Since January 2020, media outlets have again begun to report on the idea that China is “loosening urban residency restrictions,” or that provinces are “loosening [their] grip on internal migration.” Locally, some Chinese-language reporting also takes claims that restrictions will be loosened at face value, but emphasizes that a relatively low proportion of the growth of the formal population of Chinese cities has been due to migration, and that local governments are less than enthusiastic about the prospect of any hukou liberalization that would cost them money. Our outlook on these changes is even less sanguine. Hyperbolic takes on changes to China’s hukou household registration system—the core, but not the sole set of regulations affecting migrant workers’ movement—are not new. Since the 1990s, commentators have been predicting the abolition of the system, but while the structures facing migrant workers have transformed over time, the ability to discriminate between migrant and local labor markets generally and hukou registration specifically continues to be an important factor in the structure of Chinese migrant labor—which is to say, a large part of the economy as a whole. Fundamentally, management, containment, and stratification of human mobility have always been tools through which spatial differences in the cost of reproduction of labor are harnessed in service to the accumulation of capital. Barring a dramatic rupture in capital’s capacity for spatial control, this will continue to be the case––as with the vast majority of supposed “immigration reform” in the U.S. and Europe, changes to hukou policy represent not the erosion of this system of control, but its accommodation to changes in the broader structure of production.

This most recent round of expectant commentary on the demise of the hukou system results from a set of changes announced nearly a year earlier when, on April 2019, the National Reform and Development Commission (NRDC) specified that restrictions for hukou registration should be totally removed for cities with urban populations between 1 and 3 million, while cities with populations between 3 and 5 million should loosen restrictions. This includes opening hukou registration to residents with over five years of residence history in the city, as well as those with higher education qualifications. The same announcement calls for increases in the scope of new hukou registration for migrants in large, wealthy Tier-1 cities.1 Ultimately, this does not represent a change in the central state’s agenda, which was described by the State Council in 2014, in a declaration of the state’s intention to “fully remove barriers for peasants to settle in towns and small cities, reduce restrictions on settling in medium-sized cities, and set reasonable criteria for obtaining hukou in large cities, and strictly control the population size of megacities.”

This agenda has led to some changes in the legal structures that many migrants face, but does not point toward a gradual abolition of the systems of spatial stratification that have been so profitable in keeping wages low. Rather, this is better understood as the most recent in a series of adjustments to a system of spatial governance that continues to structure the geography of capital accumulation across China.

Histories and Trajectories

Household registration systems (originally implemented under the name baojia) have been used as state tools for taxation and conscription on the East Asian mainland since the Song dynasty, but were expanded substantially under the Republic of China. As an adaptation of both previous registration systems with additional influences from the Soviet passbook system, hukou registration as such was implemented under the socialist developmental regime in a piecemeal fashion throughout the 1950s. Once the system had established itself nationwide, it served as a fundamental element of state management policy throughout the duration of the regime, in particular facilitating accumulation by fixing the boundary between proto-classes of rural grain producers and urban grain consumers.2 Following the opening of urban markets, complete restrictions on mobility by economic means were no longer feasible, but hukou restrictions continued to provide the state with a tool to administer control over newly mobile laboring populations. Beyond specific bureaucratic restrictions on state services and formal urban employment, this resulted in the broader formation of a dual labor market in which holders of local hukou in cities (particularly large coastal export hubs) participated in a fundamentally different economy than that encountered by migrant workers. Those without local registration had little or no access to formal employment, significantly lower wages, and no protection, such as it is, from the local state. Further, this status differential was used to legitimize forms of labor repression that included detention and deportation of workers involved in labor struggles. Status-based detention and deportation on large scales were ended following a period of protest in response to the police killing of migrant worker Sun Zhigang in Guangzhou in 2003, but while restrictions on mobility based on hukou have become less visibly violent in the intervening years, they have not disappeared. Rather, their presence (and absence) continues to be used by both the central and local state as tools to manage labor, mobility, and the appropriation of capital.

