Rural migrant workers in urban China: living a marginalised life


Daniel Fu Keung Wong, Department of Social Work and Social Administration, The University of Hong Kong, 1317 K K Leung Building, Pokfulam Road, Hong Kong


The rural migrant worker population in China is attracting more and more attention because of its magnitude and potential economic and social impact on Chinese society. While literature abounds in describing the demographic trends and economic impacts of rural to urban migration, very few articles have been written about the psychosocial impacts of migration on the lives of rural migrant workers in urban China. Drawing on the concept of marginalisation, this article describes the nature and characteristics of marginalised living experienced by migrant workers. More importantly, it examines the underlying policy issues contributing to such marginalised living. It is argued that the Hukou system (household registration system), the process of decentralisation and the obscure role of trade unions have contributed to the experience of marginalisation of rural migrant workers in urban cities in China. Implications for policy changes are also discussed.


According to the Gazette of the State Council of the People's Republic of China (State Council, 2004), the number of migrant workers exceeded 98 million in 2003. Zhang and Song (2003) suggest that, between 1979 and 1999, the increase in the urban population was as much as 222 million people, with more than 10 million people joining the urban population every year. Most of the migrants go to the urban cities in the eastern coastal areas and are from the western and central inlands (Table 1). While Sichuan, Henan, Anhui, Hunan and Jinagxi provinces have the largest number of emigrants, Beijing, Shanghai, Guangdong, Zhejiang and Fujian provinces have the highest number of immigrants (Table 2). Various scholars have examined the driving forces behind China's rural to urban migration, which include: urban–rural income disparity (Zhang & Song, 2003); surplus labour in agriculture (Roberts, 2000); the introduction of the household responsibility system in agricultural reform, which has led to the development of township and village enterprises (TVEs) in the countryside (Iredale, Bilik, Su, Guo & Hoy, 2001); the disintegration of state-owned enterprises (SOEs) and the emergence of private enterprises and a modernised market economy (Iredale et al., 2001); and policies in some poorer provinces that favour out-migration (Iredale et al., 2001).

Table 1.  Inter-province migration trend: East, Central and West.
DestinationYearPlace of departure
  1. Source: Cai & Wang (2003).

200019.7 7.1 7.9 9.8
200015.9 8.623.915.3
Table 2.  Comparison of population growth rate between selected metropolitan/coastal areas and inland areas.
Region Population at 2000 census (10,000 persons)Population at 1990 census (10,000 persons)Comparison between 2000 and 1990Natural growth rate (per cent)
  Growth in persons (10,000 persons)Growth rate (per cent)Annual growth rate (per cent)199019992003
  1. Note: Comparing the population growth rate and natural growth rate, it is clear that metropolitan and coastal areas such as Beijing, Shanghai, Guangdong, Fujian, Zhejiang and Hainan attract large numbers of migrants, whereas inland areas such as Sichuan, Henan, Hunan see outflows of migrants in large numbers.

  2. Source: China Population Statistics Yearbook 2001 (National Bureau of Statistics of China), and China Statistical Yearbook 2004.

National 126,583113,36813,21511.661.0714.39 8.77 6.01
Metropolitan and costal areas (selected)Beijing  1,382  1,082   30027.732.40 7.2 0.90−0.01
Shanghai  1,674  1,334   34025.472.22 3.67−1.1−1.35
Zhejiang  4,677  4,145   53212.851.18 9.02 4.29 3.28
Fujian  3,471  3,005   46615.511.4117.33 5.21 5.85
Guangdong  8,642  6,283 2,35937.553.1316.5 9.92 8.35
Hainan    787    656   13120.021.7818.612.03 9.16
Inland areas (selected)Sichuan  8,329  7,836   493 6.290.5911.45 6.78 3.12
Henan  9,256  8,551   705 8.250.7718.40 7.72 5.64
Hunan  6,440  6,066   374 6.170.5816.70 4.6 4.95
Anhui  5,986  5,618   368 6.550.6218.22 8.6 5.95
Jiangxi  4,140  3,771   369 9.780.9117.05 9.49 8.09

As suggested by the 2000 census (State Statistical Bureau, 2001), rural migrant workers are: usually young, with a higher-than-average level of education than others from their place of emigration; there are more males than females; they tend to work in the private sector, holding jobs in factories and the service industries; they work longer hours, yet have lower household incomes than the urban residents, although their incomes are higher when compared with their fellow residents in the place of emigration (Table 3). Many migrant workers are temporary migrants and have ‘dual occupations’ (Hu, Wang & Zou, 2002). While they work in the fields during planting and harvest seasons, they take up jobs in the cities as restaurant employees, factory workers, construction workers or housemaids during the slack agricultural seasons.

