely to migrate themselves.

Finally, I set out a separate set of motivating factors not derived from the literature but implicitly from government actions. The government’s incentives presented in the previous section include economic and cultural measures. However, they also invoke notions of altruism and duty, which do not fit naturally with the literature’s implicit emphasis on self-interested maximisation (China Internet Information Centre, 2024a). Though the literature provides no theoretical backing to these hypotheses, I outline two more mechanisms and corresponding hypotheses by which individuals may be motivated to relocate. Firstly, I predict that those who care about improving the lives of rural people will be more likely to relocate to a rural area. As outlined in the sections above, much of the government rhetoric around rural relocation emphasises supporting local communities and poverty alleviation. The TV series The Daughter of the Mountains (Beijing Youth Daily, 2022) is a key example of this messaging emphasising ascetic and altruistic motivations. Secondly, to assess the effectiveness of government actions further, I will directly measure people’s self-assessment of the impact of the news and media on their desire to migrate. This leaves us with our final two hypotheses.

H 3.1 (altruism): Individuals who report improving the lives of rural residents as a reason to migrate will themselves be more likely to migrate.
H 3.2 (media): Individuals who report news or media inspiration as a reason to migrate are likelier to do so.

It should be noted once again that these hypotheses are by no means intended to be mutually exclusive. As mentioned above, all mechanisms may operate simultaneously within different people or the same. These hypotheses allow us to explore various possible reasons for rural relocation.

3. Data And Methods

I tested these hypotheses with an online survey of 110 Chinese citizens who were residents in cities. The survey consisted of a webpage, accessible via HTML link, with tick boxes in response to several questions. It was translated into Mandarin by Professor Seah Shuoh from the Department of Public Administration of HUST University. The translation was read and corrected by several other Chinese students to ensure an appropriate tone and message. This was essential since I am not fluent in Mandarin and needed significant input to ensure that reported motivations would translate well into the Chinese cultural context. The complete survey questionnaire is available upon request in both English and Mandarin.

I strived to ensure accurate self-reporting of motivations by assuring participants that their responses would remain anonymous. Since the survey was completed in participants’ own time, with no supervision and no physical interviewer, it is hoped that participants would disclose their motivations as honestly as possible. The survey specified that it was being conducted by an ‘intern’ but did not mention that this intern was British. This should again prevent potential power dynamics or suspicion from clouding participants’ motivation. It should be noted that online state surveillance is rigorous in China (Walton, 2001) and may affect the honesty of individuals’ self-reporting; however, the potential effect of this can be neither quantified nor mitigated.

The survey was distributed informally through academic networks and is not formally representative. However, the summary statistics will show that it captures highly educated individuals from various age groups and self-reported income brackets.

3.1 Variables

My independent variable is an individual’s self-reported propensity to migrate, which was measured with a response to the question ‘How likely are you to permanently relocate to a rural area?’ on a scale from ‘very likely’ (= 1) to ‘very unlikely’ (= 5). Please note that all questions outlined here were presented in translation. To assess my six hypotheses above, I presented participants with the question: ‘If you were to relocate to a rural area permanently, what would be the reason for this? Please mark all that apply’. Participants were then presented with eight options of migration motivations, outlined in Table 1 below. Each option assessed one of my hypotheses, apart from H 2.1 citing short-term social/cultural ties and H2.2 citing long-term social/cultural ties. The latter was evaluated by separate variables reporting city and urban lifestyles as motivations for relocation. The former was assessed by two separate variables, which recorded whether an individual experienced homesickness or felt an obligation to family members. Any responses to these questions were coded as binary checks, as represented in the summary statistics table below.

Despite my specific sample composition of highly educated individuals, I nonetheless included several appropriate demographic controls. Gender (female = 1, male = 2) has been demonstrated to affect migration decisions by Song et al. (2021), who found that of the workers who returned home during the Wuhan lockdowns, women were far more likely to continue staying at home than men, for whom no significant effect was observed. Education (middle school = 1, college = 2, college or above = 3) has been demonstrated to have a similarly strong effect by Shuangshuang Tang and Pu Hao (2019), who found that in districts around Nanjing and Suzhou, less educated migrants are more likely to return to their hometown. Perceived income sufficiency (1 = not enough at all, 2 = not enough, 3 = income just meets expenses, 4 = comfortable life) has been demonstrated to have a negative effect on the propensity to migrate at low incomes, then a positive effect once savings effects come into play. This was also evidenced by Govert Bijwaard and Wahba (2014), who used data on 16,974 labour migrants to the Netherlands. Finally, an individual’s Hukou status (1 = rural, 2 = urban – Hukou status change: 1 = yes, 2 = no) would be entirely expected to influence their return decisions, as someone with a rural Hukou, who was born in the countryside, would be expected to be more susceptible to homesickness than someone who had never lived in the countryside.

