Environment and Planning A 2010, volume 42, pages 1457 ^ 1475

doi:10.1068/a42381

Migrant workers in the urban labour market of Shenzhen, China Mark Y Wang, Jiaping Wu

Department of Resource Management and Geography, The University of Melbourne, 221 Bouverie Street, Carlton, VIC 3010, Australia; e-mail: [email protected], [email protected] Received 21 September 2009; in revised form 26 October 2009

Abstract. In view of recent changes in both migration patterns and urban labour market conditions, the permanent settlement migration paradigm has become insufficient to guide our understanding of rural-to-urban migrant workers in China's market transition. Using Shenzhen City as a case study we consider migrant workers as enabling agents who interact with the urban labour market. We therefore examine the ways that migrant workers' social capital accumulation, migration experience, and job mobility influence urban labour market dynamics. In painting a picture of migrant workers we complement existing institutional and labour market analyses. As we demonstrate, by frequently changing jobs and destination cities, migrant workers accumulate social and human capital so as to improve their opportunities in the urban labour market. This is evidenced by migrants' improved occupational positions and increased wage earnings.

1 Introduction Following economic reforms since the 1980s, China has seen a unique pattern of rural-to-urban migration. This uniqueness is due not only to the large scale of such migrationöbeing described as the largest population movement in human history (Yardley, 2004)öbut also to the hukou policy, (1) which has functioned as a `passport' for controlling peasants' migration to the city (Alexander and Chan, 2004; Chan, 1994; Chan and Buckingham, 2008; Chan and Zhang, 1999; Solinger, 1999; Wang, 2002). The hukou policy significantly impacts migrant workers' social and economic integration, by constructing their `otherness' within urban communities and labour markets (Bian, 1997; Zhang, 2001). Although the Chinese government has gradually relaxed its hukou restriction over the years (Chan and Buckingham, 2008; Wang, 2005), ongoing urban labour market segmentation, based on workers' hukou status, is still widely reported (Fan, 2008; Lu and Song, 2006). Over the years Chinese migrant workers have therefore been excluded from formal urban labour markets (Fan, 2002; 2008; Wang, 2005) and from ``the logics of urban development'' (Pan, 2007, page 303). Being outsiders, they face a range of social, political, and economic disadvantages. Therefore, migrant workers' home villages, rather than the city, provide the basis of their economic and social security (Fan and Wang, 2008). Some scholars have concluded that migrants' desire to settle in cities is not as strong as first expected (Cai, 2000; Fan and Wang, 2008; Hare, 1999). Others assume that migrants do seek to move permanently to the city, but are excluded through the `denial' of permanent residency rights (Chan and Buckingham, 2008; Rozelle et al, 1999; Solinger, 1999) and through various forms of labour-market discrimination (Knight and Song, 1999; Meng and Zhang, 2001). (1) Introduced in 1958, the hukou system divides all Chinese into either agricultural (rural) or nonagricultural (urban) populations, thereby tying every Chinese citizen to a particular place to work and reside. The hukou system has become the Chinese government's primary tool to control rural ^ urban migration. Although the system has been relaxed, with peasants permitted to work in urban centres since the mid-1980s, it is still regarded as unfair and inhumane (see Solinger, 1999; Wang, 2002).

1458

M Y Wang, J Wu

This conceptualisation of migrant workers has provided valuable insights into the social and economic status and labour-market involvement of migrant workers in comparison with their urban counterparts. The latter assumption has been confirmed, and may still hold as various benefits continue to be attached to the urban hukou.(2) However, this approach ignores migrants' human agency; consequently there are very few studies considering migrants' job mobility and migration experience, and how these shape their prospects in China's urban labour market. In this paper we acknowledge the importance of institutional factors but look beyond them and take the migrant worker as an enabling human agent, as this aspect has been ignored by previous research. We examine the dynamic interaction between the growth of labour migration and the market transition, and argue that the growing flow of migrant workers is not merely an important economic element of the market transition, but also that the migrant worker becomes an enabling human agent. We draw on a recent study involving migration-history and job-history surveys of 819 migrant workers, and interviews with 82 of them, in Shenzhen city. We first argue that the institutional view of migration is essential but inadequate. The growth of migrant workers is a dynamic process, reflected in the changing scope and scale of migrant worker flows as well as in the development of the urban labour market. Second, we argue that mobility patterns of migrant workers involve a process of accumulating social and human capital, as is reflected in the labour market. We examine both how the market transition has influenced migrant workers and how migrant workers have adapted in response to market reforms, in particular through learning by doing. This requires a research methodology moving away from the binary model of migration that sees migrant workers as either rural or urban population, towards linking migration patterns with larger social and economic structures and examining the dynamic relations between them (Massey, 1990). We begin with a brief background of rural-to-urban migration development and a review of the market transition as it pertains to conceptualising migrant workers in urban China. This leads to a discussion of various approaches to studying migrant workers and the latest developments in relation to migrant workers in urban China. We go on to describe our particular survey and discuss the changing characteristics of migrant workers in Shenzhen City, including migration experiences and human capital accumulation such as education attainment, job mobility, and training. A multivariate analysis is employed to estimate the effect of human capital accumulation on wage gains of the migrant workers. This exercise is intended to help account for the wage growth experienced by migrant workers during the labour-market transition in urban China, the understanding of which so far focuses only on institutional and labour-market discrimination. 2 Migrant workers in China: institutions, labour-market development, and human capital accumulation Rural-to-urban migration in China developed as a result of economic reform in both rural origins and urban destinations as well as policy changes between themöparticularly the hukou policy. There is a wealth of literature considering each of these factors, especially (and largely limited to) the changing hukou policy and its implications for migration and the urban labour market. (2) Many

financial benefits and subsidies for the local urban hukou population have been removed in recent years, but social medical insurance is still available to the urban hukou population (and not to migrant workers). In many large cities certain types of jobs are reserved for local urban hukou workers, and are inaccessible to migrant workers.

