6Measuring_and_comparing_construction_activity

6 Measuring and comparing construction activity internationally

Introduction

The notion that there is a predictable relationship between construction and economic development was first put forward by Turin (1969) and subsequently developed by Drewer (1980). Since then, a body of construction industry development theory has been built on that literature, including Bon (1992, 2000), Crosthwaite (2000), Ofori and Han (2003), Ruddock and Lopes (2006) and Gruneberg (2009).

One would expect wealthier economies to have larger construction outputs. Turin’s suggestion, however, was that, as an economy grows, the proportion of construction in national output would increase until, at some point, it levels out. Bon (1992), using the same kind of data, but updated, suggested what has come to be known in the literature as the ‘Bon Curve’: that, as national output continues on a growth path, after a period of levelling out, construction will tend to decline as a proportion of national output.

Two interlinked phenomena were suggested by Turin and Bon. Firstly, there is a direct relationship between construction value added per capita and GDP per capita and, second, as countries grow wealthier, construction value added tends to increase at a faster rate than growth in the overall economy. Bon’s refinement is that, at a certain point, growth in construction value added tends to slow and, eventually, begins to decline as a proportion of GDP. More recently, Gruneberg (2009) has suggested that, while there may be an increase in construction activity in the initial stages of economic development, this may subsequently slow down. Thereafter, construction may rise and fall depending on changing circumstances, but with no particular or predictable pattern.

The studies mentioned and other analyses of construction’s contribution to the economy are in terms of construction value added, presumably because data in that form is more or less readily available from national accounts. Moreover, construction data and GDP are invariably converted to a common currency – usually $US – using commercial exchange rates. The use of value-added and commercial exchange rates is questioned in this chapter. The chapter casts a fresh eye on the contribution of construction to national economies and how that contribution varies across countries and over time.

Users and uses

Construction activity is distinct from but obviously related to the property sector. However, it does not include the value of land. Construction activity is also separate from but increasingly overlaps with construction professional services. The cost of at least some construction professional services is often included in construction contracts and, therefore, construction output data, although it may be provided by consultancy firms subcontracted to construction contractors.

The main potential users of reliable national and international construction data include construction companies (contractors and consultants), construction industry clients, property agents and developers, government agencies, researchers and the press. Their needs are varied but largely comprise market or statistical data and analyses, including data on output and new orders, costs and prices and productivity. Their interest in construction data extends to levels (absolute values) and trends in:

· the size of markets for different types of construction work, for example, housing, office buildings and power projects. There is often interest locally and internationally for particularly large programmes of work or for significant falls in the demand for construction work. A number of specialist market intelligence organizations produce construction market reports, usually based on official data.

· prices for the main construction resource inputs. Rapidly rising or falling resource prices are of interest to people inside and outside the industry. Price changes are usually an indication of local demand pressures but sometimes result from international demand for particular resources. Structural steel is an example in recent years. The construction press publishes regular articles on construction price trends, for example, Gardiner and Theobald (2012), Rider Levett Bucknall (2013) and Turner and Townsend (2013); there are also annual – or more frequent – price books in a number of countries, such as Rawlinsons’ Australian Construction Handbook (2013).

· output prices for particular building types or types of work. Steep increases or decreases in these also usually follow market demand. Data on output prices is often published with data on resource prices. The effects of construction price trends should not be confused with property price trends. In the UK, for example, the effect of land prices on house prices is frequently ignored, and the cause of property price increases is wrongly attributed to the construction industry.

· construction price level differences across countries. This information is provided on a regular basis by international firms of cost consul-tants, such as Compass International (2013). Usually these include the prices of resources and different types of construction work. Priced items relate to representative local materials and products and projects that are representative of local practice and standards; and prices are usually brought to a common basis using commercial exchange rates. There have, however, been efforts recently to take a more considered approach, for example, by Faithful and Gould (2013) and Turner and Townsend (2013).

· indicators of national competitiveness and productivity. These are of great interest to industry bodies, national governments and the press and often become a source of national pride or national concern. In fact, accurate, meaningful measures can be elusive, but that does not prevent regular publication of reports providing ‘data’ ranking countries’ performance.