Table 3.  Characteristics of Chinese rural migrants: a comparison table.
Item  Population (per cent)
Urban permanentFloating (migrants)Rural permanent
  1. NA, not available.

  2. Source: State Statistical Bureau (2001).

Age 0–14 17.0519.1126.40
15–24 12.5418.7614.46
25–34 19.1730.7318.99
35–49 26.2320.4220.88
50–59 10.59 5.41 8.76
60+ 14.42 5.5610.51
EducationIlliterate  6.510.319.2
Primary school 13.524.238.2
Junior secondary school 35.952.236.9
Senior secondary school 44.113.3 5.7
Marital statusSingle 17.620.319.4
Married (first marriage) 73.673.271.7
Married (remarriage)  2.0 2.4 1.7
Divorced  1.7 0.7 0.7
Widowed  5.2 3.4 6.6
Working hours per week40–49 hoursTotal78.637.6NA
>50 hoursTotal17.856.9NA

In a small-scale survey Wong and Lee (2003) found that among the 83 rural–urban migrants in a factory in Shenzhen, in the Guangdong province of China, most respondents were young and the majority came from Hunan (41 per cent), Hubei (25.2 per cent), Jiangxi (16.9 per cent) and Sichuan (8.4 per cent). Many were married and over 61.2 per cent had children. Over 25 per cent of respondents had received a higher secondary school education (grade 10 or above). About 25 per cent had primary education or below. Their average monthly income was about 694 yuan (approximately US$1 = 8.3 yuan, as of October 2005). About 50 per cent of the workers sent about half of their incomes back home. On the whole, the respondents in this study resembled the overall profile of rural–urban migrants in China.

Migrant workers are a special group in urban cities in China. While they are needed for economic growth, particularly in the major cities, they do not enjoy the rights of urban residents and are not sufficiently protected either in or outside of work (Li, 2002). They take up jobs that the urban residents are unwilling to do and live in very poor housing conditions, and their children do not have access to public school systems in the cities. All this information points to the fact that rural migrant workers are a marginalised group in urban China. But how marginalised are they?

The concept of marginalisation

There are at least three ways of defining the concept of marginalisation in the literature on migration. First, marginalisation can be defined as an involuntary exclusion from participation in one or more spheres of life. This definition focuses on the involuntary exclusion of an individual from participating in a society. Second, according to Kuitenbrouwer (1973), it can be referred to as a state of relative deprivation characterised by poor housing conditions, lack of opportunity for education, poor health conditions and limited chances to improve income and employment opportunities. This definition points to the nature and consequences of marginalisation on the lives of individual migrants. Third, marginalisation can be conceptualised as a process of excluding an individual migrant from participation in some areas of social life that are viewed as essential in a given society. This definition highlights the importance of examining the underlying processes leading to marginalisation of a particular migrant population. The above definitions have provided the authors with a framework to analyse the experiences of marginalisation among migrant workers in urban China. The analysis will first focus on delineating the nature and consequences of marginalisation experienced by migrant workers, and then it will examine the factors contributing to the state of marginalised living of migrant workers in urban cities. It is argued here that the Hukou system (household registration system) and the policy of decentralisation are among the major policy issues that have contributed to marginalisation among rural migrants in urban China.

Nature and characteristics of marginalisation of migrants in urban cities in China

As a result of a thorough literature review, we were able to gather a great deal of evidence of marginalisation experienced by migrant workers in China. These include employment and working conditions, social security and medical benefits, education of migrant children, housing and discrimination by urban residents.

Employment and working conditions

The occupational composition of migrants and non-migrants is rather different. Little overlap is found between migrant workers and local residents in industrial and occupational distributions. Unlike urban residents, who are primarily employed in manufacturing and other industries, rural migrants are concentrated in service and construction industries (Knight, Song & Jia, 1999). Moreover, most migrant workers take up physically demanding jobs as manual labourers, textile and garment factory workers, toy factory workers and service workers. Indeed, they mainly occupy jobs that local residents disdain (Roberts, 2000).