Migrant workers in the urban labour market of Shenzhen, China

1459

Starting in the early 1980s, China's rural reforms released massive `surplus' labour from farming activities in rural areas: according to various estimates, between 50 and 150 million rural workers were reported as surplus to the agricultural sector in that decade (Liu et al, 2003; Wang and Ding, 2006). However, the hukou policy maintained strict control over rural migration to cities: the migrant workers were seen as surplus to the rural sector, but not welcomed to join the urban economy. Fortunately, the boom of township and village enterprises (TVEs) in the 1980s created employment for many rural surplus labourers. For example, between 1984 and 1988 TVEs annually created 12.6 million new jobs and absorbed 93% of rural migrant labourers, thereby preventing surplus labour from flooding into the cities (Liu et al, 2003; Wang, 1998). By the end of the 1980s most peasant workers moved to local urban centres for employment as the TVEs' declining capacity to absorb rural workers pushed massive flows of migrant labour into the cities (Chang, 1998). As these huge population movements contradicted the viewpoint of central control, migrant workers were labelled as a `blind flow' (mang liu) (Lei, 1996). They were excluded, isolated, and suffered discrimination under urban-based policies and within urban communities (Li, 2006; Solinger, 1999), being characterised as responsible for increasing traffic congestion (Lei, 1996) and urban crimes (Huang, 2007; Wang et al, 2007). However, as urban demand for migrant workers increased with the market transition in the 1990s, more migrant workers came to cities and stayed for longer periods of time (Gu et al, 2007). The number of these mang liu migrants reached 147.4 million in 2005, suggesting that they had come to constitute a significant force in the urban labour market. However, they were not real citizens of the city (Smart and Smart, 2001; Solinger, 1999) and lived with the constant threat of being captured and deported (Siu, 2007): thus they could ``be used or abandoned at any time'' (Lu et al, 2007, page 21). The growth of migrant workers has also been driven by a number of other push and pull factors, above all the growing disparities of economic development between rural and urban areas and between coastal and inland regions (Cai et al, 2002; Hare, 1999). Economic theories, especially those drawing on the labour-market analyses, have become increasingly influential on understandings of migrant workers in China. Placing migrant workers in the rural ^ urban `dual structure' in China, a vast literature has documented the political, social, and economic discriminations faced by migrant workers, including their impacts on gendered labour-market outcomes that specifically discriminate against female migrant workers (Fan, 2004; Meng and Zhang, 2001; Xu et al, 2006). The development of migration and migrant workers' labour-market experiences are seen to ``reflect their outsider status in the city'' (Fan, 2002, page 109). In recent years this rural ^ urban migration has increased in both volume and complexity. Despite the fact that institutions restricted their movement and employment, over the past three decades, migrant workers have gradually established themselves in urban labour markets, and more broadly in the market transition, in a number of ways. The most important of these is the `circular' feature of labour migration, whereby migrant workers repeatedly travel back and forth between their rural origins and destination cities. In 2006 the average time migrant workers stayed in cities to engage in the urban labour market was 268 days (Zhao et al, 2006). Importantly, their movement is often composed of several nodes, or stopping places: while circular, recurrent, or seasonal movements are common practices, workers' rural homes remain the ``center of gravity'' (Roseman, 1971). This challenges the traditional conception of migration as an ``essentially one-way and relatively permanent'' shift of residential location. Rather, as older migrant workers withdraw from cities or permanently return home to rural areas, new migrant workers replace them, effectively sustaining the market transition. For young peasants, migrating to work in the city for a period of time has become part

1460

M Y Wang, J Wu

of their life course, and is described as ``flowing into the city when young and returning to the rural when getting old'' (nian qin jin chengshi, lao lai hui nongcun). While the significance of such circulatory moves has long been recognised, early literature focused on circular migrants driven by noneconomic motives (Chapman and Prothero, 1983; Zelinsky, 1971), and emerging studies tend to refer to international migrants (Agunias, 2006; Hugo, 2003). The movement of migrant workers in China has been mainly driven by economic opportunities in the city (Fan, 2002). Cultural factors also support circulatory movement, as it is important for migrant workers to fulfil their social obligations and maintain values and ties in their rural home areas. Our survey shows that migrant workers frequently change jobs and destinations during their periods spent working in the cities, in order to improve their job opportunities and maximise earnings. The high frequency of such moves and job changes is not accounted for by key concepts like `usual residence' and `administrative unit' in migration studies. While it is very important for migrant workers to improve their social and economic status through climbing the occupational and income ladders, identifying migrant workers' moves and incomes in cities is difficult. Many migrants never register their residency in the cities, preferring to `hang around' here and there within a city or between cities (Shen, 2002). This frustrates attempts to compare the incomes of migrant workers and native urban locals. From this perspective understanding migration processes themselves has become as significant as understanding differences between migrants and native urban locals. This shift of research focus challenges our empirical knowledge both of migration and of data collection for migration studies, which tend to be based on the binary models of migration ``anchored in a permanent settlement migration paradigm'' (Hugo, 2003, page 1). In this paper migration flows are characterised as an element of market transition, and conceptualised at two levels. At the macro level the flow of migrant workers provides a stable `conveyer belt' of labour: a structure of market transition in China. Older migrant workers step off, while younger ones step on `the belt', which continually changes in response to market transition conditions. For the urban labour market the growing flow of migrant workers provides a reliable labour source, thereby enabling employers an approach of frequently recruiting and changing the workforce (changzhao changhuan), avoiding payments for labour seniority and reducing labour costs. Following the ethnographic approach of a life course perspective that traces the development of recurrent mobility (McHugh et al, 1995); at a micro level migrant workers are perceived as `enabling agents' who strengthen the conveyer belt and improve their market outcomes by accumulating human and social capital. This means that migrant workers can make their decisions and choices about migration, rather than being an inactive element as in the old institutional structure. This perspective allows for the contextualisation of migrant workers within the market transition and constructs `bilateral' changes as a result of interaction between migrant workers and the market transition over time. As Laoire noted, the process of a migrant's movement ``cannot be understood without taking into account all aspects of the personal and environmental context in which it takes place, including the relevant dimensions of power and inequality that help to shape it'' (2000, page 235). This normalisation of migration involves not only an understanding of migrants in the structural role of migration flow in the market transition, but also appreciating ``the role of human agency'' (page 241). From this view, migrant workers are active and capable agents, whose capacity ``can have important consequences not only for workers themselves but for other actors also, such as firms, states, families and communities'' (Castree, 2007, page 855).