However, because of the ways in which construction is measured and presented, what is measured and how it is presented provide opportunities to use the wrong data or to use it in misleading ways. These issues are discussed later. In addition, it is important that some kind of commentary or critique accompanies published construction data; this is particularly appropriate for the press and the wider dissemination of the data. The opportunities for misunderstanding or misinterpretation – deliberate or not – of construction data are great. More effort needs to be made to explain terms and concepts and how they should be used and interpreted. There also needs to be explicit acknowledgement of the shortcomings in the quality of much construction data.

Measures of construction activity

Two aspects of measuring construction activity are what is measured and how it is presented. A major source of inconsistency in international comparisons is the variety of definitions used, what is included and what is excluded from national measures of construction activity. As has been noted by Meikle and Grilli (1999), construction activity in the national accounts is not consistent either because some activities are included elsewhere in the national accounts (for example, construction design and management services are usually included in professional services), because they are not included at all or because they are only partially included.

A recent unpublished pilot survey undertaken by Meikle in 2011 on behalf of the African Development Bank (AfDB) revealed an even more complicated situation in African countries. Unlike most developed countries, the majority of countries in Africa have a high proportion of construction activity that is informal or undertaken by households that is often unrecorded. Recorded construction activity is usually broken down into work by registered and unregistered contractors and households and may or may not include work by small firms or individual tradesmen and may or may not include very small projects. Surveys, where they are undertaken, are usually sample surveys, and the value of most work is estimated based on material consumption or other indicators. The 2011 AfDB survey indicated that different things are measured in different countries and in different ways. In some countries, only work by registered contractors is included in the national accounts; in others, virtually everything that could be included is.

There are three commonly used definitions of construction activity: construction value added, gross construction output and contractors’ output. Construction value added and gross construction output are national accounts concepts and are supposed to include all activity that produces construction output, including work by contractors, work by government agencies, work by households and informal (black- or grey-economy) construction output. The third measure, contractors’ output, is the gross value of construction work produced by construction contractors. Because the output of construction is so heterogeneous, aggregate quantity measures of construction are not feasible, and using monetary values is the only option.

Construction value added is the component of GDP that measures the value added by construction firms to inputs from other parts of the economy to produce gross construction output. Typically, it includes compensation of employees, depreciation of capital investments and gross operating surplus, crudely, labour, depreciation of capital and profit. The concept of value added avoids double counting in national accounts, as the aggregate of all value added by all economic sectors is the measure of GDP.

Gross construction output, the second measure, appears as a component of national investment, Gross Fixed Capital Formation (GFCF). It is the aggregate of construction value added and the value of contributions from all other parts of the economy that produce construction works. Gross output is the sector’s turnover or the amount paid by its customers for its products. Strictly speaking, investment in construction excludes construction repair and maintenance activity, although it includes major refurbishments and extensions, but this distinction is not always observed. The 2005 International Comparison Program (ICP) (International Bank for Reconstruction and Development 2008), for example, breaks down GDP in expenditure terms, by investment and consumption, but includes all construction expenditure as investment.

The third measure, contractors’ output, is the total work undertaken by registered contractors and, in some cases, estimates of work by public-sector enterprises and unregistered contractors. Own-account work by commercial organizations is included in the organizations’ output.

Construction in the national accounts

Table 6.1 sets out the breakdown of construction activity in national accounts required by the United Nations International Standard Industrial Classification (ISIC).

Table 6.1 SIC activity categories for construction

The national accounts work categories are intended to include all construction activity by all agencies, not only work by contractors and direct works organizations. The categories, however, are not particularly helpful or complete. There are only two subcategories for building work (commercial and residential) when building probably represents more than half and often more than three quarters of construction activity in most countries; there is rather more detail for civil engineering work (seven subcategories); and there is no separation of public and private work or new work and work to existing buildings.

It is fairly common but not universal in national accounts, where there is a breakdown of construction output, to provide data on residential, nonresidential and civil engineering work. This is equivalent to ISIC categories 41.20/2, 41.20/1 and 42. It is often not possible, however, to obtain reliable data on construction output by trade (ISIC category 43) or preconstruction activities (ISIC category 41). There is a real need to review and revise the ISIC categories to reflect statistical, industry and other needs and modern practice.

Normalizing construction data

In order to compare value data on construction activity within countries, it is necessary to be able to adjust for differences in purchasing power in terms of type of work as well as the date of construction, location and even qualitative differences. This is done with different levels of reliability in different countries. In the UK, for example, a number of methods are used, such as cost and price indices, to bring construction values for different types of work undertaken at different times to a common price basis.