Finally, several additional questions were posed to flesh out further demonstrated mechanisms. Among these was a question about an individual’s migration destination and whether it would be their hometown. An additional question was asked about their expected occupation if they did migrate, with options including ‘work for someone else’, ‘start your own business’, ‘care for family’, ‘retire’, and ‘other’. An additional question was asked about individuals’ employment status to verify that the sample consisted of individuals in full-time employment(?) or those enrolled in a full-time study program.

Table 1: Summary Statistics.

There are a few notable features of my data to point out. Firstly, and most notably, the particularly high average education ensures that my sample is suitably representative of highly skilled urban dwellers. Occupation is not displayed in this table, but only two respondents to the survey were not either employed in full-time work or enrolled in a full-time study program, so my sample is similarly representative of the employed and educated demographic I am targeting. Desires to relocate among respondents were lukewarm overall, as evidenced by the average response of 2.35. If we look at Table 2, we see that responses are skewed toward ‘unlikely’, with only a very small group of individuals indicating that they would be very likely to move to a rural area.

Table 2: Grouped responses for propensity to migrate. The table displays answers in response to the question: ‘How likely are you to permanently relocate to a rural area?’

4. Results

4.1 Primary Model

My central estimating equation is a multivariate linear regression. I build it up progressively, first using a control model (Model 1) with only demographic variables included. A model with only independent variables is included (Model 2), and finally, the complete model is included with all variables (Model 3). As can be seen below in Model 3 of Table 3, nearly all my hypotheses are falsified (as no significant relationships are returned), except for (H 1.1), the effect of an individual feeling that it is hard to find a job and (H 2.1), a particular feeling either that they had family responsibilities that required them to move, or that they were homesick. Both were positive and increased the chances of an individual feeling they were likely to relocate. These results are robust for the progressive omission of each control and independent variable. Regression results for these robustness checks are available upon request. A complete discussion of the meaning of these results can be found below in section 4.1.

Since my sample is highly educated and employed, it is unsurprising that no significant results are returned for controlling education and perceived income. There is simply not enough variation in my sample for these controls to have a substantial effect. Similarly, it is entirely expected that Hukou status should have such a significant effect upon propensity to relocate. The Hukou status coefficient is negative, and since the coding for answers was 1 = rural, 2 = urban, this means that individuals with an urban Hukou are less likely to relocate than those with a rural Hukou. This is unsurprising and in line with our expectations that those with a rural Hukou will have stronger ties to the countryside.

The question of why these significant relationships are not displayed in Model 2 (except for homesickness) can be explained as a product of omitted variable bias. For example, the variable ‘hard to find a job’ is likely to be significantly associated with the demographic variables and thus will reflect the effect of the omitted demographic variables in addition to its effect. When the demographic variables are added back into the model, the variable ‘hard to find a job’ no longer captures the partial effect of the demographic variables but now reflects the ‘true’ effect of finding it hard to get a job, which, it turns out, is significantly associated with the outcome.

Table 3: Linear regression estimating the effect of motivations for relocation on propensity to relocate.

It is also important to note that the significant independent variables above are not the most commonly reported reasons why those who report as being likely to move justify their own decisions. In fact, their most reported decision motivations are a high cost of living, homesickness, and the attractiveness of the rural lifestyle. We understand the discrepancy between these results in the figure below, which shows relocation motivations for all respondents and respondents for whom it was possible or likely they might relocate. As we can see, the cost of living, homesickness, and an attractive rural lifestyle are motivating factors listed by all respondents. However, they were insufficient to encourage a large part of my surveyed group to relocate. My regression results identify what is unique about those likely to move to rural areas.

Figure 1: Relocation motivations for all respondents (left) and those for whom it was possible or likely they might relocate (right).

4.2 Discussion

So, what does this mean for our hypotheses? We have found that theories of ideal motivation and media influence do not hold for highly educated Chinese urban residents. Interestingly, both neo-liberal theories and theories of social ties have only been partially validated. For each one, I argue that the more pressing form has a significant effect, but the less pressing form has no effect. For example, when it comes to economic factors, the threat of unemployment is sufficient to motivate someone to relocate. However, even though many say they believe that the cost of living is lower in rural areas (see Figure 1), this does not significantly motivate them to relocate. Similarly, the pressing need that an obligation to care for a family member might exert has a significant effect, whilst what could be considered purely aesthetic considerations of cultural and lifestyle affinity has no considerable effect. The same applies to the pressing feeling of homesickness. Again, what could be seen as more pressing has an effect, and what is attractive but not urgent has none.