Despite the end of individual-level deportation policies, migrant workers—particularly those in lower-waged manufacturing, construction, and various semi-formal employment sectors—still face a legal structure that usually makes permanent settlement in the cities where they work impossible. Specifically, medical care, education, and a variety of other state services were made unavailable to those lacking local residence permits. Some of these restrictions have been loosened for some migrants, in particular through the implementation of “temporary residence permits” in most migrant-receiving cities, which allow holders to obtain some urban services—most often access to the medical system and sometimes education. However, residence permits are not the same as hukou transfer, and in most cases are only available to migrants who can demonstrate continuous residency and a history of tax payments and labor contracts, which remain unavailable to large proportions lower-income migrants.3 Even for those holding temporary residence permits, a variety of urban services are still unavailable. Notably, even where early childhood education is available, college entrance examinations must be taken in one’s place of registration, resulting in a situation where the majority of the children of migrants return to their hometowns for high school. Although employment at the large “state-owned” industrial conglomerates (jituan) has declined as a proportion of total employment, hukou also continues to mark a dividing line for those employed there, with local residence often required for employment as a formal worker, while migrants are employed as “temporary” laborers without benefits.

Beyond continuing gaps in service provision and other, more esoteric policies such as the non-portability of contributions to social insurance funds,4 direct measures continue to ensure that migrant life remains precarious and “permanently temporary,” especially in large coastal cities. In particular, housing clearance—whether politically targeted at elements of the “low-end population,” under the broad banner of reducing “dangerous” informal housing, or in the service of urban upgrading and gentrification—has become an important corollary to official, hukou-status based legal exclusion. While the demolitions of vast neighborhoods in Beijing in 2018 or earlier demolitions of Zhejiang migrant communities there between 1988 and 1995 were not directly connected to hukou policy per se, these policies have clear migration-control aims.

Contemporary Hukou Policy

Contemporary hukou policy follows in large part from the outline published in the 2014 policy paper issued by the State Council, “Views on Advancing Hukou Reform.” Like any other “migration reform” in the US or European context, we do not expect these changes to make life easier for mobile people, or to erode the legal and social structures that make exploitation of migrant labor so profitable. Instead, these “reforms” are changes or adjustments in the set of managerial tools brought to bear in the extraction of value in the process of human mobility between poorer and wealthier regions. Among the reforms that have been implemented since 2014, the formal status differential between “urban” and “rural” hukou—which was a core element of the system under the socialist developmental regime and maintained its importance in the reform period as a marker of legal exclusion from urban life—was formally abolished. The call to loosen certain hukou registration requirements was also taken up, but in most areas where they actually matter, these changes have primarily been focused on increasing the mobility of “talents” (the educated, the wealthy, high-earning employees, and entrepreneurs who have invested in the local economy), with localities competing to provide open channels to offer access in service to economic development. Although the State Council policy paper also gestures towards the dissolution of the binary local/migrant labor market in a broader sense, these changes have not touched down in the lives of low-income migrant workers in any real sense.5

Even the largest change that was effected by the reform has been more muted than one might expect based on the text of such reports. Despite the formal abolition of “rural” and “non-rural” hukou status, bureaucratic divides between rural and urban residence status remain. Former holders of rural hukou remain set apart through proxy factors, including membership in rural collectivities, which is often partially determined through hukou status and confers ownership of collective land. While the role of “urban” or “rural” status is not the same as it was under the socialist developmental regime, the surface-level abolition of “urban” or “rural” hukou has not disrupted underlying policy structures that tie rural people to their places of registration.

As for the supposed abolition of hukou status differentials based on place of registration, this process is in fact only the codification of changes that have been true in a de-facto sense for several years: Hukou transfer to smaller cities had already become relatively easy to access, as the central government, as well as the local governments of small and medium sized cities themselves, see migrants’ cheap labor as an engine of economic development, with smaller cities competing to attract migrant workers by lowering barriers to registration. The relative power of hukou access in smaller cities is limited, however, and provides little in the way of state or city services to those who hold it. It is also worth noting that many of the workers we have spoken to remark that it can be difficult or impossible to transfer hukou back to villages once urban status is obtained, and that even some transfers from one village to another are made very difficult by the associated implications for land ownership.