Spatial variation factors, the focus of this chapter, can be used to adjust for project location. Projects in dense urban or remote locations tend to have higher price levels than projects on open sites adjacent to sources of labour and materials, though other factors can come into play. For example, projects in the South-east of England tend to be more expensive than equivalent projects in the North-east. The purpose of work type and temporal and spatial adjustment factors is to permit comparisons to be made that are based on a common value. As noted in other chapters, the difficulties of making international construction comparisons at a point in time are compounded by the need to use currency convertors based on exchange rates.

Purchasing power parities (PPPs) are an alternative to exchange rates and are applied generally to the whole economy; PPPs are also available for components of the economy. The origin, development, production and use of PPPs are described in detail elsewhere in this book. It is sufficient here to note that PPPs for a range of activities, including construction, are published by the World Bank, currently every five years or so, but the intention is for this to be more frequent in the future. Table 6.2 sets out exchange rates, GDP PPPs and PPPs for selected components of GDP, all based on US$ = 1.00, for a selection of countries from different regions, income levels and population sizes.

In higher-income countries, PPPs are typically higher than exchange rates. In middle- and lower-income countries, exchange rates tend to be higher than PPPs. This means that in rich countries, the size of economies and the volume of national output tend to be overstated and in poorer countries, understated. Because construction activity is not internationally traded, construction PPPs tend to exaggerate these tendencies, whereas machinery and equipment, which are internationally traded, tend to have PPPs that are generally similar to or higher than commercial exchange rates. Food and clothing PPPs tend to be closer to GDP PPPs in higher-income countries and closer to exchange rates in lower-income countries. There are, of course, exceptions to these general statements that may be explained by specific national features. In Singapore, a rich country, for example, construction PPPs are lower than might be expected, because there is a reliance on low-cost foreign construction workers.

Table 6.2 Exchange rates and PPPs for selected countries, 2005

PPPs demonstrate the Balassa-Samuelson effect. This states that nontraded elements of the economy – like construction – will have low relative prices in the early stages of economic development but that, over time, their price levels will tend to converge with general prices. Using commercial exchange rates for all economic activities, the effect is not necessarily picked up, and this means that the real volume of construction activity in lower-income countries tends to be understated and that the real volume in higher-income countries tends to be overstated. This is an added reason for using PPPs rather than exchange rates for comparing the volume of construction in one country to that in another. However, PPPs can only really be applied to high-level economic concepts. As PPPs are an average of equivalent baskets of goods, they do not apply to particular products or projects. The next section applies the theoretical framework of PPPs to national construction expenditures.

Analysis of the International Comparison Program 2005 data

The 2005 ICP Results (International Bank for Reconstruction and Development 2008) give nominal or commercial exchange-rate adjusted data for GDP and construction expenditure in $US for 128 countries (30 African, 23 Asian, 10 Confederation of Independent States [CIS], 45 European and OECD, 10 Latin American and Caribbean [LAC], and 10 West Asian); they also give GDP and construction expenditure for the same countries using real or PPP adjusted data. Two distinct features of the ICP data are that it provides data on expenditure, not value added, and that it is brought to a common currency basis using both commercial exchange rates and PPPs, both whole-economy PPPs for GDP and construction-specific PPPs for construction expenditure.

The ICP data set allows for an examination of the effect of the use of PPPs on construction volume and on the relationship between GDP and construction expenditure. There are, of course, concerns about the reliability of such value data and PPPs, but the ICP data set is the largest single-source data set available. It is almost certain that the ICP data is more useful than comparisons based on exchange rates.

Table 6.3 sets out GDP, construction expenditure and construction as a proportion of GDP for the 128 countries in their regional groupings using commercial exchange rates and PPPs. The numbers in brackets indicate the numbers of countries for which there is construction data. Coverage is good for Europe/OECD, Asia, the CIS and West Asia but rather less so for Africa and Latin America and the Caribbean.

Table 6.3 illustrates the effect of using PPPs compared to using exchange rates on both construction volume and construction expenditure as a proportion of GDP. Using PPPs, construction expenditure increases markedly in every region except Europe and the OECD, and construction expenditure as a proportion of GDP increases most in Asia and West Asia. There are increases in Africa and Latin America and the Caribbean, but only relatively small ones, as the increases in construction expenditure in the latter are more or less matched by the increases in GDP. The results with PPPs are almost certainly more credible interpretations of the data than those with exchange rates.

