Chapter 11.

Internal Migration in China during the Reform Era:

Patterns, Policies, and Challenges[1]

Zai Liang

Department of Sociology

University at Albany, SUNY

1400 Washington Avenue

Albany, NY12222

Phone: 518-442-4676

Fax: 518-442-4936

E-mail:

ABSTRACT

This chapter reviews recent patterns of migration in China and describes major policy changes regarding migration in recent years. It also draws attention to some critical challenges ahead. In particular, the chapter shows that China’s migrant population has steadily increased since the late 1970s. China’s costal provinces, which have received millions of migrants, continue to attract even more migrants in the late 1990s and early 21st century. Although earlier waves of migrants were dominated by single males or females, increasingly family members (spouses and children) join the migration process. Another pattern worth noting is the size of China’s return migrant population and the extent to which it affects the migrant-sending communities. In reviewing migration policy changes, the chapter notes the migration related policies have been moving toward a direction that began to treat migrants as equal citizens, but major obstacles remain. The chapter also identifies some major challenges facing policy-makers and migration researchers such as issues with migrant children, health related issues for migrants, labor rights for migrants, and impact of migration on the old age support in rural China.

Although a country with a long history of urbanization, China is not particularly well-known for large-scale migrations. Indeed traditional Chinese society was dominated by the old Confucious doctrine that “if parents are alive do not travel too far (fu mu zai bu yuan yu).” This culture of staying put in one’s own birth place for the rest of one’s life was reinforced by the Chinese government’s strict control of rural-urban migration through the implementation of the hukou (household registration) system. However, since the late 1970s, there have been major changes in the size and patterns of migration (Eckholm, 2003). In fact, it will be difficult to spend a day in a Chinese city without encountering any migrants. The magnitude of migration has been accompanied by a growing research literature dealing with different aspects of this process. For example, PrincetonUniversity’s Population Index lists 34 English publications in the 1980s (on the topic of migration in China) and the number increased to 131 in the 1990s. Of course, there are even more publications published in Chinese. According to Yang (2003), the number of publications in 1985, 1988, and 1995 are 22, 41, and 65 respectively. This literature is also reflected in the number of books published on this topic. By the author of this chapter’s incomplete account, there are 8 books on migration in China published in English and 26 books published in Chinese by the early 2000s.[2]

The main purpose of this chapter is to review recent patterns of internal migration in China, discuss relevant policy changes, and identify challenging issues facing policy-makers and students of migration. Because of the large volume of this literature, this chapter cannot provide exhaustive review of current literature. Instead, the chapter will try to draw broad patterns in big stoke. In particular, it focuses on definitions, data sources, and major patterns of migration (including spatial patterns, remittances and return migration). This will be followed by some discussion of China’s migration policies and changes. The final part of this chapter presents major challenging issues for migration in China.

The Numbers Game

It has become an annual ritual now that each year in the months and days leading to the Spring Festival and (chunjie) and after, there are numerous reports about China’s migrant laborers going home for the holidays. In fact, a typical image of a migrant in China is often portrayed by the media as someone waiting in the long line at railway station, with a big bag carried on the back. Data from the Railway Ministry suggests that during the Spring Festival of 1994, the railway system transported over180 million passengers. Among the passengers, 74% are migrants (Wu, Gui, and Zhang, 1995, p.60). The number declined somewhat 10 years later with 140 million passengers transported by railway in 2004 (World Journal, 2004). The decline probably reflects the fact that some migrants chose to travel home by long-distance buses with more flexible schedules and convenient stops. Clearly, even with a slightly reduced number of passengers, China’s transportation system feels the burden of large numbers of migrant population and has developed strategies to deal with it. However, the exact size of this population is difficult to obtain for two reasons. One is that the nature of this population. Migrants are constantly on the move, from one city to another and from one residential location to another. Thus for migrants who are without hukou, they are not particularly eager to be counted in any survey or census. This is further complicated by the fact that surveys or censuses do not always use the same or “standard” definition of migrants.

Table 1 shows how migration is handled in major national surveys or census. It lists 6 major national surveys or censuses. There are three criteria in defining migrant population in China. One is space (geographical/administrative boundary). For migration to occur, an individual has to cross two administratively defined regions/boundaries. The region could be county, city, or province.[3] The second criterion is time. A definition of migration has to involve a time dimension. The question of consideration is whether to include people who have resided at the place of destination for three months, six months, or a year in the definition of migrant population. These two criteria generally apply to migration measures in any other countries. However, in the Chinese context, we need to include a third criterion which is the status of a person’s household registration (hukou). “Hukou” defines one’s legal location of residence. The use of hukou status is important at least for two reasons. One is that hukou indicates access to benefits or lack of them (Chan, 1996; Solinger, 1999). For example an urban hukou was for a long time associated with housing, employment, medical care, and pension etc. though these benefits have been much less guaranteed than before. In contrast, a rural hukou does not contain these benefits. Second, in the eyes of the government, no move is officially sanctioned until a person obtains local hukou. In the current literature, migrants with hukou are considered as permanent migrants and migrants without local hukou are often refereed as temporary migrants (or the so-called floating population). Thus depending on one’s definition of migrants (along three aspects of time, space, and hukou status), the estimated size of the migrant population can differ significantly.