For the majority of rural-urban migrants, obtaining hukou in small or medium sized cities is actually less valuable than the hukou that they hold in rural hometowns, where their status allows them to maintain a foothold in the collective land ownership structure, and in some cases also provides state benefits that compare relatively favorably to those offered by lower-tier cities. While education and healthcare systems do continue to exclude non-hukou holders, the potential gain of allowing a child to attend school in a Tier-2 or Tier-3 city must be weighed against land rights and other benefits that would be lost through hukou transfer. In this sense, urban hukou transfer is not particularly valuable anywhere but Tier-1 cities, where the majority of migrants have no tangible hope of making a permanent home.

This has been the case for many of the workers we have spoken with and worked alongside: To transfer hukou would be to lose access to the land that represents a final backstop in the face of on-the-job injury, unemployment, or old age. For many of the aging workers who make up the bulk of laborers in the construction industry, for example, despite decades of work in the same city or all across China, returning to one’s hometown is the inevitable end to a life spent building homes, roads, and infrastructure for other people. This return can be bittersweet: State pensions are, to quote one friend, “not even enough to buy cigarettes,” but a return to the village also represents a kind of freedom, leaving behind the constant pressure of the boss looking over your shoulder, endless overtime, and repetitive, damaging labor.

For more than a few, particularly in older generations of migrants, returning to the village often means living cheaply in one’s own home, farming on one’s own time, and often depending on support from a child whose education and hoped-for success was paid for through years of migrant work. Some older migrants we have spoken to, particularly those among the first to arrive in cities in the early 1980s, look back bitterly on restrictions or missed opportunities that prevented them from buying property and establishing themselves as urban residents. However, for most, even younger migrants see the loss of land that would accompany hukou transferas making life more, rather than less precarious, particularly in a wealthy urban center where costs of living are high. One younger woman we spoke with in her hometown in rural Anhui has worked as a nanny in Shanghai for years, and makes her realist view of the situation clear:

Of course, we have to overcome a lot to go out to work. We try hard to get used to other cities, other people, and regardless of if you feel wronged or oppressed, there are things you just have to deal with. […] There is some prejudice against outsiders, but not too much—but anyway, rich people looking down on poorer people is something that you have anywhere, not just in China, but all over the world. That’s something you can’t change. So right now, rural hukou is fine. People who have the option to transfer their registration aren’t willing to, because with rural hukou, you have land to plant. Urban hukou is worthless. I’ve never considered trying to get Shanghai hukou, because I know myself and I know what’s in my reach: My family is here, and I miss home anyway. […] In Shanghai, you can’t mix in with the locals. I don’t have the ability. If I had that kind of money and ability, then we wouldn’t just be talking about Shanghai, I’d want to go out and see the whole world.

If this is the prospect facing most migrant workers—“free” mobility for labor only, with no “mixing in” and an implicit assumption that education, healthcare, retirement, and most of the other reproductive functions of human life will never be transplanted into migrant-receiving cities—then what has changed in the hukou system from the 1990s or early 2000s? Two defining characteristics of contemporary hukou policy distinguishing the system’s adaptation in the past two decades are 1) the widespread implementation of points systems as a tool to quantitatively filter those who aspire to transfer their hukou to large, wealthy cities, and 2) the broad adoption of temporary residence permits.