Since the majority of migrant workers are uneducated and do not have special skills, job mobility among migrant workers is very low. Those who are occupationally mobile can compete only for less technical or non-technical jobs (Yu & Hu, 1998). Coupled with a lack of knowledge of their legal rights, migrant workers have been subjected to a great deal of exploitation. To begin with, quite a large number of migrant workers complained of delays in wage payment, which is a tactic used by the factories as a way of preventing the migrant workers from leaving freely (Liu, 2004). In a study conducted by Ma (2000), it was reported that some factories require new workers to bring capital with them. For example, all manufacturing companies require their new employees to deposit as much as 600 yuan with the firm. Another practice is to pay workers only 1/5 to 1/3 of their monthly salaries and to hold the remainder until the end of the year. Thus, it is not surprising to find that conflicts relating to overdue salaries between employers and migrant workers occur fairly frequently in China. Indeed, drastic measures such as strikes and holding managers or employers in custody have been taken by disgruntled migrant workers (Ma, 2000). The phenomenon of overdue payment is particularly problematic in the building and service (e.g. restaurants and food-related businesses) industries (Yan & Liu, 2004).

The working conditions of rural migrants are also rather disconcerting. Migrant workers are perceived as cheap labour and are paid a very low wage, averaging 300–600 yuan. Tan (2000) found that migrant workers were forced to work 10–12 hours a day, seven days a week, with no guarantee of one day of rest per week (see Table 3). More than half of the factories in Tan's survey failed to pay the government-required social security premium for their migrant workers. A number of small factories even refused to reimburse workers for their medical expenses. Also, over 80 per cent of the factories did not provide maternity leave for female workers. Yet, in some foreign-invested enterprises, the owners were not willing to do anything to protect the health of migrant workers, putting these people in hazardous environments with dust, toxic substances, noise and poor ventilation (Tan, 2000).

In a previous study on rural–urban migrants in Shenzhen (Wong & Lee, 2003), we discovered that the five most stressful work-related issues of the respondents were: a lack of consideration and understanding from the management (m = 3.43, where m = mean score), inadequate rest due to overtime work (m = 3.36), no chance to take up an additional part-time job (m = 3.29), no opportunity to move to a higher paid job (m = 3.26), and unfair treatment by the management (m = 3.24) (Table 4). This information appears to convey similar findings to other studies on the lives of migrant workers in urban China. Essentially, management has considerable power over their migrant employees and can pose unreasonable demands on them. Moreover, migrant workers are powerless and have little government protection. Invariably, this is related to the status of migrant workers in urban cities in China.

Table 4.  Chronic work stress of rural–urban migrants in China (n = 83).
RankWork stressMean* (SD)
  1. Notes: * On a scale of 5, with 1 = no stress and 5 = a great deal of stress. Two focus groups were held at the beginning of the study to generate a list of items on the work stress of migrant workers. Coupled with the work stress mentioned in the literature, we developed this scale. SD, standard deviation from the mean.

1A lack of consideration and understanding from management3.43 (1.52)
2Inadequate rest due to overtime3.36 (1.22)
3No chance to take up an additional part-time job3.29 (1.18)
4No opportunity to move to a higher paid job3.26 (1.41)
5Unfairly treated by the management3.24 (1.55)
6Feeling financially insecure3.19 (1.16)
7Work procedures and duties are too complicated3.10 (1.41)
8Work pace is set at a high level3.06 (1.26)
9Working hours are too long3.04 (0.96)
10Heavy and repetitive work makes me feel exhausted3.03 (1.36)
11Delay in payment3.01 (1.32)
12Lack of social security protection3.00 (1.32)
13Work always piles up2.93 (1.55)
14No chance to raise opinions on improving work procedure2.88 (1.06)
15There is no friend to support me when I encounter work-related problems2.85 (1.26)
16No choice in meals and poor nutrition2.68 (1.47)
17I have no say in working overtime or not2.61 (1.25)
18I feel bored and monotonous about the repetitive work2.60 (1.18)
19Repetitive work makes me feel tired easily2.57 (1.21)
20Accommodation is crowded2.41 (1.16)
21Inadequate protection against industrial hazards2.34 (1.22)
22Having arguments with fellow workers2.25 (1.28)
23Long queue for using accommodation facilities (e.g. toilet or hot water)2.11 (1.22)

