• No results found

Rural to Urban Migration in China: 


Academic year: 2023

Share "Rural to Urban Migration in China: "


Loading.... (view fulltext now)

Full text


Rural to Urban Migration in China: 

An Overall View.



* Paul Frijters, ** Xin Meng

* School of Economics and Finance, Queensland University of Technology, Brisbane, Australia

** RSSS, Australian National University, Canberra, Australia



China is not merely growing at more than double the rate of the European countries during the Industrial Revolution, it is also urbanising at double the speed. Using a unique dataset of rural-to-urban migrants in 15 major Chinese cities, we give preliminary answers to some of the most pressing policy questions: how many migrants are there and what are their attributes? Are they dissatisfied or are their children doing worse than the children of others? Are they discriminated against in the labour market and, if so, what are the mechanisms via which this discrimination works and where are the market forces to undo the discrimination?

JEL Classification:

Keywords: migration, economic growth, urbanisation, Tiebout, political economy, discrimination.

We would like to thank the Australian Research Council and the Australian Ministry of Development (AusAid) for the funding that made the entire project possible. All errors and opinions in this piece are our own.



China is transforming from an agricultural society which witnessed almost no growth in the 1950-1978 period, to a modern society that is dominated by industry and services. At the end of the 1990s the major cities started to run out of capable cheap labourers and started to import more and more rural migrants. In 2005, the Chinese National Bureau of Statistics (NBS) estimated there to be 130 million workers in the city who originate from the countryside, out of a potential rural workforce close to 600 million. One of the groups of scholars examining the socio-economic aspects of this human population movement is organised around the ‘Rumici project’ and this paper brings together the preliminary answers that this research group has unearthed concerning migrants in China.

The Rumici project (Rural Migration In China and Indonesia) mainly gathers data on actual migrants and is the first project of its kind in China to attempt to track a representative sample of migrants over time. The first wave of data, tracking 5,000 migrants and 15,000 non-migrants in the city and in the countryside, became available in 2008 with the second wave coming out in November 2009. The data includes information about the wages, health, family circumstances, migration history, education history, wealth, and key opinions of migrants.

We pose the following questions in this paper, all posed with an eye on the importance of internal migration in China for ‘the West’:

1. General economic background: how fast has China been growing, when can it be expected to overtake the West, and how dominant will China become in terms of the size of the economy?

2. General migration background: what has been the pattern of migration so far and how does this compare to the urbanisation seen during the Industrial Revolution in the West?

3. Who are the migrants, what kind of role do they play in the economy, and what is happening to their families?

4. Are the rural migrants being discriminated against, how does this discrimination work, does the discrimination make the migrants disgruntled and is the discrimination sustainable?


5. What can we expect to happen in terms of the political economy of migration within China and how will this affect the West?

We realise each of these questions is worthy of a library of its own. Throughout, we will therefore be explicit as to the limited information we base our preliminary answers on.

General Economic Background: Chinese GDP Growth Now  and in the Future. 

The following three graphs show the bare bones of the Chinese economic miracle.

Figure 1: Real GDP growth in China 1981-2009 (source: China Yearbook)

0 2 4 6 8 10 12 14 16

1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009

%   Re al   Ch an ge   pe r   annum  


China's  Real GDP Growth  


Figure 2: Nominal Chinese exports 1982-2009 (source: China Yearbook)

Figure 3: Aggregate composition of the Chinese economy 1981-2009

(source: China Yearbook)

0 200 400 600 800 1000 1200 1400 1600 1800

1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009

Bil   US D


Nominal  Exports of Goods & Services

0 10 20 30 40 50 60

1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009



Composition of Economy

Agriculture/GDP Industry/GDP Services/GDP


These three figures show that the Chinese economic miracle has been a classic export- lead story closely mimicking the experience of the Asian tigers in the 60s to 80s.

Exports have risen faster than its level of GDP, with the total level of exports overtaking that of Germany in the mid 2000s such that China’s export is now only second to that of the US.

In terms of GDP, when China relaxed the central control system in 1978, its nominal GDP was roughly comparable to that of Costa Rica, standing at around 200 US (1990) dollars per person per year. Currently, the best PPP estimates available put Chinese GDP at roughly 5000 US dollars per person per year, having increased roughly 10% each year in the whole 30 year period. Nominal GDP is four times smaller, but this is mainly due to low prices for non-tradable goods (housing in particular) and an active policy of buying up US government debt in order to keep the exchange rate artificially low. Hence, China is already the second largest economy in the world.

