Kam Wing CHAN

Accurate Chinese migration statistics are notoriously hard to come by. United Nations researchers once called the Chinese floating population—the largest migrant group inside China—“statistically invisible.” While reforms to statistical reporting systems have begun to emerge, the concept of interregional migration in China remains difficult to record.

Any meaningful analysis of Chinese migration must begin with an understanding of the hukou (household registration) system and its relationship with migration. In China, migration was an area of heavy state control in the past and active state regulation exists at present. People wanting to change residence permanently are required to get approval from one or more authorities. A change in residence is deemed legal only when it is formally approved and registered with the public security authorities. For urban dwellers, changing residence within the same city or town (to move to a new apartment, for instance) and residential changes caused by marriage are generally permitted.

Similar freedom is also given to rural residents moving within their rural area because of marriage or other family reasons. But formal (permanent) moves crossing city, town, or township boundaries are heavily regulated and require the possession of a migration permit issued by the public security authorities. The permit is granted only sparingly when there are extraordinary reasons and when it is seen as serving (or at least is not at odds with) the central or local state interests defined in various policies, such as controlling the growth of large cities. Hence, to an ordinary person without state connections, getting a migration permit for moves from rural to urban areas or from smaller cities to larger cities is still very hard, if not totally impossible. The hukou system in the prereform era functioned as a de facto internal passport mechanism; today it still serves many similar functions, although peasants can now travel to many places to take up jobs or stay with relatives temporarily. The catch is that temporary migrants are ineligible for many of the benefits and rights of permanent local residents.

Massive rural labor outflows (workers in search of employment) have been the most important aspect of China’s geographic mobility in the last quarter century. “Rural migrant labor” is defined as a working population (from the countryside) that moves to a destination without a local hukou. (See table 1.) By inference, it is a subset of the “floating population.” Most rural migrant labor is unskilled labor; only a small percentage are skilled craftsmen and traders, and these are often self-employed. A portion is seasonal, operating in synchronization with farm work schedules (the outflow is larger in winter when there is not much work on the farm). Numerous large-scale national surveys of rural migrant labor have been conducted, especially since the early 1990s, when this group started to increase quite rapidly. Many surveys of this kind are one-time studies and are not strictly comparable; the national rural migrant labor estimates are usually derived from the percentages generated from these sample surveys.

Migration Trends since the Early 1980s

Some general migration trends are identifiable in the past twenty-five years. Despite the general surge in migration, the annual volume of hukou migrants has remained quite stable, at between 17 and 20 million. The rate has actually declined slightly, relative to the size of the Chinese population. An analysis of the figures from the Ministry of Public Security from 1993 through 2005 shows that between 80 and 90 percent of all hukou migrants were recorded in urban areas. The stability reflects strong government intervention in this area.

On the other hand, the size of the non-hukou migrant population has clearly been rising since the early 1980s. The “floating population” started to grow rapidly in the mid-1980s to about 70 million in 1988, then dropped somewhat from 1989 to 1991 due to an economic austerity program; it then regained momentum around 1992 through probably 1997, reaching 100 million at that point. The current figure is about 150 million. Similarly, between 1992 and 2006, the number of rural migrant laborers has more than doubled, from 53 million to 115 million.

The trend among the “temporary population” subset of the non-hukou migrants is much less consistent: two data points, 2000 and 2005, are far larger than the other years’ figures. The 2000 and 2005 figures are more supposedly accurate than the figures in other years because they are from a full census (in the case of the 2000 figures) or a 1 percent sample (in the case of the 2005 figures), while the data in other years draw from a 1-per-1,000 sample. Earlier research suggests that the 2000 census figure is, however, likely to be overcounted; it is also likely that the 1-per-1,000 samples may have undercounted the migrant population. If this same logic can be applied to the figures in 2005 and 2006, then it is reasonable to believe that the temporary population, defined as such, was between 130 and 150 million from 2005 to 2006.