Social security and medical benefits

In China, social welfare benefits are closely tied in with one's residence status within the Hukou system. Therefore, migrant workers are largely excluded from social security and medical benefits in the urban cities because they are not official residents of those cities. In a 2003 study on rural migrants in Shanghai, Feng, Zuo and Ruan (2002) found that of the 4,714 migrants surveyed, only 14 per cent had health insurance and 10 per cent had pension plans, whereas 79 per cent and 91 per cent of local employees had health insurance and pension plans, respectively. Although it is compulsory for employers in urban cities in China to contribute to unemployment funds set up by the government, these funds apply only to local residents. Migrant workers who work in the same units are not entitled to unemployment benefits. Thus, it is not surprising to find that the respondents in our study mentioned the lack of social security protection as a source of stress (Table 4).

Education of migrant children

According to Xinhua News Agency (2004), China's migrant children drop out of school at a rate of 9.3 per cent, and those who do go to school are mostly over-aged. Migrant children in China who have never been to school represent 6.85 per cent of the total migrant children population. The greatest difficulty migrant parents face is that educational expenses are too high. This stops them from sending children to the public schools. Indeed, the opportunity for education is also closely tied in with the Hukou system. Children who are registered residents in the cities are entitled to 9 years of schooling. In one study, Fleisher and Yang (2003) found that migrant families have to pay an annual fee ranging from 3,000 to 30,000 yuan per child to have their children enter public schools. Although the private schools charge quite low fees, these schools are usually rather small, lack qualified teachers, and do not have standard teaching materials or adequate sanitation facilities. In 2003, the General Office of the State Council issued a document that aimed to provide access to education for migrant children and prohibit the public schools from charging extra fees. However, this policy has proven, so far, to be ineffective. Although the central government stipulated the policy, the local governments do not carry it out properly. Take Baoshan district in Shanghai as an example; for migrant children to be admitted to the schools, their parents must fulfil four conditions. First, they must obtain four certificates: temporary resident certificate, identification certificate issued by the police station at the place of origin, employment certificate issued by the labour bureaus of destination and certificate of student status from the place of their origin. It should be borne in mind that these certificates cost a migrant worker as much as 200–300 yuan, and deter some migrant parents from putting their children in public schools (Zhao, YH, 2000). Second, the migrant workers must have a fixed abode. Third, they must have been living in Baoshan district continuously for more than 1 year. Lastly, they must have worked in the area for more than 6 months. In a study conducted by several universities in Shanghai (Cheng & Zhang, 2004), it was found that no more than 20–30 per cent of the migrant families could fulfil all of the above-mentioned conditions, and only 15 per cent of those who fulfilled the conditions had actually entered their children in the public schools.

There are two reasons for the stringent responses made by some local governments regarding the admission of migrant children to public schools. First, in Article 8 of the Education Law of the People's Republic of China, it is stated that compulsory education is the responsibility of the local government in urban cities. Migrant children without residency status are not supposed to receive these educational benefits. Such responsibility rests with their local government of origin. This puts migrant children in a very awkward situation because neither of the local governments is willing to fulfil the responsibility (Ju & Ji, 2004). Second, the central government gives no incentive or disincentive to the local governments to implement this policy (Ju & Ji, 2004). Thus, due to limited resources, local governments have no enthusiasm for carrying out the education policies for migrant children stipulated by central government.


Migrant workers are confronted with housing problems in the cities. Since the Hukou system links registration status with social welfare and employment, migrant workers are not entitled to subsidised rental and cannot purchase housing at subsidised prices (Wang & Murie, 2000). In urban China, there are limited numbers of commercial apartments on the market and those available are extremely expensive. For example, a one-bedroom apartment costs at least 1,000 yuan a month in Beijing (Zhao, YH, 2000); the great majority of migrant workers cannot afford to pay for such housing. As a result, many migrant workers are forced to live in the dormitories provided by their employers. This type of accommodation rarely has indoor sanitary facilities (Chai & Chai, 1997; Shen & Huang, 2003).