Migrant workers in the urban labour market of Shenzhen, China

1461

This contradicts existing economic literature, which centres on capital and casts migrant labour as merely a kind of commodity. Using this alternative understanding of migrants, in this study we analyse mobility and migration experiences, and assess the labour market outcomes of migrant workers. One critical factor for workers' migration and their job changes is the acquisition of human capital through learning by doing (Heckman, 2000). This includes migrants learning from their own experience as well as from others in the urban labour market. As Heckman et al (1998) have noted, postschool learning in firms accounts for almost half of all skill formation in modern economies. Labour market experience and lifelong learning are increasingly important for personal fulfilment and wage growth (Altonji and Williams, 1998; Gladden and Taber, 2000; Williams, 1991). The effect of learning by doing, which in this case corresponds to migration experience, is proportional to the stock of human capital, which is then reflected in the labour market. Oettinger (1996) observed that, even if no wage gap exists at the time of labour force entry, one will develop as workers accumulate experience in the labour force. Wage growth is the expected outcome of job changes, due to the assumed improvement in job status and increased labour productivity (Jovanovic, 1979; Mincer and Jovanovic, 1981). In a highly influential paper Topel and Ward (1992) show that, for young white men, job mobility produces a large positive influence on wage growth. They attribute over one third of men's average wage growth during their first decade of market activity to job `shopping'. Similarly, comparing the wage gains of US workers who consistently work for the same employer with those who change employers, Gottschalk (2001) finds that mean wage growth between jobs is considerably greater than wage growth while working for the same employer. The only existing research on the mobility of migrant workers in urban China seems to support this view: that is, that migrants' mobility improves job status and raises their wage earnings (Knight and Yueh, 2003). Knight and Yueh (2003) have contributed the only existing empirical study of migrant workers' job mobility and migrant experience in China's urban labour market. However, their study was based on by-product information from the Household Survey of Urban Hukou Residents in 1999. Such data are not only ``not designed specifically to examine labor mobility'' (Knight and Yueh, 2003, page 10) but are also outdated, considering the urban labour market transition and changes in labour migrant flows over the past decade. Also, as the authors noted, the migrant workers they investigated were more established ö so the study is unlikely to represent the situation of rural ^ urban migrants as a whole. By contrast our study focuses directly on migrant workers' strategies for migrating and engaging with the labour market, and how this influences their wage earnings. More specifically we use Shenzhen City as a case study to explore migrant workers' differential migration experience, human capital accumulation, and urban labour-market outcomes. 3 The Shenzhen survey The most important part of China's market transition has been the designation of Special Economic Zones (SEZs) as windows to the outside world, where new policy on market reform is experimented. Shenzhen, the first and most successful SEZ in China, has been the `capital city' of `red capitalism' in China's market transition. Located immediately north of Hong Kong, Shenzhen has used its geographical advantage, making Hong Kong the largest source of foreign investment over the years (Chen, 1986; Wang and Meng, 2004). As shown in figure 1, Shenzhen is composed of SEZ parts (four urban districts: Nanshan, Futian, Luohu, and Yantian) and non-SEZ parts (two suburban districts: Bao'an and Longgang). Until the mid-1990s, to prevent migrants

1462

M Y Wang, J Wu

Figure 1. Location and structure of Shenzhen City.

flooding into the SEZ, entrance into and exit from the SEZ was strictly controlled through border points: to cross, individuals needed either local Shenzhen identity cards or special permission. Upon realising the stable labour market in Shenzhen in the mid-1990s, the municipal government removed border-check points for entry to and exit from the SEZ. Migrant workers now have unrestricted access to the Shenzhen SEZ. Shenzhen's development has been largely due to the combination of foreign investment (especially investment from Hong Kong) with the cheap labour supply of migrant workers. Investment has concentrated on the labour-intensive sector dominated by san lai yi bu enterprises (that is, export processing activities like processing materials on clients' demand, assembling parts for clients, producing products according to clients' samples, and receiving compensation trade in exporting final products). In 1983 82% of total overseas investments were invested in san lai yi bu sector. The growth of san lai yi bu investment has rapidly stimulated an influx of migrant workers into the city (Liu, 2007). The total population of Shenzhen leapt from fewer than 20 000 in 1978 to 3.45 million in 1996 and to 8.46 million in 2006; 76.7% of these people did not possess a Shenzhen hukou (Shenzhen Bureau of Statistics, 2008). The hukou policy has therefore become less relevant to the understanding of migration and labour market transition; the economic development of the city now relies much on bill-led industries, and the consequent seasonal and temporary labour-market dynamics. For example, table 1 shows the age structure of labour demand in Shenzhen in 2008, when 76% of jobs required workers of under 35 years of age. Table 1. Age structure of labour demand in Shenzhen in 2008 (source: data supplied by the Shenzhen Bureau of Labour and Social Security, SBLSS, 2008). Age required < 24 25 ± 34 35 ± 44 45 and above No age requirement Total

Labour demand (%) 47.2 28.8 8.1 2.5 13.3 100.0

Migrant workers in the urban labour market of Shenzhen, China

1463

Shenzhen is an ideal site for analysing migration processes and the market transition: industrial development has created new jobs opportunities particularly suited to young migrant workers, and the growth of young migrant workers has further facilitated industrial development. While the examination of a specific case may not support a major reformulation of general migration theories, Shenzhen's development could be considered a prototype for the market and institutional transition in the whole of China over the past three decades. Many new reform policies have first been introduced and experimented in Shenzhen, before being implemented nationwide. For example, the `contractual employment' system was introduced in the early 1980s, and is now widely employed in China. Similarly, san lai yi bu enterprises have spread to the whole nation. This study is based on data from a migration-history and job-history questionnaire survey conducted in Shenzhen City during December 2008 ^ January 2009. As discussed earlier, this approach is different from existing research which has already examined the institutional factors of hukou policy and market development. Instead, this survey focused on underresearched aspects of migrant workers, especially their mobility patterns and migration experiences. This approach involved tracing migrants' paths through the time and space of their migration experiences, linking changes in the life courses of individual migrants to the broader environmental context. The survey was complemented by ethnographic methods such as focus group interviews, migration-history interviews, personal observation through absorption in local life, and discourse analysis. These methods were intended to build up the development of migration and migration histories of each individual, including information on their paths through education, migration, employment, training, and career progression. Our migration questionnaire survey consisted of two parts. Initially, a subset of units (1000) was generated from a database provided by the Shenzhen Bureau of Labour and Society Security (SBLSS), which contained information about 43 573 migrant workers, collected from 1814 firms in the city. We selected the 1000 workers using systematic random sampling techniques öthe units were selected at an equal interval space listed in the SBLSS dataset. This was to guarantee the survey to be random and representative. A migration and job mobility survey was then conducted. The selected migrant workers were approached by the mixed modes of telephone, visits, and mail for questionnaire and interview. As a result 776 valid migration and job mobility questionnaires were produced. Intensive effort was made to communicate with managers of enterprises and organisations and to visit the participating migrant workers with the help of local assistants for distributing the questionnaires and managing the quality of the survey. The official database provided by the SBLSS collected for the development of a social security policy in the city contained little information about migration and the job mobility of the migrant workers. Part of the data was collected through reports of industrial organisations. The migrant workers who engage in domestic services such as nannies and housekeepers are severely underrepresented because they do not work in large industries. Our interview confirmed that many such domestic service workers are introduced by friends and relatives. These jobs are not generally recorded. To overcome such underrepresentation of this group in the data, and for broad comparison purposes, efforts were made to reach such people. Using the same questionnaire, domestic and community workers in two large residential blocks in the city (one within the SEZ and one in the Baoan district) were surveyed, producing forty-three valid samples. As a result, a total of 819 surveys were collected, covering state-owned enterprises, collective, private, and foreign funded enterprises, totalling thirty-one enterprises