The sectoral composition of the Chinese economy, as shown in Figure 3, is far more industry-intensive than the economy of any OECD country, which typically constitutes about 70% services and 25% industry. This underscores the fact that the Chinese economic miracle has occurred via export-lead manufacturing. If China follows the pattern typical of the Asian tigers (including Japan, South Korea, Singapore, Taiwan, and Hong Kong), it will soon start to expand its service industry to eventually dwarf its now dominant manufacturing sector.

Figure 4 below shows the relative size of various regions of the world in terms of PPP dollars, past, current, and future. The Appendix details how these figures were arrived at, but we can here mention the key aspects: we took the Maddison tables on PPP- adjusted growth for all regions up to 2004, extrapolating the economies for further years using the estimated UN-World Book reported growth rates. We grouped countries together into blocks in order to prevent the well-known phenomenon of having small countries with high recent growth rates expected to grow beyond reasonable bounds. We then extrapolated ahead using a very simple rule of thumb for developing countries, based on the experience of the Asian tigers. In particular, we presumed growth rates to stay stable for the US and other developed regions, with the growth rates of China and India also held stable until they hit about 66% of the per- person GDP of the US, after which they are presumed to follow the world growth rate. The 66% figure is the average level of GDP at which Japan, Korea, Singapore,


and other Asian countries stopped outgrowing the US and there was no marked period of slower or faster growth between the moment they clearly started developing (defined as 5 consecutive years of outgrowing the US) and the period they stopped developing (less that 4 years out of 5 in which they outgrew the US). As to population growth rates, we simply adopt the extrapolations published by the UN. We realise that to take historical patterns as an indication of the future is somewhat arbitrary, but will mention that one can easily rationalise this with a macro-model of country-specific institutional limits to growth rates.












1950 1980 2004 2040


World GDP by regions

(million in 1990 int Geary-Khamis dollar)

Africa Other Asia India China

L America EE/fUSSR Japan AUS/NZ/CAN

USA Western Europe

5,329,719 20,047,814 43,553,044 216,697,44


Figure 4: Relative size of the major economies in the world (source: Frijters et al. 2008).

The main interesting aspect of the figures above is that ‘The West’ produced about 60% of the world economy in 1950, close to 75% if we would include the regions of Eastern Europe and the former USSR (EE/fUSSR) into our definition of ‘the West’.

China accounted for less than 5% of world GDP then, even though it accounted for more than a quarter of the world population. By 2004, China already accounted for close to 12% of the World economy, projecting to account for close to 35% of the world economy by 2040, which is a little over the size of the whole of the ‘extended West’ combined. Interestingly enough, by 2040 India will be the other main economic powerhouse, accounting for an additional 15% of the World economy and projected to eventually overtake China. As India’s population grows faster than that of China it will catch up, even if its growth spurt is less spectacular than that of China. With these growth rates, China is expected to overtake the US as the biggest single economy around the year 2019.

We can make a couple of remarks here as to whether there are any obvious signs that the Chinese economy will run out of steam in the coming decades, i.e. anything that would make it likely that it will not reach the 66% per-person GDP of the US. Reflect on the following figures in this regard:

- According to the China Yearbook, average saving rates in China are close to 35%, which are exceptionally high for any country and time period.

- Education levels are soaring, with the current generation having at least 3 years on average more education than the generation of 25 years ago (see further sections for direct evidence).

- Very ambitious long-term investments, including an ambitious program of setting up more universities, covering the whole country with highways and modern communication technologies.

- Rapid modernisation of the central and city bureaucracies itself, witnessed by the introduction of new regulators, the gradual opening of China’s capital market, and the enormous expansion in city-based higher education (Tobin, 2005).


- A boom in agricultural production and an increase in food prices (Meng et al.


In short, each of the production factors one usually thinks of (human capital, technology, government, physical capital, resources) is accumulating with no major impediment to sustained growth in sight. If China were to follow the examples of Korea and Japan, then the somewhat vulnerable reliance on foreign exports as a source of growth should also start to diminish soon as Chinese industries start to cater for Chinese consumers.