The lack of opportunity to improve one's housing conditions may be further hampered by the policies set by the various local governments. For example, in Beijing, there is a regulation stipulating that any institution or person leasing housing to non-Beijing residents must obtain a house-leasing certificate from the district or county government, and renew it annually. However, to be eligible for renting, the house or apartment must be privately owned and must be certified by the police bureau as meeting safety standards. The owner must sign an affidavit with the family-planning agency and the police bureau, agreeing to be responsible for the prevention of crime in the house or apartment. In addition, the owner must pay a fee equivalent to 2 per cent of the annual rent. Undoubtedly, this regulation has raised rental costs and deters owners from considering renting their houses or apartments to migrant workers.

Discrimination by urban residents

Although migrant workers play an indispensable role in economic growth in China, they are frequently portrayed negatively in the media. They are perceived as a threat to social stability and are often linked to the increase in crime rates in the cities. They are also perceived as competing with unemployed urban residents who have been laid off from the state-owned enterprises. Furthermore, the general public holds the view that migrant workers are stupid and ignorant, and should be blamed for their misfortunes (Davin, 2000). Thus, it is not uncommon for migrant workers to experience unpleasant social encounters such as verbal disrespect, deliberate avoidance or being looked down upon by the urban residents (Guo, 2004). Indeed, the hostility displayed by local residents has hurt migrant workers’ selfhood and self-respect (Guo, 2004), and has widened the social gap between migrant workers and local residents.

Adverse psychological health as a result of marginalisation

A cursory look at the literature will tell readers that most studies on Chinese migrant workers come from such disciplines as geography, economics and sociology. Very little has been done to explore the psychological consequences of rural–urban migration in China. This is somewhat surprising because the above analysis should have conveyed a strong impression that migrant workers live in a socially disadvantaged position and could easily be suffering from poor psychological health. Indeed, data from one very small-scale study suggests that migrant workers experience a high level of anxiety, and feel nervous, upset and unhappy (Fu, Ye & Chen, 2002). Another study by XF Li (2004) shows that 60 per cent of the 106 migrant workers interviewed experienced mental health symptoms related to anxiety, depression, hostility and interpersonal sensitivity. In view of the lack of studies on the psychological health of migrant workers, it should be considered a necessity to conduct large-scale studies on the psychosocial consequences of rural–urban migration among migrant workers in urban China.

In our previous study (Wong & Lee, 2003), we used the validated Chinese version of the General Health Questionnaire (GHQ-30) (Shek, 1987) to explore the mental health conditions of 83 rural–urban migrants. Using six symptoms as the cut-off point, some 63 per cent of the respondents were considered to be at risk of developing poor mental health (Table 5). The most prominent clusters of symptoms found among migrants were sleep disturbance (e.g. difficulty falling asleep) and anxiety (e.g. feeling anxious all the time). Our study showed that the rural–urban migrant participants suffered from poor mental health and that such mental health conditions might be linked to the work stress experienced by the migrant workers.

Table 5.  Mental health of rural-urban migrants by five factors of the Chinese version of the general health questionnaire (GHQ-30) (n = 83).
Five factors of GHQ-30Mean (SD)
  1. Notes: * We have used the 0-0-1-1 scoring method of GHQ-30 to separate ‘case’ from ‘non-case’. Based on this scoring method, if an individual rates a current symptom as worse, or much worse than before, then, he/she will be considered as having a distress symptom. An individual with 6 or more distress symptoms is classified as falling into the at risk category. According to Shek (1987), there are five subscales in GHQ-30. We have used the 1-2-3-4 scoring method to estimate the mean of each item. SD, standard deviation from the mean.

Sleep disturbance1.31 (0.79)
Anxiety1.28 (1.25)
Depression0.99 (0.64)
Inadequate coping0.91 (0.40)
Social dysfunction0.89 (0.41)
Overall percentage of migrants at risk of developing mental health problems*63 per cent

Reasons leading to marginalisation among migrants in urban cities in China

There are at least three reasons contributing to the experience of marginalisation among migrant workers in urban China. These include the Hukou system, the decentralisation policy and the obscure role of the trade unions.