Aside from regional surveys such as the Shanghai Floating Populaiton Survey and the Beijing Floating Population Census, all national surveys or censuses cover both permanent and temporary migrants. Thus, among national surveys/censuses, the main variations lie in the other two criteria, space, and duration (time). For example, the 1987 China 1% Population Sample Survey uses six-month duration for migrants to be counted at the places of destination. However, when the 1990 Census was conducted, the time dimension dictates that only migrants who have stayed at places of destination for a year or more are defined as migrant population.[4]

When China’s National Bureau of Statistics conducted the 1% Population Sample Survey in 1995, the time/duration criterion was changed to six months, migrants need to have resided at places of destination for no less than six months to be counted as migrants (NBS, 1997). Since then, six-month duration seems to become the standard and was used in the most recent 2000 Chinese Population Census (NBS, 2002).

Given the different definitions of migrant population contained in different national surveys and censuses, it is no wonder different estimates of China’s migrant population exist. For example, in a popular book that portrays different aspects of migrant population, it cited “80 million floating population tribes” in China in the 1990s (Hao, 1996). Deriving from newspaper reports or academic sources, the Chinese economist Cai Fang cited additional estimates of floating population as 80 million to 120 million in the mid-1990s (Cai, 2001). Although not always made explicit, these estimates often refer to the size of China’s floating population (migrants without local hukou). However, when these estimates are cited, it is difficult to figure out how the population is defined in terms of space and time. It was often unclear whether the estimates include migrants between counties as well as within counties. Even less clear is methodologically how they are derived, either from national surveys or extrapolation from local surveys. As a result, the validity of these estimates is often unclear except that the size of China’s migrant population is huge.

The approach taken in this chapter is to use data from national surveys or censuses. The main advantage is that we have a clear definition of the population being discussed so that the degree of change over time can be clearly assessed. The main focus is on China’s floating population because this is the migrant population that has experienced the largest increase since the late 1970s. The data sources include various years of the Chinese censuses along with national surveys of population dynamics, all conducted by the National Bureau of Statistics (NBS, 1997, 1998, 1999, 2000).[5] When the first census since the economic reform was conducted in 1982 the floating population was around 11 million. By the time of the 1990 census, the floating population increased to about 30 million. The early 1990s saw another big jump in the size of the floating population. Data from the 1995 China 1% Population Sample Survey show that the floating population reached 56 million in 1995. [6] (Liang Zai: could you please specify whether all these estimates include intra-county migrants? Also for the estimates in the following paragraphs.)

To see whether the commonly reported size of floating population in the mid-1990s makes sense, the following simple exercise can be conducted. From the 1994 Beijing Floating Population Survey, it was reported that 36.7% of the floating population with duration of residence of less than six months(Source of citation here?)(Zhou, 1996). Assuming the distribution of duration of residence among floating population in Beijing is the same as China as a whole, the total floating population in China(migrants without local hukou) in 1995 could be derived as 88.46 million (56 million+22.46 million). This figure is close to the lower bound estimates of floating population in the 1990s commonly cited in the literature, i.e. 80-120 million.

Data from the 2000 Census reveal a major increase in the floating population, to nearly 80 million, counting only inter-county temporary migrantsfloating migrants with duration of residence of no less than six months (source of citation hereLiang and Ma, 2004)). Again, if the same methodology is applied, the total size of China’s floating population could reach 126 million in 2000.

Spatial Patterns of China’s Floating Population

Data from the 2000 census show that the province that sent largest inter-provincial floating population is, not surprisingly, Sichuan (6.9 million, excluding Chongqing), followed by Anhui (4.33 million), Hunan (4.30 million), Jiangxi (3.68 million), and Henan (3.07 million). These provinces are among the most populous provinces in China (Liang and Ma, 2004)source of citation here).

Where do these migrants go? Early data in mid-1990s suggest that migrants choose coastal region destinations (Liang, 2001). This pattern seems to further intensify in the late 1990s and early 21st century. The province received the largest number of floating population is Guangdong province (21 millions in 2000). Guangdong is a quintessential migrant destination province in China, more than a quarter of China’s total floating population migrated to Guangdong according to the 2000 census. The dominance of Guangdong’s position in attracting the floating population in the new century has accelerated and shows no signs of abating. The number of floating population in Guangdong reached 21.3 million in 2004 (People’s Daily, 2004). Other two coastal provinces Zhejiang (5.4 million) and Jiangsu (5 million) are in distant second and third places. Shanghai ranks fourth place with 4.3 million floating population in 2000 (Liang and Ma, 2004). The size of Shanghai’s floating population is much more significant than it appears, knowing the size of its total population of 16.4 million (add the size of Shanghai total population here). What is also noteworthy is the fact that Shanghai had only 1.7 million floating population recorded in the 1995 China 1% Populaiton Sample Survey (using identical definition as that of the 2000 Chinese census) (Liang, 2001). Five years later, the size of the floating population in Shanghaihas more than doubled, a major increase largely attributable to Pudong’s (eastern Shanghai) development. (Are all these estimates from People’s Daily, 2004? If not, please add the sources of citation for Zhejiang, Jiangsu and Shanghai estimates.)