At face value therefore, there is no major impediment to report on the macro- economic side that would prevent China from rapidly overtaking the West as the largest economy in the world. This of course will have geo-political implications on many fronts. The dominant economy in the world will undoubtedly be dragged into local regional conflicts in other parts of the world if these conflicts threaten its supply of raw materials or access to consumer markets. Though internally more homogeneous than any other major country (some 92% of the Chinese population is part of the Han-Chinese ethnic group, all using the same script and speaking very similar languages), China does have its own share of ethnic and religious minorities who can potentially drag it into various conflicts with its neighbours or with world religions.

Here, we can briefly mention a few other salient aspects of the Chinese growth experience:

1. Though dominated by export-industries located in the major cities, there has also been spectacular growth across the more rural provinces of China (Unel and Zebregs 2006), partially fuelled by the development of inland cities like Chongquing or Luoyang. When the Chinese economy becomes more inwardly-oriented via an expanded service sector there is no good reason to expect this trend to be unsustainable in the medium-run. This will also open up new opportunities for rural workers to migrate to.

2. Following a whole set of reforms in the 90s the urban welfare state has effectively been dismantled, except for free education. Unemployment, old age, and health are no longer insured by the state and are now only insured via employers or within extended families. This partially explains the extremely high rates of savings (Meng, 2007).


3. The Chinese Communist party has changed from being a vehicle for communist ideology to an accommodator of economic networks.

Entrepreneurs were welcomed into the party in large droves in the mid 2000s;

membership has grown to about 6-7% of the population; and embryonic democratisation has occurred within the party. This has arguably removed political impediments to economic growth.


General Migration Background. 

In the figure below, we show the estimates of the World Bank regarding the number of rural migrants in China. They were generated by asking individuals in rural villages how many members of their village were now in the cities.

Figure 5: Total rural to urban migration 1997-2005 (source: World Bank)

The most glaring aspect of this figure is the sheer magnitude of the difference between the level of migration at the start in 1997 and 8 years later. The level of migration more than tripled during the period, showing that before the late 1990s the cities effectively grew without many rural migrants. However, since the late 1990s it can be seen that they started absorbing migrants on a grand scale. If we were to simply take a low estimate of the linear projection of these trends (say, 10 million net additional migrants per year), then there would have to be around 165 million migrants in 2009. Put otherwise, urbanisation increased 10% in about 12 years solely due to rural to urban migration (hence neglecting the transformation of villages into cities).

Compare this level of exceedingly rapid urbanisation with the closest historical analogy we can think of, the industrial revolution in Europe, where GDP growth rates


49.4 52.0


78.5 84.0




20 40 60 80 100 120 140

1997 1998 1999 2000 2001 2002 2003 2004 2005 Migration

population (millions)

Total rural migration …


were on average 2% with short periods of higher growth during the period 1800-1950.

The picture below, which uses data from De Vries (1984), Grauman (1977), and the UN (with a threshold level of population density as a rough definition of an urban community) shows the speed of urbanisation in Europe.

Figure 6: Urbanisation in Europe 1500-2025 (source data: Jan de Vries 1984;

John Grauman 1977; UN calculations (after 1950))

This picture highlights that before the industrial revolution, which roughly started around 1800, Europe was still 90% rural. At its fastest, in the period 1900-1950, it took Europe 40 years to increase from 30% urbanised to 50% urbanised. Compared to this figure, it is clear that China is currently urbanising at least twice as fast.

We may also mention here that the period of urbanisation in Europe was one of emerging ideologies. It was the time when nationalism became firmly established and a period that witnessed the advent of the temperance movement, the puritans and extreme forms of Protestantism, etc. We cannot claim that these developments were causally linked, but it is an interesting hypothesis to think that the extremist ideology had something to do with the large population of rural migrants who felt alienated in the big cities and who found a home in the emerging ideologies of the day. This overview paper is not the place to argue this in greater detail, but from an overall

Urban Rural












1500 1525

1550 1575

1600 1625

1650 1675

1700 1725

1750 1775

1800 1825

1850 1875

1900 1925

1950 1975

2000 2025


perspective the link between mass-migration to mass-ideology via mass-loneliness is perhaps the most important in terms of the possible dangers migration poses to stability.