The Hukou system (the household registration system)

A meaningful analysis of rural–urban migration in China cannot be made without making reference to the Hukou system, which affects the migration experiences of many migrant workers in the cities (Chan, Liu & Yang, 1999). The Hukou system was established in the 1950s. It registers every person at a specific place (usually their place of birth), and requires all changes in residence to be registered with and approved by both the government of the place of origin and that of the destination. This policy became even more restrictive when the Chinese government introduced the food stamp system, and provided low-priced rationing of foods to each individual residing in his or her place of residence. There were two main consequences resulting from implementing these policies. First, it became almost impossible for an individual to move from one place of residence to another. Second, the division between rural farmers and urban city dwellers became wider, with rural farmers lagging behind in economic and social resources (Zhao, YH, 2000).

Indeed, entitlement to public and social services depends entirely on the type of Hukou the individual has. As an urban resident, a person is entitled to employment, health care, housing, pension and food subsidies. None of these privileges, however, are available to people with a rural registration (Cai, 2003). With the introduction of a socialist market economy and the establishment of the responsibility system in rural China, it has become very difficult to restrict people from migrating to the cities. Unfortunately, the Hukou system has done little to accommodate to the changing situation. Indeed, as the previous analysis indicates, the Hukou system has been largely responsible for creating a marginalised group of migrant families who are not allowed to enjoy the same employment, housing, health and welfare benefits as the urban residents in China.

The decentralisation processes in economic development and social welfare

There is a tendency for the central government in China to decentralise control over economic development and social welfare. Economic and social welfare decentralisation came about when, in the late 1970s and 1980s, local government officials called for more autonomy and control over local affairs (Iredale et al., 2001). The fiscal contract system introduced at the end of the 1980s further increased the revenue share retained by localities, and thereby allowed them to implement their own priorities and policies (Iredale et al., 2001). While there are obvious advantages to decentralising the power to local government, problems occur when local governments have different understandings of the laws and ‘notices’ stipulated by central government. To make matters worse, it is difficult for the central government to enforce these laws and notices. Local governments may respond to these policies with little enthusiasm, particularly when such policies appear to jeopardise local interests. After all, the policies lay down only principles, and there is no punishment if local governments do not strictly observe them. Even if central government has detailed regulations in place, the resources needed for enforcement are seriously inadequate. Some notices have no budgetary or personnel support and, in some cases, do not have a government department legally responsible for implementing them.

Taking overdue wages as an example, the central government has issued ‘notices’, such as The Circular of the General Office of the State Council on effectively settling the issue of overdue payment in the construction industry, 2003, Issue 94. General Office of the State Council November 22, 2003, demanding that local governments ensure that private enterprises follow the notices. Unfortunately, local governments do not want to enforce this because ‘there is a common fear among township and village governments that enforcing the Chinese law too strictly will destroy the investment environment and scare away foreign investors’ (Tan, 2000: 302).

Another example is that, although the central government has urged the local governments to provide education for children of migrant workers (e.g. Notices from State Council: The Circular of the General Office of the State Council on Transmitting and Issuing the Proposals of the Ministry of Education and Other Departments for the Provision of Regular Education Assistance (Issue 31), General Office of the State Council December 30, 2003 Issue No. 36, Serial No. 1107), few local governments have taken action to enforce the ‘notices’. Those who have responded to the policies have devised various strategies to protect the interests of local residents and have set stringent requirements for migrant workers to receive the services. For example, as cited above, to limit the number of children who can study in the public schools in the Baoshan district of Shanghai, the local government has set very stringent criteria for migrant children to be admitted to the schools.

There are at least three reasons why local governments do not enthusiastically uphold the notices set by central government. First, both local governments and local people understand that their interests are directly related to foreign investors. It is important to uphold rather than upset those relationships (Tan, 2000). Second, migrant workers are perceived by local authorities as the ‘three withouts’: without legal documents to live in the city, without legal work permits and without legal sources of income. In relation to rural–urban migrants, the primary tasks of the local governments are to collect fees, issue permits, supervise migrants through public security organs, and detain and deport them back to the countryside if they commit any crime (Davin, 2000). Third, there are little incentives or disincentives for local governments to either carry out or not carry out the policies.