1464

M Y Wang, J Wu

and organisations across the city. (3) Each migration decision reflects how the migrant's individual choices and the market transition intersect and interrelate in different ways, and highlights the role of human agency in negotiating a path through them. The data collected are exceptionally comprehensive, including demographic features, wage levels, occupation, and employment status for each move of the migrant workers from the date of leaving their rural home town until the time of the interview. This enables us to assess the impact of job mobility and levels of migration experience on the wage earnings of migrant workers in the city. The data also include detailed information on how migrant workers obtained each of their jobs, any job training, their social networks, as well as their career progression and future plans, allowing us to investigate the alternative channels by which human capital and market opportunities influence migrants' wage growth. This analysis provides insights about changing migration flows and how the interrelation of two processesönamely, labour migration and the market transitionö has positioned migrant workers in the urban labour market. We pay particular attention to the impact of social and human capital accumulation, including learning by doing and `learning from others', and job mobility on migrants' labour-market outcomes. Table 2 summarises the gender and age structure of the surveyed migrant workers in Shenzhen. The average age of surveyed migrants is 25.3 years and over 94% of surveyed migrant workers were under 35 years, echoing an SBLSS survey (25.9 and 91.8%, respectively) of 68 400 migrant workers in 2006 (documents supplied by SBLSS, in 2008). Table 2. Demographic profile of migrant workers in Shenzhen. Own survey (%)

SBLSS survey a (%)

Gender Male Female

47.8 52.2

45.9 56.1

Age 4 18 19 ± 26 27 ± 35 >36

4.9 61.3 27.5 5.9

6.7 61.4 22.7 9.2

a Data

supplied by the Shenzhen Bureau of Labour and Social Security (in 2008).

In terms of industrial structure, the surveyed migrant workers were disproportionately distributed in the manufacturing sector (see table 3); this proportion far exceeds the national average of 24.4% in this industry (Xinnongmen, 2008), thereby reflecting the industrial structure of san lai yi bu in the city. Shenzhen and other SEZs and cities in the Pearl River Delta have become well known as the world's workshop. Domestic and community services workers accounted for 6.7% of the total, while the combined share of other service industries öincluding retail and wholesale and hotel and cateringöwas just 6.1%. Migrants working on construction sites and in urban infrastructure development consisted of merely 5.8% of the total. Geographically, the surveyed migrant workers came from all over the country apart from Tibet. Over three quarters of our survey migrant workers had come from provinces in central China (Hunan, Henan, Hubei, and Jiangxi), the neighbouring province Guangxi, and provinces of Sichuan and Jiangxi. These provinces are also the main sources of migrant workers in the country (Liu et al, 2003). Migrant workers in Shenzhen from other parts of Guangdong province consisted of 7.2% of the respondents. (3) Owing

to space constraints, the analysis in this paper is based mainly on the survey data.

Migrant workers in the urban labour market of Shenzhen, China

1465

Table 3. Industry of employment and geographic origins of migrant workers in Shenzhen. Industry of current employment

Percentage

Geographic origin

Percentage

Manufacturing Domestic and community services Construction and urban infrastructure Retail and wholesale Hotel and restaurant Agriculture Other

75.4 6.7 5.8

Hunan Henan Hubei

17.1 12.3 12.2

Guangxi Sichuan Guangdong Other

11.2 10.9 7.2 29.1

3.2 2.9 0.1 5.9

4 Migration experience and educational attainment The average migration experience of the total samples is 5.1 years. Among the surveyed migrant workers, about 17.7% of the surveyed migrants migrated to cities before 2000. However, only 11.6% of them came to Shenzhen during the same period (table 4). Also indicating the respondents' migration experience, over 10% of the sample had worked in at least three cities before they came to Shenzhen. Over half of the surveyed migrants arrived in Shenzhen during the period between 2005 and 2008. During the same period of time only one third of the migrant workers left their rural origins in the samples. The reasons for moving from city to city included `for better earnings including welfare' (51.9% of respondents), and `to experience new places including different working environments' (27.8% of respondents) (Xinnongmen, 2008). Table 4. Migration experience and job mobility of migrant workers in Shenzhen. Years of migrating away from rural origins to cities (%) 1981 ± 94 1995 ± 99 2000 ± 04 2005 ± 08 Total

Years of migrating to Shenzhen City (%)

4.4 13.3 46.9 35.4

3.4 8.2 36.8 51.6

100.0

100.0

Number of jobs being taken by migrant workers (%) 1 2 3 54 Total

27.2 37.5 20.9 14.4 100.0

In terms of job mobility, the surveyed migrant workers had experienced an average of 2.6 jobs, and one third had experienced at least three jobs at the time of survey. The average duration of each job was 2.8 years. This suggests that workers' migratory urban life had already developed during the short period they had been working in the city, and had involved strategies of adapting to urban settings, integrating with urban locals, and educating their childrenöall of which differs greatly from their life in the rural area and from those of urban locals. Migrant workers experience different places and working in various jobs:

1466

M Y Wang, J Wu

``We meet various kinds of people from different parts of the country. We learn from each other and help each other, like exchanging job and income information'' (personal interview, 2008). The importance of educational attainment as a crucial determinant of migrant income points to human capital as the key to a successful career in a modern economy. Table 5 presents the changes in migrants' ages, education attainment, and monthly wage levels over the past three decades. The average age of a migrant worker entering the urban labour market has fallen steadily from their mid-thirties in the 1980s and early 1990s, to their early twenties in recent years. Such a general transitional trend of demographic characteristics is similar to the conclusions reached by Xinnongmen (2008) and Zheng and Huang-Li (2006). It should be noted that not all the first jobs of the surveyed migrant workers were necessary obtained in Shenzhen. The new generation of migrant workers is different from those who were released from the agricultural sector in the early 1980s. The new generation of migrants were born in rural areas, but most never had any agricultural knowledge or experience. Instead, they entered the urban labour market immediately after finishing junior or senior high school. Compared with the older generation, they also have a higher educational attainment, with different migration prospects and urban labour-market opportunities. The survey data reveal that the proportion of migrants who had completed senior high and above increased steadily from 25.8% in the 1980s to 34.6% in the past few years, while the proportion of those who had only finished primary school or were illiterate fell. However, data regarding the 1980s and 1990s represent a relatively small proportion of migrant workers in the survey, and span a long time period. This may also be because most of the less-educated migrants had already withdrawn from the urban labour market. Table 5. Migrants' age, educational attainment, and monthly mean income by time period.