Even without a possible dangerous cultural change component to migration, there are immediate reasons to believe that mass migration will seriously change the internal dynamics of China. The main reason is that the countryside is emptying of young people, consequently undermining the traditionally important role of the elderly within village life. The land ownership of the elderly is now less important than the income-generating role of the young workers in the cities, thus reversing the traditional balance of power within families (see Chang, 2008, for an insightful account of the changing social relations in the city and in the countryside due to migration).


The Rumici Project: Who are the Migrants? 

The Rural Urban Migration In China and Indonesia (Rumici) Project is a panel data study into migration. The Chinese part of the project consists of tracking about 5000 migrants and their families for 4 years, starting in 2008. As comparison information, the Rumici Project also tracks 5000 urban households and 10000 rural households in the same period, asking the same questions.

The questionnaire administered to these samples is found at rumici.anu.edu.au and includes the usual questions of a panel dataset, i.e. information on jobs, incomes, health, education, and social relations. The main difficulty in the case of China was to find the migrants, for they are not officially registered anywhere, implying that a mini-census had to be undertaken in each of the 15 cities in order to find representative samples. More information can be found in the book produced on the first wave by Meng et al. (2009, eds).

The key definitions are that rural migrants report to have non-urban hukou and are nevertheless working and living in the cities; urban individuals have urban hukou; and rural non-migrants are member of households that do not include migrants and that have non-urban hukoui (on the importance of hukou see Meng et al. 2009).

The following three figures reveal information about the age, education, and income of the migrant households in the first wave (2007-2008).

Figure 7: The age distribution of migrants, rural non-migrants, and urban individuals.


Figure 8: Average education by age for migrants, rural non-migrants, and urban individuals.

Figure 9: Average per capita income for migrant, rural non-migrant, and urban households.


Figure 7 reveals that the migrants are disproportionately young compared to urban individuals or rural non-migrants. This is of course closely related to the fact that a large proportion of the migrants have only relatively recently arrived in the cities (see Figure 5). One can also see that young children older than 4 are disproportionately missing from the migrant sample, which is because in most cities migrant children are not allowed to go to school if they do not have urban hukou. Hence the children of migrants often remain in the countryside looked after by family members.

Figure 8 shows average education by age for the three samples and shows many interesting trends. From a macro-perspective, the most important fact is perhaps that the total number of years of schooling has increased tremendously for all three groups, with the youngest cohorts having at least 3 more years of education than their grandparents. It also shows that the elderly belonging to the migrant households are the relatively higher educated elderly amongst the countryside, implying that it is the children of the relatively well-educated who have been first to take advantage of the opportunity to migrate. Another salient feature is that there is not a substantial difference in terms of education between those who leave for the cities and those who stay behind in the countryside. This indicates that the majority of the gain in education is universal. Finally, Figure 8 shows the large gap between the countryside and the city in terms of levels of education, which is partially because university education is almost solely a city-phenomenon. Indeed, there is cream-skimming of migrants in that those migrants who pass the university entrance exams are offered urban hukou, implyingthat they are no longer defined as migrants (we unfortunately have no statistics on how many of these there are, but it is likely that there are few because the expansion in tertiary education was only recent and focussed on the cities).

We can briefly mention some other salient characteristics of the migrants. Less than 3% are from an ethnic group, even though 8% of the whole Chinese population is from an ethnic group ( even though the urban population itself has about double the concentration of ethnicities compared to the migrants), implying that the economic advantages in our 15 cities are mainly taken up by Han-Chinese. In addition, the average number of working hours of the migrants is about 58, compared to 43 hours for urban workers, indicating that the migrants work exceptionally hard.


Policy Relevant Aspects of the Migration Experience The children of migrants: are they better or worse off ?

One important policy question is whether the migration of a large part of the prime- age population of China adversely affects the children of migrants. If so, this would create costs in the future in terms of reduced education, health, etc. Conversely, one might expect them to do better because of the additional income generated by the migration of their parents.

The next figure summarise the more extensive information on this issue generated by Kong and Meng (2009).

Figure 10: Physical height of the children of migrants compared to children of urban hukou holders.

Figure 10 illustrates that there is a difference in height between urban children and the children of migrants, but that the difference is quite small. This small difference is also evidenced by the absence of a negative difference in the degree of health problems experienced by children as reported by their parents (see Kong and Meng 2009). This is despite the fact that the amount of health spending is far greater for urban children. One possible reason for this apparent contradiction is that the countryside is in fact healthier in terms of pollution levels and opportunities for exercise than the urban environment.