4.3 Further Models

As mentioned above, participants were also surveyed on what they expected to be doing if they did relocate. I now use these results to further explain the pragmatic mechanisms I have outlined above. This is particularly relevant in the case of homesickness and family responsibilities. One could argue that homesickness does not necessarily represent a pressing need to relocate but could be associated as much with aesthetic considerations. When we reconsider the data, this claim is shown to be unfounded. As shown below in Table 4, there is a significant positive association (0.276*) between a respondent feeling homesick and intending to relocate to care for their family. This is also the case for those who report family responsibilities as a reason for their desire to move to the countryside. It is also interesting to note that having an urban Hukou has a negative effect (-0.268**) on the likelihood that an individual feels homesick, as someone who was not born in a rural area could not be expected to feel homesick about it.

It is also significant to note that for those who report difficulties finding employment as a reason to relocate, this disposition is not significantly associated with any intended occupation upon relocation. This can be observed in column 3 of Table 4 below. This is surprising, given the Chinese government’s emphasis on and expectation that young students will see entrepreneurship in rural areas as an alternative to a competitive and tough job market in urban areas. It appears instead that those who struggle to find a job do not generally intend to return to rural areas for one particular occupation but may instead opt for a variety of options.

Table 4: Linear regression estimating the effect of planned relocation occupations on selected relocation motivations.

5. Conclusion

My results indicate that theories maximising both income and social ties are, to a certain extent, correct. However, when applied to highly educated individuals with secure employment and livelihoods, they only apply in their strongest and most immediate form. This is shown by the fact that the only motivations for relocation which displayed significant association with actual relocation possibilities were difficulties finding a job, family responsibilities, and homesickness.

These findings can inform current policies promoting rural revitalisation and urban-rural migration in China. Firstly, the Chinese government should moderate expectations about the efficacy of both government rhetoric and incentives to convince educated individuals to migrate. My results indicate that the most promising reasons why highly educated and employed people move are largely outside of direct government control: individual family and employment situations. Many of the policies currently implemented by regional and local administrations, such as entrepreneurship funds, relocation subsidies, cultural activities, and media rhetoric, may work for those with little to lose from migration, but my data firmly suggests that the highly educated are not easily convinced. This is further reinforced by the fact that there is no significant relationship between either participants’ desire to help the rural poor or their experience of inspiring media and a desire for rural relocation. People move only when they must, and those who are highly educated and secure rarely have a real reason to move. In this vein, it is crucial to bear in mind the overall proportions of my survey. Only less than 2 per cent of my respondents reported that it was very likely that they would move to a rural area. My data suggests that large-scale urban-rural migration of the highly educated in China is highly unlikely to take off, and this is also likely to become an increasing policy challenge as China’s population grows more educated. Indeed, China has recently seen unprecedented growth in per capita higher education (Fleisher and Jia, 2024), the precise demographic investigated here.

However, my results are generalisations, and so a more modest expectation of urban-rural migration would be entirely in line with these findings. Accordingly, these findings clearly suggest certain paths will be more convincing for the highly educated than others. The lack of any significant relationships between rural and city lifestyles suggests that the marketing campaigns upon which the Chinese government has embarked will be largely futile. People will not move solely because they think the countryside is beautiful. Similarly, whilst the stories recounted in my introduction are inspiring, it would be unjustified to expect that many people will have the drive or will to do the same. Instead, two key policy directions must be pursued to promote urban-rural migration. Firstly, it is of key importance to offer stable and secure employment opportunities in rural areas to ensure that those who cannot get a job in cities can find one in the countryside (Eckhardt, 2023). In addition, proper advertising of these positions is essential, as those with no knowledge of the opportunities presented to them will not move. Secondly, support must be offered to those who would return home and care for their families. Since the aim is rural revitalisation, this support should be paired with incentives to engage in part-time entrepreneurship and employment to continue to stimulate the local community. My results suggest that following these two policy directions is the most likely way to effectively incentivise urban-rural migration and promote rural revitalisation in China.

Luke Dorman

© The Author(s) 2024. This is an open access article under the terms of the Creative Commons Attribution 4.0 International licence, which permits use, distribution, and reproduction in any medium, provided the original work is properly cited.