The implementation of points systems, which began in Shanghai in the 2000s and spread to other large cities, replaces previous schemes in which hukou transfer could be directly purchased or obtained after long periods of residence. The systems standardized and unified a set of characteristics (often including wealth quotas, employment in SOEs, educational attainment, and a wide variety of other class-based factors) required for hukou transfer. In Beijing and Shanghai, hukou is all but unattainable on the basis of the points system, which can set extremely high barriers to entry: In Beijing, regulations issued in 2016 are estimated to have only allowed approximately 10,000 residents to obtain residence permits. Even those who meet points criteria are not necessarily guaranteed status, as is made evident in Shanghai, where out of a million applicants for residency in 2015, approximately a third met the minimum qualifications, but only 26,000 were granted status. This was an infinitesimal fraction of the city’s estimated population of over 24 million at the time and represents only about a quarter of one percent of the city’s non-registered population.

Temporary residence permits, on the other hand, create a mid-level of bureaucratic recognition, which serves to provide some necessary (and less-costly) services to a middle stratum of migrant workers, while also providing a mechanism for surveillance and diminishing the risk to social stability implicit in a large undocumented population. This is accomplished without providing what in other contexts would be referred to as a “path to citizenship.” However, even temporary residence permits should not be understood as a universal system. Rather, they are simply another level of incomplete access that further stratifies the migrant population. For those on the lower end of the migrant labor system, where formal labor contracts are rare, even temporary residence permits are unavailable. If anything, the implementation of points systems and temporary residence permits have made the hukou system more, not less, systematic, despite the supposed “abolition” of its key distinction between urban and rural residents.

As filters, these systems serve two functions: On one hand, they make it functionally impossible for the vast majority of migrants in wealthy cities to ever attain permanent resident status in wealthy cities, maintaining a barrier that has, with minor gaps, been in place since the 1950s. On the other, points systems are intended to smooth out complicated transfer policies, while temporary residence permits provide minimal inclusion as a lever to roll migrant populations into the state’s systems for management and surveillance. The primary “opening up” that has occurred, at least in large cities, is focused on making it easier for “talents” who are identified through high point totals (directly corresponding to higher class status) to migrate. As such, these systems serve to selectively exclude working migrants, while raising wealthy cities’ attractiveness to investors and so-called “skilled migrants,” who are often offered not only a painless registration process, but also direct and preferential access to housing and social welfare benefits. While the specific sets of requirements established under points systems differs from city to city, sometimes dramatically, the barriers to hukou provision are closely tied to the economic prospects of a given city, with Tier 1 cities establishing barriers to entry that are intentionally unobtainable to the vast majority of migrants who live there. This, again, is directly in-line with the seemingly contradictory goals outlined by the State Council and NRDC, of dramatically restricting population growth in super-large cities, while also gradually opening channels for hukou transfer in those areas.


The needs of capital change from moment to moment, and hukou registration policy has become a flexible tool that can maintain a cheap labor force by preventing workers from settling in wealthy cities where they are employed, continuing to cultivate the core elements of a dual labor market by driving down wages for migrants, without sacrificing the management and control provided by registration. Finally, the emphasis of the hukou system as a whole has partially shifted to provide “pulls” to encourage certain kinds of desirable migration. In short, the hukou system in most respects mirrors migration-based labor restrictions and dual labor markets worldwide, which suppress the general price of labor and serve to control unrest.

There are, however, a few key distinctions in the Chinese case. The first is the internal character of this migration system, which is why it has often been characterized as similar to a system of apartheid. At the same time, the attempt to extrapolate apartheid systems and their explicit connection to race onto the Chinese case tends to draw parallels between systems in an inaccurate way. In particular, it misses the ways in which the hukou system is, in fact, utilized in the process of racialization in China, but not along the lines one might expect were it similar, in form and function, to a system of racial apartheid. One early and prominent example of the use of loosened hukou registration, for instance, has been the work to incentivize migration into Xinjiang. Since the 1950s when the system was still forming, barriers to access hukou across Xinjiang have been minimized, in a strategy to encourage Han migration into the province. While relatively unsuccessful in terms of the number of migrants who transfer hukou, this strategy follows up on earlier efforts to mobilize those who otherwise lie at the lowest rungs of the system in a broader settler-colonial project with the intention to construct a Han demographic majority in the region.6 In this respect, the system bears more homology to the intricate class-based legal exclusions and inducements that have historically driven colonial settlement under capitalism.