60 70 80 90 100 110 120 130 140 150 160 170

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

Height (CM)


Urban children migrated children


Kong and Meng (2009) similarly depict trends in education. The children of migrants are not more likely to drop out of school or have problems at school, as reported by their parents. Indeed, the self-reported performance of school children is slightly higher amongst migrants, but this may be partially due to ignorance on the part of migrants as to how their children are doing or the greater level of ambition on the part of urban parents.

The bottom line of these observations is quite straightforward though: at present there is no indication that the children of migrants are seriously worse-off than the children of non-migrating households, and that there is little difference in terms of health status. What is still true is that the children of migrants, like other children connected to rural hukou, can be expected to end up with much less education than urban children. This is a difference that is yet to be addressed by policy makers.

Family Formation and Migration

The question emerges whether the current pattern of migration is ‘costing’ China a lot of unborn children. Moving to the city and working exceptionally long working hours can be expected to crowd out other activities in the 20-35 age bracket and one other important activity is family formation. Hence, the next figure gives information on the relationship between migration and family formation (see Ning et al. 2009).

Figure 11: The percentage of women with at least 1 child by age

-0.2 0 0.2 0.4 0.6 0.8 1

10 20 30 40 50 60

The percentage of women with child


The percentage of women with child by age (moving average)

Rural Women without Migration Experience Rural Women with Migration Experience Urban Women


Figure 11 shows that there is quite a significant degree of delay in child-bearing age for migrants compared to their non-migrating rural counterparts or the urban hukou holders. Whilst the recent glut in migration has occurred it is still too soon to say for certain that lifetime fertility is going to be affected for migrants. Nevertheless, it is a near certainty that it observed decreases in fertility will exist because the ability to conceive itself goes down as womenage and we can thus be certain that there will be more childless migrants in the future due to the delay in childbearing. If these trends continue and it becomes a general pattern that the majority of the prime age workers from the countryside work in the cities and delay their childbearing, then Chinese overall fertility is certain to drop substantially in the coming decades.


Migration and Discrimination

Equity amongst migrants and urban households is perhaps the most immediate and pressing policy issue. Unlike the migration in Europe during the industrial revolution, the Chinese migrants already self-identify in a similar way to the urban households:

their script is the same and the differences in dialects are relatively small. Also, as mentioned above, they self-identify as Han-Chinese and receive exactly the same education in school regarding their ancestry. This means that any strong and open degree of discrimination between rural migrants and urban households has the potential to raise enormous levels of political unrest should the migrants start to feel they are unfairly treated.

The key information on differences in hourly total compensation across the 15 cities is given in the figure below, where the data comes from Frijters et al. (2009):

Figure 12: Median hourly compensation for urban workers and migrant workers in the 15 Chinese cities of the Rumici project.

Figure 12 shows the raw difference in median hourly compensation for workers across cities, where hourly compensation includes wage earnings, employer-provided welfare payments (primarily unemployment, health, and pension insurance

Chengdu 9.86

4.39 Chongquing 10.27


Wuxi 12.01 Luoyang 9.56

9.65 3.36

Guangzhou 19.96


Shenzhen 19.20

7.22 Hefei


4.70 Ningbo

18.19 5.95 Hangzhou

15.07 7.37 Zhengzhou

9.63 4.33

Bangbu 6.41

3.70 Shanghai

19.55 6.52 Nanjing

14.29 6.73

Wuhan 11.23


Dongguan 13.22



contributions), and in-kind food and housing support. These hourly differences do not take account of the fact that urban workers are more highly educated, nor the fact that the migrants are of prime-age compared to the urban workers who are often past their optimal age. Nevertheless, the wage differences are striking. For one, the wage differences between cities are simply staggering: the median urban person in Shanghai earns about 3 times more than the median urban person in Bangbu. Such a large difference is probably only maintained due to the difficulty of moving across cities even for urban hukou holders (hukou is specific to a place).

The difference in hourly compensation between urban workers and migrant workers is also striking, with the migrant workers earning no more than about 37% of an urban worker’s hourly rate .