Another careful examination of the data will show that there was a slowing down of migration from 1996 to 1999. For example, the temporary population in those years only inched up slightly; the average growth rate of rural migrant laborers also dwindled to 4.8 percent per year in the years between 1995 and 1998, compared with 7.3 percent per year from 1992 to 1995. It is believed that this slowdown in rural outflow was related to the sluggish performance of the urban economy, job competition from laid-off workers of urban state-owned enterprises (SOEs), increasingly protectionist policies used by local governments against recruitment of outsiders, and improvement in the rural economy, at least between 1996 and 1999.

The Geography of Migration

Significant disparities in wages between the urban and rural sectors and among regions underlie a great portion of labor migratory flows in China. The bulk of migratory flows in the last twenty-five years involves predominantly those who are not changing their hukou location (non-hukou migration), mainly rural migrant labor. The root cause is lack of sufficient gainful employment in the countryside in many agricultural provinces. Because of serious institutional barriers—mainly the hukou system—the rural and urban population segments and the labor markets operate as two largely separate circuits or strata. The choice for rural migrant labor is mainly between a farm job (or no job) at home and a low-end job in a city. Rural migrant labor moves across different geographic regions to benefit financially, which can be broadly considered in terms of the balance of the wage differentials and living cost differentials between the origin and the destination. Most rural migrants go to nearby towns outside the villages, but others cross thousands of miles to big cities on the coast. Two major sources of available data of a different nature allow us to examine the national geographic patterns of migration in the 1990s and beyond.

Population Data: Migration Importers and Exporters

The first source is the population data derived from full censuses (1990 and 2000) and the 1 percent national population surveys in 1987, 1995, and 2005. In the data covering migration flow, “migrant” is defined as a resident (staying more than six months or one year in an administrative unit) who lived in a different administrative unit five years earlier. Those data, plus data from the 1982 census, also provide information on the size of the non-hukou population (migran
t stock), based on roughly similar but not exactly the same criteria. The 1995 survey reports a total of 33.23 million migrants crossing county-level boundaries in the preceding five-year period. Seventy-two percent of the intercounty migration was within provinces; the remaining 28 percent (9.2 million) was interprovincial migration (IPM).

From 1995 to 2000 the volume of migration increased substantially. With the caveat that the 2000 census probably overcounted migrants, it appears that intercounty migration doubled between 1995 and 2000. Using information from a 1 percent microdata, one can also classify and estimate the flows by origin and destination. The predominant flow was from rural to urban areas (50.32 million), followed by urban to urban flows (45.70 million).

Total IPM volume has increased significantly since 1990, from only 9.2 million between 1990 and 1995 to 38 million between 2000 and 2005. The different definitions and procedures used for collecting the data obviously account for a small part of the increase; the remaining increase can be assumed to be real. IPM also accounts for an increasingly greater share of all intercounty migration in the 1990s, for example, from only 28 percent between 1990 and 1995 to 44 per cent between 1995 and 2000. It is very likely that the same is also true in the first few years of the twenty-first century.

The largest IPM flows are overwhelmingly toward the coastal provinces, with Guangdong being the most popular destination and the lower Changjiang delta being the second-most-popular destination. Flows involving major IPM players (provinces, autonomous regions and municipalities) are basically unidirectional. In other words, these regions have overwhelming one-way flows: either in-migration (such as Guangdong) or out-migration (such as Sichuan), with only very small flows in the opposite direction.

Over time, and especially between the first and second halves of the 1990s, a greater concentration of IPM went to the coastal provinces. In the second half of that decade, the five coastal provinces with the largest net IPM received 54 percent of all IPM, compared with only 39 percent in the first half of the decade. The high concentration in the coastal provinces continued to be maintained between 2000 and 2005 (55 percent). In the 1990s there was also significant convergence into one single province (Guangdong), which received 34 percent of all IPM between 1995 and 2000, compared with only 20 percent in the earlier period. The pattern was slightly altered from 2000 to 2005 by Zhejiang’s rapid rise to become the second net importer of labor, with a net intake of 11 percent. Guangdong dropped slightly to 27 percent in this period.