Beyond Xinjiang however, there are broader goals for the use of hukou to encourage migration, both at systemic levels and in terms of the geographic redistribution of specific populations. On the part of the central government, reforming the hukou system helps accomplish two important objectives: On one hand, reforms such as the issuance of temporary residence permits serve to eliminate one potential source of social instability associated with migration—the existence of a large pool of low-waged workers massed in urban areas but bureaucratically separated from the local state. On the other, the opening of hukou in all but super-large cities is part of a broader strategy that hopes to replicate parts of the migrant-driven development of those cities in smaller areas, all while pushing the migrant working class out of the large cities that they were instrumental in building.

In terms of management, temporary residence permits and the provision of nominal urban status offer a direct channel for local governments to perform state functions of cataloguing, statistics, planning, and taxation in direct contact with migrants themselves, without relying on employers as intermediaries. The minimal services provided by these measures also make room for a certain amount of urban “buy-in” on the part of migrants, particularly those with aspirations toward eventual urban citizenship. This is part of a more general attempt to build up state capacity, which has been perhaps the primary project of the Xi administration. The goal is to functionally connect the central and provincial state apparatuses “all the way down” to these lower levels. Historically, one hurdle to this—alongside de facto decentralization and the autonomy of local bureaucrats—has been the fact that many local governments simply do not effectively perform many of the functions of a modern state. The core of these functions are, as with states everywhere: surveillance, statistics, taxation and policing, after which comes some degree of planning and economic management. For decades, the bulk of these tasks in many areas have either been abandoned or inconsistently delegated to local employers and/or gangsters, with officials either serving in these functions or living a parasitic life on the sidelines. This has meant that attempts to integrate the lower and upper levels of the state finds a mismatch between the two, since so many state functions have either been abandoned or delegated away. Further state functions – in particular, the provision of a social safety net – mark a significant point of divergence between the interests of the central and local state. In most migrant-receiving cities, local tax budgets have essentially been subsidized by migrant labor, while the costs of social reproduction for this sector of the population have been borne by local governments in their place of registration. While we should be skeptical of any government’s claims of an inability to pay for social welfare, calculations of the costs of regularization of status show that an abolition of the hukou system thatwould provide existing migrants with the same level of benefits as the currently-registered population would significantly exceed many local governments’ ability to pay,7 particularly under the existing tax-sharing system in which locally-collected taxes flow primarily to the central government while social benefits are paid for in a decentralized manner by local governments. So far, these shortfalls have been covered as local governments produce large portions of their revenue through land leases, but any real abolition of the hukou system would require either that the central government dramatically change the structure of tax revenue distribution, or that local governments manage to significantly accelerate the pace of land appropriation and exploitation for development, neither of which seems likely.

Nevertheless, the shift in managerial function played by the recent hukou reforms has been a necessary change, since the percentage of the Chinese GDP occupied by large-scale manufacturing has declined in relation to both speculative investment in the real-estate market and the growing service economy. Both have led to shifts in common migration trajectories, as well as central government priorities. Hard boundaries established by previous iterations of the hukou system were crucial first in the establishment of a planned economy defined by local autarky and, following the opening of migration, in the maintenance of a system of production for export under which a larger proportion of temporary migrant workers were employed in medium- to large-scale manufacturing industries, largely living in dormitories and under close supervision of industrial bosses who could in turn be managed by the state. While small-scale employment and informal housing have been part of the migrant experience since migration became possible, the decades following the 1990s have seen the average size of employing units decrease across China, with an increasing proportion of migrants working in small-scale, decentralized employment configurations. At the same time, an increasing proportion of migrant workers find housing on rental markets, as opposed to in employer dormitories. Under these conditions, and particularly following the end of direct deportation policies, the state has needed to find policy levers that can allow for intervention directly in the lives of migrant workers, rather than relying primarily on management by employers. This has been accomplished both through material interventions such as demolition campaigns, including the most recent clearance of migrant housing in Beijing as well as instances such as the demolitions of Zhejiangcun in Beijing, and in managerially-oriented policy transitions, such as the implementation of residence permits.