Before we present wage-differences corrected for the human capital variables we observe, we first want to address the question of whether there is actual evidence of discrimination and what channels the discrimination would go via:

- Are the migrants legally discriminated? Indeed they are. They do not have urban hukou and hence their children are usually not allowed to go to school in the cities. Furthermore, each city has its own regulations regarding the jobs that migrants are allowed to do, with most cities reserving government jobs for urban hukou holders (see Frijters et al. 2009). Worse, in many cities, migrants are not allowed to have managerial positions, which implies that their options within the competitive sector are also limited to the lower-paying jobs. In some cities, there is even an explicit guest-worker system with migrants staying in specialised accommodation.

- Do the migrants have equal access to credit? On the whole, no; because they have no clear legal status they basically cannot access urban credit markets.

Their high degree of mobility and inability to use any land back home as collateral prevents them from credibly taking out loans at banks. As a result, Frijters, Liu, and Meng (2008) found that about half of the migrants who want to be self-employed are self-employed. Preliminary information on the second wave shows that those who are not self-employed indeed self-report to have difficulty gaining access to credit. What credit there is, is informal, i.e.

migrants borrow from family and friends or self-finance via savings.

Interestingly, self-employed migrants earn about 40% more than non-self- employed migrants and migrants who are self-employed estimate they would


earn about 40% less if they were not self-employed! Also, some 25% of migrants are self-employed compared to only about 5% of the urban population. All these ‘strange’ facts make sense if one sees self-employment as an attempt to circumvent the legal restrictions migrants face with regards to having high-paying jobs. Starting ones own business is then a means of realising ones actual potential, something the urban workers have no need for because they are not discriminated against within the cities.

We would thus want to argue that there is indeed direct evidence of market imperfections when it comes to the labour market migrants live in. They face legal discrimination in the cities regarding the entitlements of their children, and they face a missing credit market due to their higher mobility, uncertain legal status, and inability to use any assets they own in the countryside as collateral. These ‘restrictions’ benefit urban insiders who can thus be expected to capture the administration of their cities in order to protect the rents these restrictions bring them.

Are the migrants unhappier due to their discrimination?

Figure 13 illustrates the distribution of happiness across migrants, urban individuals, and non-migrating rural individuals.


Figure 13: Happiness amongst migrants, urban individuals, and non-migrating rural individuals.

Findings show some happiness differences between the three groups do exist.

Migrants are a little unhappier than the urban individuals and the rural individuals are the happiest. Though each of these differences is certainly statistically significant given the large numbers of individuals involved, they are not really economically significant. Compared to the differences in happiness seen across countries (see Clark et al. 2008), these differences are not larger than those between closely related OECD countries, such as France and England. Hence, the overt discrimination faced by the migrants, the differential treatment of their children, the fact that many cannot access the credit that would allow them to escape their wage-paying jobs in favour of self- employment, and the differential pattern of lifetime fertility that their situation has lead to, has not translated into a clear degree of resentment with their own life.

Despite using a non-representative sample of migrants that probably oversamples the more successful migrants, Frijters, Liu and Meng (2008), find that a large part of the migrants’happiness is attributable to the high levels of expectations they have regarding their incomes. This suggests that the lack of open dissatisfaction is gained through the high sustained levels of economic growth.

The figure also shows that the happiest group is in fact made up of the rural residents who do not have migrated family members. This is remarkable because they are also the poorest group, with the gap in wealth and wages relative to urban individuals increasing over time (Meng et al. 2005; Meng 2007).

Market forces undermining the discrimination

The central government in recent years has undertaken various legal initiatives to address the legal entitlements of migrants, which we will only briefly mention.

Amongst these are compulsory insurance schemes for all workers, an embryonic migrant registration system, and legislation outlawing particular forms of legal discrimination within cities. We do not yet know whether these regulations will be enforced and how widely they are actually intended to be implemented.

One may wonder if there are market forces that could break the discrimination of migrants even in the absence of government regulation. After all, restricting migrants from attaining their potential leads to market opportunities for those able to find a way


to circumvent the restrictions. Some possibilities we can only speculate about, such as the possibility of attaining urban hukou through marriage. There is one possibility we can directly provide information on, which is the possibility of individual cities to undermine the rent-seeking opportunities of the insiders of other cities.