On the net exporter side, Sichuan, the largest exporter province, was the single dominant net exporter between 1990 and 1995 with ?14 percent, far exceeding the second-place net exporter (Anhui, whose net IPM was ?7 percent). Sichuan’s dominance was slightly eroded between 1995 and 2000 (net IPM ?12 percent), partly because of the split of Chongqing from the province. More importantly, the second-, third-, and fourth-largest net exporters (Hunan, Anhui, and Jiangxi) all had net IPM values much closer to those of Sichuan between 1995 and 2000 (?7.6 to ?9.0 percent), and that trend continues between 2000 and 2005: the four largest net exporters have about the same net IPM values (?7.4 to ?8.4 percent). In other words, in those fifteen years, while there was a convergence of the IPM flows into one or two provinces, sources became more diverse. These changes seem to be related to the intensification of the regional industrial restructuring beginning in the late 1980s, whereby inland provinces lost proportionally more manufacturing jobs to the coastal provinces in the second half of the 1990s, giving rise, in particular, to the emergence of Guangdong as the industrial leader, indeed the “world’s factory.” The pattern appears to be consistent with the diffusionist paradigm of migration in which migration has been adopted as a labor strategy by an increasing number of households in a greater number of provinces in the noncoastal provinces.

Certain provinces experienced some interesting ups and downs during these three periods. Most striking is Zhejiang. For a time, migrants from this province went almost everywhere in the country (and to many parts of Europe too). But the province has gone from being a major net exporter of migrants (the seventh-largest net exporter between 1990 and 1995) to a top net importer of migrants (the third-largest between 1995 and 2000 and second between 2000 and 2005). This shift is related to the economic success and industrial job growth of the province. On the other hand, the position of Xinjiang Uygur Autonomous Region has slipped in the ranking in the years since 1990. Its net IPM percentage dropped from about 5 percent between 1990 and 1995 to only 1 percent between 2000 and 2005.

The fact that the poorest and relatively populous provinces like Guizhou are not among the lowest in migration rank shows that it is not simply abject poverty that drives long-distance migration. While most provinces are predominantly either importers or exporters, there are some notable exceptions, such as Jiangsu, Fujian, Shandong, Hubei, Hebei, and Zhejiang between 1990 and 1995; Jiangsu between 1995 and 2000, and Shandong and Hebei in both periods from 1995 to 2000 and 2000 and 2005. The fewer number of provinces in the more balanced in- and out-migration category is consistent with the greater regional economic specialization or polarization that took place in the 1990s. It is also interesting to note that Guangdong also registered a significant out-migration (1.7 million) in the latest period (2000–2005). A detailed examination of the 2005 data suggests that the out-migration was largely to the provinces from which in-migrants had originally come (such as Hunan and Jiangxi). The out-migration from Guangdong is likely a return migration rather than a diversion of migrant labor from Guangdong to other coastal provinces.

Rural Labor Data and Surveys

The second major set of migration data draws from national surveys of rural households done by the National Bureau of Statistics and rural migrant labor surveys undertaken by other agencies or research institutions. The data provide some direct and useful information about this special group. Because rural migrant labor, defined on the basis of migrants’ hukou status, is a rather unique Chinese phenomenon, some of the migration statistics compiled are less standard. The national estimates are all derived from large sample surveys that were conducted at different times, often using only broadly the same definition of rural migrant labor. As mentioned earlier, this kind of rural labor migration study covers only labor migration from the countryside, regardless of a migrant’s length of stay at the destination. Therefore, these surveys capture more comprehensively all labor migrants, in contrast to the census data, which exclude migrants staying in the destination less than six months or a year.