Average age at first job (years) Monthly wage at first job (yuan) Educational attainment Senior high and above (%) Junior high (%) Primary and illiterate (%)

1981 ± 89

1990 ± 94

1995 ± 99

2000 ± 04

2005 ± 08

35.1 786.7

33.2 1120.6

27.9 1027.1

24.8 1224.3

23.8 1388.0

25.8 60.4 13.8

27.3 65.2 7.5

31.2 63.1 5.7

33.3 64.8 1.8

34.6 61.8 3.5

We also examined the possible relationships between migrant workers' educational attainment and their migration experience, including the frequency of their job mobility. For both migration experience (years of migration) and job change frequency (jobs per year), the coefficients are ÿ0:21 and ÿ0:12; respectively, but are not statistically significant. This suggests that migration experience and job mobility are not closely related to migrants' education levels. Does a greater educational attainment and increased migration experience broaden migrant workers' means of job obtainment? Figure 2 presents the means of migrant worker's job obtainments. As figure 2(a) illustrates, overall, the main means of job obtainment did not change much over time, while the magnitude of relying on each means did change with an increase in migration experience. The proportion of jobs that migrants obtained through their `own job search' has increased slightly from 34.5% before 1995 to 38% during the period between 2006 and 2008. Job agents are relatively new to Chinese labour markets and have played some role in migrant workers' job placements. Direct recruitment by employers as a proportion of migrant workers' job placements

Migrant workers in the urban labour market of Shenzhen, China

1467

Percentage

40

30

20

10

0

(a)

Pre-1995

1996 ± 2000

2001 ± 05

2006 ± 08

Self-find Relatives and friends 40

Percentage

Job advertisement 30

Employers recruitment Job agents

20

Government

10

Others 0

(b)

First job

Second job

Third job

Figure 2. Changing means of migrant workers' job obtainment (a) by time, and (b) by job mobility.

declined in the second half of the 1990s but increased in the new century. This can be explained by the fact that Shenzhen and the rest of the Pearl River Delta experienced a significant shortage of migrant workers in the urban labour markets from 2000 to 2006 (Liu, 2006). Efforts were made by Shenzhen employers personally to recruit workers from the urban labour market, or even directly from major migrant-sending areas. Confirming the common findings from many other studies, our data indicate that relying on relatives and friends has been a consistently dominant avenue of job obtainment for migrant workers over time. Interestingly, it increased from 31.6% as migrants' means to obtain their first job to 42.4% to obtain their third job [figure 2(b)]. Greater migration experience therefore does not diminish migrants' ongoing reliance on relatives and friends in the labour market. In fact, all other means of job obtainment decreased between migrants' obtainment of their first job to their third job. This suggests that maintaining connections is critical for migrant workers in locating new job opportunities in the urban labour market. As an important avenue of job

1468

M Y Wang, J Wu

placement, relatives and friends may reserve desirable jobs for people within their network. They are also reliable sources as to the quality of workplaces and income information, thereby influencing migrant workers' job choices. Means of job obtainment is not influenced by migration experience, but migrant workers with different levels of education tend to locate jobs through different channels. Table 6 shows how migrant workers with higher education levels are more proactive when job hunting, while those with lower education levels are more passive. The more educated the migrant worker, the more likely he or she is to be self-recommended, or to secure a job through their own job search, through the Internet, or at special job fairs. More highly educated migrants know how to use the Internet and sell themselves in the labour market, while migrant workers with limited education find jobs through friends or relatives, are recruited by employers, or have their employment arranged by the governments or agents. Table 6. Migrants' means of job obtainment by education level (%) (source: Xinnongmen, 2008). Education level

Passive a (%)

Proactivea (%)

Relative/friends (%)

Through others (%)

Primary or lower Junior high Senior high College diploma University

35.0 23.7 22.9 19.6 11.9

13.6 15.1 29.2 42.1 50.9

38.0 50.0 34.7 24.9 20.1

13.4 11.2 13.2 13.4 17.1

a `Proactive'

job hunting refers to situations where job information is obtained by migrant workers themselves or mainly through their own initiative during the job hunting process; `passive' refers to circumstances where job information is mainly given or passed on by others. The former includes jobs obtained through: (1) self-recommendation (migrants approach potential employers asking about available jobs); (2) searching for job opportunities on the Internet; (3) attending recruitment fairs (where curriculum vitae and interview are normally required); or (4) self employment. Passive means to obtain jobs include: (1) recruitment by employers; (2) job arranged through community service or by the government; (3) through job advertisement.

5 Mobility, human capital accumulation, and wage earnings To further understand labour-market processes involving migrant workers, we analysed job and occupation mobility and how these affect migrant workers' wage earnings. 5.1 Job change, occupational mobility, and income change

Using our survey data we first plotted migrants' wage changes over time and their occupational obtainments with job mobility (figure 3). Figure 3(a) compares migrant workers' `current' monthly wages each year with their earnings when they first entered the urban labour market (that is, each year shows mean wages of migrants in their first urban jobs, against all other migrants' wages). Reflecting the contribution of migration experience to wage levels, the mean current monthly wage of workers with a longer migration history in the urban labour market is consistently higher than that of recently arrived migrant workers. However, the survey data also reveal an interesting contradiction: the current monthly mean wage level of migrant workers who started working in cities between 1995 and 1997 is slightly higher than that of recently arrived migrant workers (as would be expected), but is lower than that of workers who started working in cities between 1998 and 2001. More research is required to understand why this older generation of migrant workers currently earns less than those who began working in the city between 1998 and 2001. It may be that, although most of these

Migrant workers in the urban labour market of Shenzhen, China

1469

2000

Monthly mean wage (yuan)

Current 1600

First job

1200

800

400

0 (a)

1996

1998

2000

Year

2002

2004

2006

2008

Manufacturing worker Other manual worker

Percentage

40

Intermediate sale and service person Skilled craft and trades worker

30

Semiskilled manual worker 20

Manufacturing supervisor Crafts and trades supervisor

10 0 (b)

Skilled sales and service worker First job

Second job

Third job

Technician

Figure 3. Changing mean monthly income of migrant workers (a) by migration experience, and (b) by job mobility.

older generation migrant workers in Shenzhen reached their income ceiling several years ago (their salary has not increased since their early 30s), they continue living in Shenzhen for family reasons, such as their children's education and/or their spouses' employment. In either case workers' acceptance of a plateaued income (by staying in Shenzhen) may be compensated by their spouses' income (double income) and/or the quality of their children's urban schooling. However, this is mere speculation: further research is clearly needed on this point. Figure 3(b) presents migrant workers' occupational attainment with job changes. The occupational attainment data require careful interpretation, as migrant workers' occupational changes are closely related to industrial restructuring and structural characteristics of the local labour market. Industrial structure, as one of the important organising principles of the labour market, varies considerably over regions and cities. In this case city, upon first entering the urban labour market, 85% of migrant workers were engaged in three sectors: manufacturing, other manual sectors, and intermediate services. After two job changes, 7% of these manufacturing workers had been promoted to supervisor positions of the manufacturing assembly lines. The proportion of migrant workers engaged in `other manual work' and intermediate service work had