If one city does not share the same restrictions, what can be expected to happen? In most cases, over time prospective migrants will learn that they face better opportunities at the ‘fair’ city, implying that that city would enjoy an influx of highly able migrants from which their employers could choose. Given that we are still talking of a situation in which cities grow so fast that employers only need to pay marginal productivity and thus can enjoy the difference between the average productivity of the migrants and their marginal productivity (i.e. credit markets haven’t caught up to the extent that no profits are being made), it should be clear that the deviating city may well enjoy high rents from undercutting the restrictions imposed by other cities. This is of course a particular example of the Tiebout theorem at work (Tiebout 1956), where cities in competitive equilibrium are forced to offer the amenities that attract enough residents because its residents will otherwise migrate elsewhere.

Is there any evidence for the Tiebout theorem at work? Consider the evidence of the percentage of workers whose employers indeed pay the ‘compulsory’ workers insurance by city:


Figure 14a: Percentage of migrant and urban workers with pension coverage.

Figure 14b: Percentage of migrant and urban workers with unemployment insurance coverage.

Figures 14a and 14b show that Wuxi, a medium-wage level city, has indeed started to offer similar legal packages to migrant workers in that the difference between insurance coverage is small in that city. For the other cities, the difference in insurance coverage is often enormous. In Shanghai for instance, almost no migrants are covered by the ‘compulsory’ insurance, whilst nearly all urban workers are covered. In yet other cities, like Dongguan, insurance coverage of migrants is about half that of rural individuals.

We now show the percentage of the wage differences in Figure 12 that are not explained by observable characteristics (taken from Frijters et al. 2009, using standard Oaxaca wage decompositions), where the observed characteristics include age, tenure, schooling, health, sector, gender, firm-size, marital status, and height:


Figure 15: Percentage of the raw median wage difference not explained by characteristics.

The unexplained difference is calculated by predicting the wage of the median urban individual in each city if they were paid according to the wage schedule of the migrants and then calculating how much of the difference between the median urban and the median migrant worker in that city was not explained by it.

The most striking aspect with respect to Figure 15 is that the percentage unexplained is actually negative for Wuxi, which we interpret to mean that a migrant there has a higher unobserved quality component than the urban worker. This is in direct agreement with the prediction from the Tiebout model that offering a fair legal system would entice the better migrants to go to that city.

Figure 15 also shows that in most cities, the observed characteristics explain very little. In Shanghai for instance, nearly 80% of the wage difference is not explained by the observable characteristics, concurring with the fact that the percentage of migrants receiving their ‘compulsory’ employer contributions is almost 0% in Shanghai.

Conclusion and Discussion 

Chengdu 70.36%

Chongquing 68.91%

Wuxi -28.22%

Luoyang 63.74%


74.41% Dongguan


Shenzhen 74.30%

Hefei 66.55%

Ningbo 70.62%

Hangzhou 68.30%


63.53% Bangbu

29.75% Shanghai


Nanjing 59.76%

Wuhan 75.58%


In this paper we have attempted to argue that, according to the limited evidence we have available, the largest human peacetime migration ever is so far happening without significant obstacles. In particular, we have argued that:

- The migrants are not much unhappier than their rural or urban counterparts, even though they work very long hours (58 per week compared to urban workers’ 43) and are usually not allowed to stay permanently.

- Their children are not experiencing too many problems in that their levels of health as reported by their parents are not worse than those of urban children, and that the levels of reported problems at school are slightly less than that reported for urban children.

- The fertility is affected in that migrants significantly delay their time till first child, but it is an open question whether the decreased lifetime fertility that this is likely to lead to is a good or a bad thing.

In terms of the economic forces we found to be at work on the labour market we can say that:

- The migrants are legally discriminated against within cities, in that the jobs they can do are restricted, and they usually do not get the welfare insurance they are entitled to (pensions, health, unemployment). As a result, their median hourly compensation is on average only 37% of that of urban workers, of which only a small part is explained by observed characteristics like education and tenure.

- The Tiebout theorem is at work in that a city has been identified (Wuxi) which seems to be able to attract the better migrants by offering a fairer legal system. If other cities follow suite, then the legal discrimination would be overcome by market forces in the absence of central government intervention.

What can we further expect to happen? Here, we enter the realm more of speculation, but we would expect to see the same issues emerge in Chinese capitalist cities as is evident in Western capitalist cities and in other cities in Asia (such as Singapore). In particular, if the government will not provide welfare insurance and include migrants in joint city identity, then we expect religious groups and gangs to enter the market for these psychic needs. Iannaconne and Berman (2006) have an explicit model of


religion that argues this basic mechanism, even though they were more concerned with explaining Muslim fundamentalism than future cultural changes in China.