Migrants (and control over the geographic distribution of migration patterns) are also explicitly seen by the central government as a tool to manage both the process of moving the Chinese economy up the value chain and the threats of economic and social instability represented by a ballooning speculative asset market, particularly in real estate. Initiatives such as plans to move polluting manufacturing industries from Beijing to the newly-created Xiong’an New District in Hebei province clearly demonstrate that moving up the value chain, industrial upgrading, and the development of domestic consumer demand to diminish reliance on export production are not geographically uniform processes. At the same time, increases in domestic consumptionare at their most intense in Tier 1 cities. Meanwhile, the largest traditional employers of migrant labor are moved out of sight, either to built-from-scratch development zones like Xiong’an, or more commonly to smaller cities in the interior. This shift might be seen as an attempt to replicate the success of migrant-driven development that already took place in coastal cities like Shenzhen, Guangdong, and Shanghai by moving cheap-labor production to poorer cities that are now linked to export routes by effective transportation infrastructure. But this phenomenon is itself merely another instance of capital’s perennial flight away from countries or regions where the price of labor and the costs of its social reproduction have grown too high to support. From either angle, the maintenance of a cheap production base while moving up the value chain elsewhere will require exactly the kind of geographical separation that new hukou regulations make explicit: The movement, free or coerced, of low-waged labor out of large cities and into precisely the smaller, poorer cities that are now opening their hukou registration policies.

At the same time, while most migrants are still reluctant to lose access to land, it is worth considering the possibility that, by opening hukou policies in smaller cities to migrants, policymakers will be able to push migrants to settle down and enter into the lower ranks of urban consumers, particularly through the purchase of newly developed housing. While property markets in Tier-1 cities are subject to significant controls intended to prevent a housing bubble (including restrictions based on hukou status that in some cases prevent non-locals from purchasing housing altogether), the sharp rise in property prices in lower-tier cities has been worrying to both the central state and international capital. With over 40% of housing purchases in 2018 being made up of speculative purchases of 2nd homes by wealthy urbanites, mostly in second and third tier cities, one way to prevent a potential crash in housing prices would be to ensure a stronger segment of real demand for the millions of newly-constructed housing units in smaller cities. Under these circumstances, the abolition of hukou restrictions in these jurisdictions is a necessary, if insufficient, step toward filling at least some portion of the stock of newly-built housing, although a large population of migrant workers will remain unable to afford these properties without a significant increase in wages. It’s also important to recognize that elements of this process of urbanization are already underway in the redevelopment and consolidation of rural land under urban control, in which members of dispossessed rural collectivities are commonly offered low-quality “urban” housing as compensation for the loss of their land. As with current dispossessions, any such transition to urban residence will not be easily facilitated through a simple incentivized purchase of urban housing or even rising wages, but rather require the continuation of a violent and fiercely contested process of dispossession that has already been underway for decades and has inspired countless instances of local resistance.

In sum, while previous iterations of the hukou system served to facilitate a series of development models through the exploitation of low-income ruralites, from industrialization under the socialist developmental regime to export-oriented production in the past 40 years, strict barriers to mobility are no longer as desirable in the context of an increasingly service-focused economy oriented toward the construction of domestic demand. At the same time, however, the central government has deep and abiding concerns about the stability and governance of extremely large urban areas, from economic and environmental perspectives as well as direct incidences of labor and social struggle. Under the current equilibrium, then, the NRDC calls for a contradictory expansion of hukou provision by Tier-1 cities, while at the same time those cities continue to firm-up centrally-mandated population cap policies that will not only ensure that the provision of local status to new migrants remains minimal, but also necessitate the expulsion of significant portions of the so-called “low-end population” of migrant workers. Ultimately, the current emphasis on managerial, points-based systems for the provision of hukou in Tier-1 cities is at least partially an effort to shift the scale at which China’s spatial hierarchy functions: Rather than profiting from a stark divide between production in urban centers and reproduction in the rural periphery, current policies at both urban and central government levels function to push the migrant periphery of industrial work and proletarian life out of Tier-1 cities entirely, re-orienting these circuits of production and migration toward smaller cities.8 What remains to be seen, however, is what future will be built by younger generations of migrant workers on the front-lines of the urban service and logistics economy, whether in the kitchens, gyms, and hair salons, or in the vast and growing platform economy of warehouse workers, drivers, delivery staff and home-service workers. These workers will continue to be crucial to the functions of urban life throughout China, and unlike many in older generations have direct connections to the urban social worlds they make possible. Still, at least in Tier-1 cities, they have little to no chance of obtaining full residency in the places where they work.