1470

M Y Wang, J Wu

declined from 32.1% and 15.4%, respectively, in their first jobs to 12.75% and 9.5% in their third jobs. This suggests that many migrant workers moved to the city with no specific job expectation and accepted whatever menial jobs were available. Over time in the urban labour market (that is, as their experience grew), migrants gradually established themselves in more skilled employment. This analysis is confirmed by the fact that, after two job changes, 10% of migrant workers became skilled craft and trade workers, compared with only 5% at their first job. Although small in proportion, migrant workers also become technicians (2.7%), skilled sale and service workers (2.4%), and managers (0.3%). Therefore, after several job changes many migrant workers are able to climb the occupation ladderöeven to rungs considered reserved. This indicates that occupational attainment is not entirely blocked to `outsiders'. 5.2 Human capital accumulation and wage incomes

To quantify the relationships between social and human capital accumulation and migrants' position in the urban labour market, we used a multivariate analysis to estimate the effects of job mobility and migration experience on wage earnings. Monthly mean income (MinC) at the time of the survey was selected as the dependent variable. Our model specification is based on the theoretical and empirical discussions in section 2, with an awareness of potential problems in analysing data on the earnings impact of human and social capital accumulation that may arise due to a wide range of exogenous variables. Bearing these in mind, the changing MinC is assumed to be a function of three factor sets: (1) migrants' human capital, efforts, capabilities, and social capital accumulated through their engagement with the urban labour market; (2) structural factors; and (3) individual characteristics of the migrant worker. As discussed in section 2, structural factors are not analysed here önot because their impact (and particularly the hukou policy) on migrant workers' earnings is less significant, but because considerable knowledge about these connections already exists (eg Fan, 2002; Lu and Song, 2006). Therefore, our explanatory variables refer to the accumulation of human and social capital, and individual factors that could facilitate social capital accumulation during the migration process in the city. These include: (1) educational attainment (EduA), classifying individuals into four categories based on their highest level of school completion (primary, junior high, senior high, and college and above); (2) migration experience (MexP), which refers to number of years since leaving the rural home; and (3) job training (JobT), representing days of training during the entire period of work in the urban labour market. It was expected that individuals with higher educational attainment, more migration experience, and more job training would have a higher wage income. The effect of job mobility on income was expected to be ambiguous: while job mobility can help improve job matches and raise income, frequent job changes may also undermine human and social capital accumulation (Knight and Yueh, 2003). We use two variables to capture migrants' job mobility in the urban labour market: number of jobs undertaken (JobN) and average job duration (JobD). The impact of migrants' individual characteristics on earnings is captured by age of migrant worker at survey time and dummy variables of gender where male is denoted 1 and female is 0, marital status where 1 represents married migrant workers while 0 is never married or divorced, and geographic place of origin where 1 denotes intraprovincial migrant worker while 0 represents interprovincial migrant to Shenzhen City. Pearson correlation and a scatter plot matrix of human and social capital accumulation variables were initially conducted to detect possible presence of heteroscedasticity, effects of nonlinearity, and for model consideration. These analyses indicate that

Migrant workers in the urban labour market of Shenzhen, China

1471

Table 7. Regression results. Variable

Standardised coefficient

Human and social capital accumulation JobT 0.054* EduA 0.237 JobN 0.085* MexP 0.304** JobD ÿ0.095** Individual factors Geographic place of origin Marital status (married) Gender (male) Age R2 (model 1) R2 (model 2)

ÿ0.004* ÿ0.075 0.110* 0.008* 0.295 0.319

**significant at 95%; *significant at 90%.

reasonable linear fits can be obtained by regressing MinC against the selected variables individually. The null hypothesis of no heteroscedasticity was further tested by a White's test, carried out by running an ordinary least squares regression in SPSS. The null hypothesis cannot be rejected, which suggests that the model specification is accepted. Table 7 presents the results of regression with the control of individual characteristics. According to the estimated results of both models, joint effects of the selected explanatory variables on MinC were significant, but effects of some individual variables were not. R 2 for model 1 (controlling for individual factors) was 0.295 (significance ˆ 000), and 0.319 (significance = 004) for model 2. Our basic hypotheses on the earning impacts of migration experience and job mobility are supported by the regression result with significance level of 95%, suggesting that human capital and social capital accumulation do play a significant role in shaping the wage levels of migrant workers in urban China. The estimated coefficient of the length of migration is positive and significant (0.304), which indicates that, all other variables being constant, one more year of migration experience corresponds to a 41 yuan increase in migrant workers' monthly wages. The significant but negative coefficient for job duration confirms the traditional view that job mobility does improve job matches. This may be a reason why migrant workers change jobs frequently in the urban labour market: migrant workers who have had more jobs in the urban labour market will have higher wages. Job training slightly improves migrant workers' wage levels. In the survey samples, 57.2% of respondents had experienced at least one day of job training during their time in the urban labour market. The average training period was nine days, and 6% reported having experienced over six months' traineeship for accumulation of job skills for specific tasks. Migrants' participation in training was predominantly within their first year of engagement in the urban labour market, and decreased over subsequent years. According to the regression result, an increase of one day of job training improves monthly wages by 1.1 yuan. Education level shows no significant effect, even though the coefficient for education attainment is relatively high. This may be due to a coding limitation in our questionnaire design, which classified education attainment of migrant workers into four categories rather than by total years of schooling.