Finally, will any of the issues signalled above affect the West in any material way? It seems very unlikely, simply because the level of internal unhappiness is quite low at the moment and none of the indicators of future growth (education, health, investment, savings, etc.) suggests the Chinese growth miracle has run its course.

Frijters, Liu, and Meng (2008) argue that the low level of unrest stems from the high level of realistically optimistic expectations fed by the high level of actual growth, again implying one should not expect much internal unrest in the near future as long as growth persists.

Perhaps there is one mechanism which will directly affect the West, which is that a great many Chinese will be able to afford to travel outside of China as tourists or migrants. This would by and large seem to herald a positive externality for the ‘West’, i.e. it would constitute a form of brain drain.



Chang, L.T. (2008), Factory Girls, Picador, Pan McMillan.

Clark, A., Frijters, P., Shields, M.A. (2008), ‘A survey of the income happiness gradient’, Journal of Economic Literature, 46(1), March 2008, pp. 95-144. also IZA, NCER, and DELTA discussion paper.

Frijters, P., Dulleck, U., Torgler, B. (2008), Economics for decision makers, Cengage Publishing.

Frijters, P., Johnson, D., and X. Meng (2009) ‘The mental health cost of long working hours: the case of rural Chinese migrants’, mimeo.

Frijters, P., Lee, L., and X. Meng (2009), ‘Jobs, Working Hours, and Remuneration Packages for Migrants and Urban Residents’, in Xin Meng, Chris Manning, Tadjuddin Effendi and Shi Li (ed.), The Great Migration: Rural-Urban Migration in China and Indonesia, United Kingdom, Edward Elgar Publishing.

Frijters, P., Liu, A.Y.C., and X. Meng, (2008). "Are optimistic expectations keeping the Chinese happy?," NCER Working Paper Series 37.

Grauman, J.V. (1977), “Orders of Magnitude of the World's Urban and Rural Population in History,”. United Nations Population Bulletin, 8, 16–33.

Kong, T., and Meng, X. (2009), ‘Children of Migrants in China: Health and Education Outcomes’ in Xin Meng, Chris Manning, Tadjuddin Effendi and Shi Li (ed.), The Great Migration: Rural-Urban Migration in China and Indonesia, United Kingdom, Edward Elgar Publishing.

Meng, X, Gregory, R., and Wang, Y. 2005. "Poverty, inequality, and growth in urban China, 1986-2000," Journal of Comparative Economics, Elsevier, vol. 33(4), pages 710-729


Meng, X. (2007) "Wealth Accumulation and Distribution in Urban China," Economic Development and Cultural Change, University of Chicago Press, vol. 55, pages 761- 791.

Meng, X., Manning, C. (2009), ‘The great migration in China and Indonesia: trends and institutions’, in Xin Meng, Chris Manning, Tadjuddin Effendi and Shi Li (ed.), The Great Migration: Rural-Urban Migration in China and Indonesia, United Kingdom, Edward Elgar Publishing.

Meng, X., Manning, C., Effendi, T., and S. Li (2009, ed.), The Great Migration:

Rural-Urban Migration in China and Indonesia, United Kingdom, Edward Elgar Publishing.

Ning, D., Frijters, P., Meng, X. (2009), ‘Fertility and migration: a duration analysis’, mimeo.

Tiebout, C., "A Pure Theory of Local Expenditures", Journal of Political Economy, 64(5):416-24, 1956

Tobin, D. (2005), ‘Economic Liberalization, the Changing Role of the State and

“Wagner’s Law”: China’s Development Experience since 1978’, World Development (33), Issue 5, Pages 729-743.

Vries, Jan de (1984), European urbanization 1500–1800, Associated book Publishers.

i Note that this category is not entirely representative of the rural population since the rural population also includes households with missing members who migrated. These are excluded from the

comparison categories for ease of interpretation.


Related documents

• Teachers learn quickly to not trust students • Most of the rules, policies and procedures we teachers have set up in the classroom are because of perhaps only one student that

Managers responsibilities: to ensure that staff are aware of the Public Interest Disclosures Act 1994 and that any report of corrupt conduct or maladministration and serious and