  1. “Tier lists” for Chinese cities are informal rankings of developmental difference between urban areas. Traditionally, Beijing, Shanghai, and Guangzhou are universally agreed upon as “Tier 1” cities, with Shenzhen as a new addition. In 2018, media attempts to market a list of “15 new tier-1 cities” caused significant debate – that list included cities like Hangzhou, Nanjing, and Chengdu. Many provincial capitals are commonly understood as Tier 2 cities (Hefei, Harbin, Dalian) while smaller cities in wealthier provinces might also be seen as belonging to that grade, such as Foshan or Xiamen. The tiered categorization continues, with perceived levels of opportunity and quality of life decreasing from rank to rank.
  2. We elaborate on the history of the hukou system in Sorghum & Steel, and discuss the implications of this proto-class designation in significantly more detail in Red Dust. It is worth mentioning that while some elements of the Chinese system were adapted from the Soviet propiska passport system, the hukou registration system is widely regarded as having been significantly more effective at actually limiting population mobility. In Sorghum & Steel, we explore how this enabled the state to manage and ultimately constrain the pressure toward urbanization that built through the 1950s.
  3. For one example, see the difficulties facing construction workers in Shenzhen who contracted silicosis and have been struggling to receive compensation despite lacking formal labor contracts.
  4. Social insurance policies are administered locally, with contributions automatically deducted from pay for formal workers. Generally speaking, social insurance policies which receive government funding are unavailable to anyone who does not hold local hukou. While programs such as retirement insurance that are funded through employee contributions are often open to participants with temporary residence permits, or at least contributions made by their employers, these funds are sometimes difficult to access and/or impossible to transfer for older migrants. As a result, accumulated funds become unavailable upon migrants’ return to their hometowns. In some cases, such as in Shenzhen and Guangdong, the benefits provided by cities’ retirement insurance policies are effectively subsidized by large sums accumulated through unclaimed contributions by millions of migrant workers. For a more detailed explanation of the social insurance system, see our translation: Jiangxia, “Left to Rot: The Crisis in China’s Pension System,” Chuang, 2 March 2020.
  5. Calls to “break the binary labor market” have been prominent in Chinese academic and policy papers since the 2000s, often based on rationales that are closer to a right-libertarian belief in the efficiency of a “free labor market” in encouraging economic development. These calls are set against worries on the part of both the central and local state, that managerial capacity would be insufficient to retain control of the population without continued migration management, as well as local governments’ claimed inability to provide real services to migrants.
  6. Since 1949, there have been multiple forms of forced and voluntary Han migration into Xinjiang some of which involved forced hukou transfer. Currently, barriers to registration are low, but migrant workers rarely transfer their registration. It has also become difficult for both Han settlers and locals to transfer hukou out of Xinjiang. Local officials in Urumqi and other cities have refused to drop registrations, and many destination cities refuse to process transfer applications for those with Xinjiang hukou.
  7. See Zhang and Li’s calculations for a sample of 45 cities from 2016.
  8. For an exploration of enterprise relocation and the shifting geography of industry in China, see: “The Changing Geography of Chinese Industry: Data Brief”, Chuang, 5 August 2019.