1472

M Y Wang, J Wu

As regards a migrant's individual characteristics, neither marital status nor age had statistically significant effects on migrant workers' wage levels. Interestingly, being Cantonese was not an advantage for migrant workers in Shenzhen City. There are two possible reasons for this: first, the majority of Shenzhen's population is composed of migrants, among whom Mandarin is widely spoken and accepted; second, intraprovincial migrant workers come mainly from poorer areas in northern and western Guangdong. Concerning the variable of gender, male migrant workers tend to earn more than females. Interestingly, according to the survey data, the monthly average income of female migrant workers' first jobs was the 1227.1 yuan, which was slightly higher than 1221.1 yuan of male migrant workers. However, the monthly average income from female migrant workers had become, by the time of the survey, 1482.8 yuan, which was 9.7% lower than that of male migrant workers. This suggests that women face genderspecific disadvantages in the migration process. These result in loss of potential economic gains due to irregular migration and employment, arising from interruptions such as bearing children and other social responsibilities assigned to women in rural origins. 6 Discussions and conclusion The urban influx of migrant workers has emerged as a specific but very important structural factor of the market transition in China. In essence, labour migration continues to be constrained by centrally planned institutions, especially through the hukou policy, which excludes migrants from opportunities associated with permanent urban residency rights and restricts them to a secondary segment of the urban labour market. The central contribution of this paper is its extension beyond `the permanent settlement migration paradigm', instead viewing migration development rather than institutional restrictions, and focusing in particular on the ways that migrant workers have established themselves in the market transition. This involves taking migrant workers as a normal part of market transition, and understanding their own development patterns within the continuously changing labour market. Their social and human capital accumulation is not only important for migrants' engagement in the urban labour market but also has implications for market transition. By illustrating both changing migrant worker flows and the migratory histories of individual migrant workers, the paper sheds light on several of these important aspects of migration development ö namely, the development of migration patterns and labour market processes. This includes the changing quality of migrant workers, their strategies for engaging in the urban labour market, and the broader implications of those patterns. Our main findings include that rural ^ urban labour-migration flows have changed in response to the urban labour market, which has become an important social and economic structure of the market transition in China. The latest contingent of incoming migrants is made up of younger workers, with higher educational attainment. They potentially face different migration prospects and experience. This shift requires further research attention. Our case study shows that the growth of migrant workers has been related to the market transition, particularly through the development of special forms of industry in the city. The development of the city has been dominated by the san lai yi bu industries, providing the main employment sector for migrant workers in the city, and creating a labour market of `frequent recruitment and changing labour'. This is another important reason why the permanent settlement migration pattern has become less relevant. The rise of economic globalisation is expected to make more economic functions footloose, especially with respect to the choice of lower cost production

Migrant workers in the urban labour market of Shenzhen, China

1473

locations and responses to rapid changes in the global market. This is likely to continue to reshape employment patterns, suggesting that institutional policy reform will not be the only force influencing workers' migration patterns in China. The established patterns of labour migration feature migrants' high mobility during their time in the urban labour markets, including frequent changes in jobs and destination cities. Through these strategies migrant workers have managed to accumulate social and human capital so as to improve their labour-market outcomes. Some migrant workers have even increased their job status and opportunities. Our statistical analysis demonstrates that greater migration experience positively impacts on migrants' wage incomes. Job changes have to some extent also improved migrants' job matches, though they have undermined the efficiency of migrant workers' human capital accumulation. These findings have important implications for policy. Traditionally, policy development and related research have adopted a simplified concept of rural and urban populations in migration studies in China. To a large extent this view has facilitated the isolated treatment of migrant workers as an `other category' of the population in the urban labour market. Our research illustrates that, in addition to hukou policy reforms to improve migrants' engagement with the urban labour market and to create opportunities for migrant workers to establish themselves in cities, further efforts are needed to understand the continuing changes in the migration patterns of migrant workers. Acknowledgements. Mark Wang acknowledges the support of the ARC grant DP0880244, which supported the fieldwork on which this paper is largely based. We both wish to thank our friends in Shenzhen ö Dr Xiying Hu and Ms Xinwen Chi ö for their help in undertaking the fieldwork. And we are, above all, grateful to the migrant workers of Shenzhen for their hospitality: they were generous with their time and knowledge. References Agunias D, 2006,``From a zero-sum to a win ^ win scenario: literature review on circular migration'', Migration Policy Institute, Washington, DC, http://www.migrationpolicy.org/pubs/ CircularMigrationLitReview 9.06 DAgunias.pdf Alexander P, Chan A, 2004, ``Does China have an apartheid pass system?'' Journal of Ethnic and Migration Studies 30 609 ^ 629 Altonji J G, Williams N, 1998, ``The effects of labor market experience, job seniority, and job mobility on wage growth'' Research in Labor Economics 17 233 ^ 276 Bian Y, 1997, ``Bringing strong ties back in: indirect ties, network bridges, and job searches in China'' American Sociological Review 62 366 ^ 385 Cai F, 2000 Zhongguo Liudong Renkou Wenti [China's floating population], Henan renmin chubanshe, Zhengzhou (in Chinese) Cai F, Wang D, Yang D, 2002, ``Regional disparity and economic growth in China: the impact of labor market distortions'' China Economic Review 13 197 ^ 212 Castree N, 2007, ``Labour geography: a work in progress'' International Journal of Urban and Regional Research 31 853 ^ 862 Chan K W, 1994 Cities with Invisible Walls: Reinterpreting Urbanization in Post-1949 China (Oxford University Press, Hong Kong) Chan K W, Buckingham W, 2008, ``Is China abolishing the hukou system?'' The China Quarterly number 195, 582 ^ 606 Chan K W, Zhang L, 1999, ``The hukou system and rural ^ urban migration: processes and changes'' The China Quarterly number 160, 818 ^ 855 Chang S, 1998, ``The floating population: an informal process of urbanisation in China'' Population, Space and Place 2 197 ^ 214. Chapman M, Prothero R M, 1983, ``Themes on circulation in the third world'' International Migration Review 17 597 ^ 632 Chen X, 1986, ``Magic and myth of migration: a case study of a Special Economic Zone in China'' Asia-Pacific Population Journal 2 57 ^ 76 Fan C, 2002, ``The elite, the native, and the outsiders: migration and labour market segmentation in urban China'' Annals of the Association of American Geographers 92 103 ^ 124