The major inter-provincial floating migrant streams that dominate the patterns of migrant origins and destinations is reviewed to further examine the spatial patterns of floating population. The origin-destination linked patterns are characterized by two major migration streams: one is Guangdong-centered migration stream (Pearl River Delta region) and the other is Shanghai-centered (or Yangtze River Delta region) migration stream. Within Guangdong-centered migration stream, three adjacent provinces, Hunan, Guangxi, and Jiangxi sent largest numbers of inter-provincial floating population. In fact, Hunan to Guangdong migration represents the largest migration stream in China, with 3.3 million floating population from Hunan were residing in Guangdong province in 2000. Among six provinces that have floating migration streams of more than one million, three are the neighboring provinces (as discussed earlier) and other three provinces are the populous provinces,Sichuan (2.8 million), Henan (1.5 million), and Hubei (1million). The migration streams in the Yangtze River Delta region are characterized by the dominance of Anhui province which sent over 1 million migrants to Shanghai and Jiangsu province (If possible, please also add other major source provinces to Shanghai to balance the discussion with Guangdong-centered migration stream. Please add sources of citation for all figured in this paragraphLiang and Ma, 2004)).

Longer Duration of Residence and Increase in Family Migration and Migrant Children

Students of migration have long realized that migration is demographically a highly selective process. Migration theory suggests that migration is a developmental process and the selectivity of migration changes over time (Massey et al., 1987). Migration often begins with young and single males. Over time, as migrants secure stable employment and find affordable housing, they will gradually bring spouses and children to the places of destination to settle. Thus it is expected that family migration will increase over time and the number of migrant children should follow the same trend. In China’s largest city Shanghai, so far six surveys of floating population have been conducted since early 1980s (1984, 1985, 1986, 1987, 1993, and 1997) (Zhang et al., 1998). The data from the most recent two surveys are used to document the increase in duration of residence in Shanghai, increase in family migration, and migrant children. It should be noted that identical definition of floating population is used in all six surveys, i.e. to be counted as floating population one has to have stayed in Shanghai for at least one day. This feature of identical definition across surveys allows researchers to analyze changes over time.

The first finding from the 1993 and 1997 Shanghai Floating Population Surveys is that migrants have increased their duration of residence in Shanghai. In 1993, the proportion with 1 year6 months to 5 years of duration in Shanghai was about 22.6 percent. In contrast, by 1997, the same proportion has increased to 37.1 percent. Percentage for individuals with duration of residence in Shanghai for 5-10 years has more than doubled (from 5.7 percent in 1993 to 13.6 percent in 1997) (Zhang et al., 1998). The increasing duration of residence indicates that substantial portion of migrants were clearly settling in Shanghai(Please specify whether all these figures are from Zhang et al., 1998, or other sources).

.

The increasing duration of residence for the floating population in Shanghai is, not surprisingly, accompanied by another trend: the increase in school-age children. Overall, the size of floating population in Shanghai did not experience significant change from 1993 to 1997 (the size of floating population actually declined slightly from 2.81 million to 2.76 million). [7] However, during the same time period, the number of school-age children increased from 280,000 to 340,000, an increase of 21 percent. This is not unique to Shanghai. Similar pattern shows up in Beijing as well. Similar to Shanghai, Beijing also conducted several surveys of floating population, more recently in 1997 (OBFPC, 1998). In the 1997 Beijing data, the types of households floating migrants live in can be identified, namely Beijing local resident households (households heads with Beijinghukou), migrant households (households headed by migrants), and institutional households. Based on the 1997 Beijing Floating Population Census and Beijing portion of the 2000 China population census, the changes in the types of household floating migrants live in over time are clearly identified (see Figure 1 Notes: please submit the original file that the figure 1 is generated). In 1997, 32.84 percents of migrants lived in migrant households. By 2000, the number of floating population who lived in migrant households increased to 45 percent, an increase of 12.2 percentage point in a matter of only three years from 1997 to 2000 (Zhang, 2003).

The implications of this change could beare enormous. Increasing proportion of migrants stay longer and bring their families suggest that migrants are in for a longer stay. This is consistent with results from the recent surveys. For example, results from a survey of floating population in Beijing in 2004(notes: which year?) reveal that 54% of the floating population prefers to stay in Beijing if possible (World Journal, 2005a). Thus the mentality of seeing the floating population as people who come, make money, and then leave does not reflect this new demographic reality. In addition, as more family members move to cities, it increases the demand for other services, such as school, heath care, and service for reproductive health (prenatal care and other health delivery). The presence of a large number of school children help create a new urban social phenomenon: the mushrooming of dagong zidi xuexiao (migrant children schools) in urban China. In Beijing alone, there are over 200 such schools in existence (Han, 2003). These schools often are located in areas with inadequate sanitary conditions. Usually owners of these schools are migrants themselves and do not always have experience in teaching or management of schools. Teachers and classrooms are rarely up to the standard of urban public schools (Liang and Chen, forthcoming). This poses a major challenge to China’s Law of Mandatory Education.