The obscure role of the trade unions in foreign-invested enterprises in China

Since the 1980s, the National People's Congress and the central government have instituted a series of laws and regulations to regulate labour relations in foreign-invested enterprises in China. Essentially, all foreign-invested enterprises must abide by Chinese laws. In 1990, the Chinese government adopted the International Labour Organisation Convention No. 144 on tripartite negotiation, and promised to take a neutral stand and allow the labourers and the enterprises to negotiate and resolve their conflicts and interests (Tan, 2000). Moreover, the government also tightened up its regulations and set up labour arbitration and inspection bodies to conduct regular inspections of the factories. Despite all of these efforts, the rights and interests of migrant workers are still not being upheld (Tan, 2000).

An analysis of the practice of the trade unions can give us some idea as to why it is difficult to uphold the rights and interests of migrant workers in China. In China, workers have the right to set up or join trade unions irrespective of their employment, gender, nationality, race, religious belief or education. However, a large number of migrant workers work in small groups and are scattered in various private enterprises (e.g. those in service industries). These people may not belong to any union or are unaware of the existence of unions. Furthermore, in foreign-invested enterprises, the foreign investors may refuse to allow the establishment of a trade union within the enterprise. Once a trade union is set up, the trade union representative usually becomes the Chinese manager in the enterprise (Tan, 2000). This puts the representative in a rather ambiguous position. He or she not only has the responsibility to safeguard the interests of the labourers and the enterprises, he or she must also uphold the interests of the local government. As mentioned above, since local governments are more interested in pursuing economic growth, the trade union/Chinese manager in the enterprise is more inclined to maintain a good relationship with the foreign investor and sacrifice the rights and interests of the migrant workers (Tan, 2000). Thus, the tripartite relationships among local governments, foreign investors and the migrant workers, as stipulated in the International Labour Organisation, cannot be effectively exercised. As Zhu (1995) puts it, in the Chinese world of industrial relations, unions have become largely a part of the bureaucratic control system, unable to distinguish themselves from management.

Implications for social welfare

Although a number of people have urged the central government to abolish the Hukou system, this does not seem to be possible at the present time. First, unless the rural–urban income differential can be narrowed, it is likely that a large group of rural migrants will continue to drift to the urban cities. Second, the ongoing trend to close down state-owned enterprises will result in a large number of unemployed city dwellers. This situation is expected to further deteriorate because China has now entered the World Trade Organisation (WTO) (Leung, 2003).

However, the situation of migrant workers in urban China cannot be left unattended. The increase in conflicts between employers and migrant workers and between local residents and migrant workers is of growing concern to both central and local governments. Moreover, from a humanitarian point of view, the marginalised living of migrant families ought to be addressed. As mentioned, they are deprived of decent housing, health care, social security, education etc. One way to improve their situation is for the central government to tighten up the implementation of the various regulations and notices that protect the rights of migrant workers. While the central government may have to provide incentives for the implementation of the public and social services, it must also establish tough measures against the lack of implementation. In addition, the establishment of a specific ‘unit’ in local governments is needed for the regulation and implementation of these services for migrant workers. At present, migrant workers are the shared responsibility of several ministries and commissions: the Ministry of Public security, Ministry of Labour and Social Security, Ministry of Agriculture, Ministry of Civil Affairs, Ministry of Education, the State Development and Reform Commission, Ministry of Finance, National Population and Family Planning Commission, etc. This new unit could streamline or specify the procedures of the various regulations and notices, develop specific plans for the implementation of these regulations and notices, and coordinate the above ministries and commissions.

Another way of improving the lives of the marginalised migrant workers is to help them organise themselves into supportive networks. According to S.K. Zhao's analysis, there are various types of informal mutual-aid organisations being run by migrant workers (Zhao, SK, 2000). While some of these organisations are economically oriented (e.g. informal organisations for migrants in the transport and loading industries), some are socially oriented (e.g. fellow villagers, kinsmen or relatives). Apart from satisfying basic material needs, such as food and housing, these organisations facilitate the collection and exchange of information about job opportunities and provide a sense of security for the migrant workers. At present, some of these organisations are loosely organised. It would be a sensible alternative for the central and local governments to guide and strengthen the development of these organisations so that they could help migrant workers settle in urban cities (Zhao, SK, 2000).