1474

M Y Wang, J Wu

Fan C, 2004, ``The state, the migrant labour regime, and maiden workers in China'' Political Geography 23 283 ^ 305 Fan C, 2008 China on the Move: Migration, the State, and the Household (Routledge, London) Fan C,Wang W W, 2008, ``The household as security: strategies of rural ^ urban migrants in China,'' in Migration and Social Protection in China Eds R Smyth, I Nielsen (World Scientific, Hackensack, NJ) pp 205 ^ 243 Gladden T, Taber C, 2000, ``The relationship between wage growth and wage levels'', WP 173, Northwestern University and University of Chicago Joint Center for Poverty Research, http://econpapers.repec.org/paper/wopjopovw/173.htm Gottschalk P, 2001, ``Wage mobility within and between jobs'', WP 486, Boston College Working Papers in Economics, http://escholarship.bc.edu/cgi/viewcontent.cgi?article=1139&context =econ papers Gu S, Zheng L, Yi S, 2007, ``Problems of rural migrant workers and policies in the new period of urbanization'' China Population, Resources and Environment 17 1 ^ 5 Hare D, 1999, `` `Push' versus `pull' factors in migration outflows and returns: determinants of migration status and spell duration among China's rural population'' Journal of Development Studies 35 45 ^ 72 Heckman J J, 2000, ``Policies to foster human capital'' Research in Economics 54 3 ^ 56 Heckman J J, Lochner L, Taber C, 1998, ``explaining rising wage inequality: explorations with a dynamic general equilibrium model of earnings with heterogeneous agents'' Review of Economic Dynamics 1 1 ^ 58 Huang S, 2007, ``On the crimes by the floating population from the perspective of social support network'' Beijing Social Science 2 60 ^ 64 Hugo G, 2003, ``Circular migration: keeping development rolling?'', Migration Information Source, http://www.migrationinformation.org/Feature/display.cfm?ID=129 Jovanovic B, 1979, ``Job matching and the theory of turnover'' Journal of Political Economy 87 972 ^ 990 Knight J, Song L, 1999 The Rural ^ Urban Divide: Economic Disparities and Interactions in China (Oxford University Press, Oxford) Knight J, Yueh L, 2003, ``Job mobility of residents and migrants in urban China'', DP 163, http://www.economics.ox.ac.uk/Research/wp/pdf/paper163.pdf Laoire C N, 2000, ``Conceptualising Irish rural youth migration: a biographical approach'' International Journal of Population Geography 6 229 ^ 243 Lei H, 1996, ``An analysis of current development of `Mang Liu' in Chinaöconflicts, problems, shocks and influences'' Socialism Studies 6 58 ^ 62 Li B, 2006, ``Urban social change in transitional China: a perspective of social exclusion and vulnerability'' Journal of Contingencies and Crisis Management 13 54 ^ 65 Liu C, 2006, ``Shortage of migrant workers and structural undersupply of labour in the urban labour market'' Population Journal 155 3 ^ 6 Liu K, 2007, ``The Shenzhen miracle: the relationship between the migrant labor force and Shenzhen's economic development'' Chinese Economy 40 47 ^ 75 Liu S, Li X, Zhang M, 2003, ``Scenario analysis on urbanization and rural-urban migration in China'', Institute of Geographic Sciences and Natural Resources Research, Chinese Academy of Sciences, Beijing, http://www.iiasa.ac.at/Admin/PUB/Documents/IR-03-036.pdf Lu S, Song S, 2006, ``Rural ^ urban migration and wage determination: the case of Tianjin, China'' China Economic Review 17 337 ^ 345 Lu Z, Huang Q, Lu T, Zhou W, 2007, ``The process and problems of industrialization and urbanization in China'' Chinese Economy 40 6 ^ 30 McHugh K E, Hogan T, Happel S K, 1995, ``Multiple residence and cyclical migration: a life course perspective'' The Professional Geographer 47 251 ^ 267 Massey D S, 1990, ``Social structure, household strategies, and the cumulative causation of migration'' Population Index 56 3 ^ 26 Meng X, Zhang J, 2001, ``The two-tier labour market in urban China's occupational segregation and wage differentials between urban residents and rural migrants in Shanghai'' Journal of Comparative Economics 29 485 ^ 504 Mincer J, Jovanovic B, 1981, ``Labor mobility and wages'', in Studies in Labour Markets Ed. S Rosen (Chicago University Press, Chicago, IL) pp 21 ^ 64 Oettinger G S, 1996, ``Statistical discrimination and the early career evolution of the black ^ white wage gap'' Journal of Labor Economics 14 52 ^ 78

Migrant workers in the urban labour market of Shenzhen, China

1475

Pan Z, 2007 Society, Subjectivity and Order: Spatial Turn on Peasant-worker Studies (Social Sciences Academic Press, Beijing) Roseman C C, 1971, ``Migration as a spatial and temporal process'' Annals of the Association of American Geographers 61 589 ^ 598 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 number 158, 367 ^ 393 Shen J, 2002, ``A study of the temporary population in Chinese cities'' Habitat International 26 363 ^ 377 Shenzhen Bureau of Statistics, 2008 Shenzhen Statistical Yearbook 2007 (China Statistics Press, Beijing) Siu H F, 2007,``Grounding displacement: uncivil urban spaces in postreform south China''American Ethnologist 34 329 ^ 350 Smart A, Smart J, 2001, ``Local citizenship: welfare reform urban/rural status, and exclusion in China'' Environment and Planning A 33 1853 ^ 1869 Solinger D, 1999 Contesting Citizenship in Urban China: Peasant Migrants, the State, and the Logic of the Market (University of California Press, Berkeley, CA) Topel R, Ward M, 1992, ``Job mobility and the careers of young men'' The Quarterly Journal of Economics 107 439 ^ 479 Wang F L, 2005 Organizing through Division and Exclusion: China's Hukou System (Stanford University Press, Stanford, CA) Wang J, Ding S, 2006, ``A re-estimation of China's agricultural surplus labouröthe demonstration and modification of three prevalent methods'' Frontiers of Economics in China 1 171 ^ 181 Wang M Y, 1998 China Mega Urban Region (Ocean Press, Beijing) Wang M Y, 2002, ``Small city, big solution? China's Hukou system reform and its potential impacts'' DISP 150 23 ^ 29 Wang M Y, Meng X C, 2004, ``Global ^ local initiatives in FDI: the experience of Shenzhen, China'' Asia Pacific Viewpoint 45 181 ^ 196 Wang Z, Cai R, Zhang D, De B, 2007, ``An investigation of crimination of migration people in Beijing'' Journal of the People's Public Security University 126 9 ^ 15 Williams N, 1991,``Re-examining the wage, tenure and experience relationship'' Review of Economics and Statistics 73 512 ^ 517 Xinnongmen, 2008 2007 zhongguo nong ming gong (lanling) bao gao [China peasant worker (blue collar) report], Xin Nong Men Research Institute, Hangzhou Xu W, Tan K-C, Wang G, 2006, ``Segmented local labor markets in postreform China: gender earnings inequality in the case of two towns in Zhejiang province'' Environment and Planning A 38 85 ^ 109 Yardley J, 2004, ``In a tidal wave, China's masses pour from farm to city'' The New York Times 12 September, page 6 Zelinsky W, 1971, ``The hypothesis of the mobility transition'' The Geographical Review 61 219 ^ 349 Zhang L, 2001 Strangers in the City: Reconfigurations of Space, Power, and Social Networks within China's Floating Population (Stanford University Press, Stanford, CA) Zhao C B, Wu G, Zhang C, 2006, ``An analysis of employment of rural migrant labour in 2006'', National Rural Fixed Site Survey Office, Research Center for Rural Economy, Ministry of Agriculture, http://www.rcre.cn/userArticle/ArticleFile/20071210154317735.doc Zheng G C, Huang-Li R L, 2006, ``Zhongguo nonmingong wenti: lilun panduan yu zhengce silu'' [China's migrant workers' problems: theoretical judgment and policy considerations] Journal of Renming University of China 6 2 ^ 13

ß 2010 Pion Ltd and its Licensors

Conditions of use. This article may be downloaded from the E&P website for personal research by members of subscribing organisations. This PDF may not be placed on any website (or other online distribution system) without permission of the publisher.