Available geographic data from three separate major national sample surveys undertaken in late 1993 (and early 1994), 1998, and 2004 are particularly useful. According to the 1993 study, the stock of rural migrant labor (those who participated in work outside their village, including seasonal labor) at the end of 1993 and early 1994 reached 51 million, accounting for about one-eighth (12.5 percent) of the country’s rural labor force. The flows were predominantly toward urban areas (77.9 percent). China’s central region was the largest source of rural migrant labor, having the highest labor out-migration rate (15.9 percent) and volume (22.8 million), followed by the western region (13.5 percent and 15.4 million). The eastern region had the lowest rate (8.5 percent) and t
he smallest volume. This pattern is broadly consistent with the findings of other studies of the early and mid-1990s. Because of the large size of the labor force in central provinces, this region accounted for 44 percent of the estimated total outflows. The low rate of out-migration in the eastern region is attributed to the high level of development of rural nonfarm enterprises in many villages and townships, which absorbed local and nearby rural labor. This is not the case for the central or western regions.

Nationally, a great portion of the movement was within counties (36 percent) and even more within migrants’ own provinces (71 percent) in 1994. A decade later, the overall labor out-migration rate almost doubled the rate in 1994 (from 12.5 to 23.8 percent). Among the three regions, the eastern region has gained a significant share of the migrant labor over time. Rural migrant labor was found almost entirely in cities and towns (94.3 percent), with a large percentage in large cities (62.4 percent).

A comparison of the 1993 data with another broadly similar national rural migrant labor survey from 1998 to 1999 reveals some interesting trends. Both of them report a stock of rural migrant labor of about 50 million. While the size and percentage of within-county migration remains quite stable (17–18 million, or 34–36 percent), there is a significant increase in the migration to other provinces, mostly to another region, between 1993 and 1998. This means that rural migrants moved to farther destinations over time. This is consistent with what has been shown earlier based on 2000 census and 2005 mini-census data. Drastic increases in the number of migrants crossing both provincial and regional boundaries are obvious. In 1998, this group accounted for 31 percent of the migrant stock, whereas it accounted for only 18 percent in 1993.

Another regional comparison of the data in 1993 and 1998 in shows that the central region further consolidated its role as the largest source of rural migrant labor crossing provincial boundaries (55 percent in 1998 compared with 46 percent in 1993), and the eastern region is the destination of the vast majority of interprovincial rural migrant labor (increasing from 70 percent to 83 percent). Interprovincial rural migrant labor generated in and from the western region has witnessed the most rapid growth, with its share rising from about one-quarter to one-third in those five years. The share of out-of-province rural migrant labor in the eastern region, however, dwindled from about 30 percent to only about 11 percent of total interprovincial rural migrant labor in the same period.

Most of the interprovincial rural migrant labor in the eastern region stayed within the region (71–72 percent) throughout the 1990s. A large, and increasing, majority of the interprovincial rural migrant laborers from the central and western regions moved to the eastern region (87 percent from the central region and 79 percent from the western region). In terms of the regional outflow pattern, migrants from the western region followed the footsteps of migrants from the central region. Five years earlier, a large portion of out-of-province rural migrant labor (38 percent) from the western region moved within the same region. In the late 1990s, a much smaller percentage still did (14 percent), while almost three quarters of them moved to the eastern region. It has been argued that such moves placed migrants in the best position to benefit from the largest geographic wage disparities possible, and one would also expect that migration would narrow the spatial disparities.

Another comparison between 1998 and 2004 shows that the eastern region has further concentrated rural migrant labor, accounting for 70 percent of all migrants, as compared with only 38 percent in 1998. The two net exporting regions (central and western) show a similar trend; in terms of the destination distribution of all rural migrant labor, the central region still had a much higher percentage in the eastern region than the western region did (because the central region had a higher rate of out-of-province migration).

Guangdong and Sichuan, which are the two provinces with the largest net migration change (in-migration and out-migration, respectively) between 1995 and 2000, are also the provinces with the lowest and highest per capita GDP growth rates, respectively, in the same period. If we add the remittances migrants sent back to their hometowns to our calculus, the overall economic gains of migration to the sending provinces would be even greater. This postulate is consistent with the general pattern of higher rural income growth rates in locales associated with higher rates of out-migration (after controlling for other factors) in China.

Considering the Future

Although previous works have shown that China’s rising migration went hand-in-hand with an increase in regional disparities, one may argue that interregional migration actually helped to narrow spatial economic disparities. From a human capital perspective the Chinese government must continue to promote education and migration as a way to narrow the gaps between the coastal and inland provinces.

More importantly, migration is also closely tied to the reforms of the hukou system. Despite a good deal of official rhetoric about abolishing the hukou institution, the reality is quite different. Almost all the changes to the hukou system and new initiatives have had only marginal impact on weakening the foundation of the system—that is, the separation of two segments of population and discrimination based on that separation. The hukou system, directly and indirectly, continues to be a major barrier that prevents China’s rural population from settling in the city, maintaining the rural–urban “apartheid.” This problem has become more acute as rural migrant labor has become more and more permanent (rather than seasonal) with an increasing proportion of women and children. Increasingly, the problem migrants face is not just employment, but also education (for their children), health, and social security. Despite the good intentions of the central government, local governments do not seem ready to implement any fundamental change to the hukou system.

Further Reading

Cai, Fang. (1999). Spatial patterns of migration under China’s reform period. Asian and Pacific Migration Journal, 8(3), 313–327.

Chan, Kam Wing. (2003). Chinese census 2000: New opportunities and challenges. China Review, 3(2), 1–12.

Chan, Kam Wing. (2004). Chinese hukou reforms and rural-urban migration (in Chinese). Zhongguo laodong jingji [China Labor Economics], 1, 108–124.

Chan, Kam Wing. (1994). Cities with invisible walls: Reinterpreting urbanization in post-1949 China. Hong Kong: Oxford University Press.

Chan, Kam Wing. (2001). Recent migration in China: Patterns, trends, and policies. Asian Perspectives, 25(4), 127–155.

Chan, Kam Wing, Ta Liu, & Yang, Yunyan. (1999). Hukou and non-hukou migration: Comparisons and contrasts. International Journal of Population Geography, 5(6), 425–448.

Chan, Kam Wing, & Wang, Man. (2008). Remapping China’s regional disparities, 1990–2006: A New Assessment of de facto and de jure Population Data. Eurasian Geography and Economics, 49 (1), 21–56.

Chan, Kam Wing,
& Buckingham, W. (2008). Is China abolishing the hukou system? The China Quarterly, 195, 582–606.

Fan, C. Cindy. (2005). Interprovincial migation, population redistribution, and regional development in China: 1990 and 2000 census comparisons. Professional Geographer, 57(2), 295–311.

Fan, C. Cindy. (2005). Modeling interprovincial migration in China, 1985–2000. Eurasian Geography and Economics, 46(3), 165–184.

Lin, J., Wang, G., & Zhao, Y. (2004). Regional inequality and labor transfers in China. Economic Development and Cultural Change, 52, 587–603.

Rozelle, S., Guo, L., Shen, M., Hughart, A., & Giles, J. (1999). Leaving China’s Farms: Survey Results of New Paths and Remaining Hurdles to Rural Migration. China Quarterly, 158, 367–393.

Fallen leaves return to the root.

落叶归根

Luò yè guī gēn

Source: Chan, Kam Wing. (2009). Migration, Interregional. In Linsun Cheng, et al. (Eds.), Berkshire Encyclopedia of China, pp. 1456–1461. Great Barrington, MA: Berkshire Publishing.

Migration, Interregional (Qūjì rénkǒu liúdòng 区际人口流动)|Qūjì rénkǒu liúdòng 区际人口流动 (Migration, Interregional)

Download the PDF of this article