The data, of course, represents regional averages, and countries in each region will vary around the mean. Nevertheless, using averages helps to identify broad regional trends. These are in line with expectations, although it is probable that the regions with more complete data – Asia, CIS, EU/OECD and West Asia – will be more reliable than Africa or Latin America and the Caribbean, where there is less complete coverage.

The differences in Asia and West Asia are a result of overvaluation of price levels using commercial exchange rates, particularly for construction. The reasons for this distinction in the two regions are subtly different. Asia comprises mainly middle-income countries with a few high- and low- income countries, a high proportion of locally or regionally manufactured materials and access to low-cost labour. West Asia also comprises high- and middle-income countries with access to low-cost resources and low-cost foreign workers in the case of the high-income countries. The similarities in both GDP and construction exchanges rates and PPPs in the European and OECD countries are as expected in developed economies.

Table 6.3 GDP and construction expenditure in 2005 using exchange rates and PPPs

The reasons for the decrease in construction expenditure as a proportion of GDP in the CIS states are possibly more complex and deserve further study. Although both GDP and construction may be overvalued using commercial exchange rates, this appears to be more so in construction than in GDP. Using the data in Table 6.3, there is a decrease in the proportion of construction to GDP using PPPs (9.47%) compared to exchange rates (10.40%). The difference, however, is less than 1%, which may not be significant bearing in mind the data quality. Overall, using PPPs for 128 countries, GDP increases by around 29%, but construction expenditure increases by 67%, and construction expenditure as a proportion of GDP increases by 30%.

It seems, therefore, that the volume of construction expenditure using exchange rates compared to PPPs tends to be overstated in developed countries and understated in developing countries. Previous research, based on commercial exchange rates, looked at the relationship between construction output (measured as value added) per capita and GDP per capita. Figures 6.1 and 6.2 present the relationship between GDP per capita and construction expenditure per capita using exchange rates and PPPs.

The trend line in Figure 6.1 indicates that, on average, construction expenditure represents around 12% of GDP using commercial exchange rates. Figure 6.2 uses separate GDP and construction PPPs and indicates that the percentage increases to 15%. The figures reflect Turin’s view that there is a direct relationship between GDP per capita and construction expenditure per capita; they also reinforce the theory behind the Balassa-Samuelson effect. They do not, however, strongly support Bon’s proposition that the proportion of construction expenditure increases in the early stages of development but then levels off or declines in maturity.

The following figures look in more detail at the data by region, by income level and by population. Figures 6.3 to 6.6 show GDP and construction expenditure for Africa, the Americas, Asia and Europe In all four figures, the scales are the same, in $US millions using PPPs; the vertical scale is construction expenditure per capita; the horizontal scale is GDP per capita.

Figure 6.1 GDP and construction expenditure for 128 countries in 2005 in $US using commercial exchange rates

Figure 6.2 GDP and construction expenditure for 128 countries in 2005 in $US using PPPs

Figure 6.3 GDP per capita and construction expenditure per capita in 2005 $US PPPs by region: Africa

Figure 6.4 GDP per capita and construction expenditure per capita in 2005 $US PPPs by region: the Americas

Figure 6.5 GDP per capita and construction expenditure per capita in 2005 $US PPPs by region: Asia

Figure 6.6 GDP per capita and construction expenditure per capita in 2005 $US PPPs by region: Europe

In Africa and the Americas, with the exception of Canada and the US, both GDP per capita and construction expenditure per capita are low. National income appears to be a major constraint on construction expenditure. In Asia and Europe, on the other hand, greater construction expenditure appears to be associated with higher national income. The regional data generally confirms that the relationship between GDP and construction expenditure indicated in Figures 6.1 and 6.2 holds at the more detailed level.

Figures 6.7 to 6.10 group the countries by national income per capita in US$ PPPs. The 128 countries have been divided into five groups, four with 25 countries each and one with 28 (the income bands for these five groups are set out in Appendix A). In these figures, the middle group has been omitted. The labels for the vertical and horizontal scales are the same as in Figures 6.3 to 6.6, but the values vary to suit the groups.

Figure 6.7 GDP per capita and construction expenditure per capita in 2005 $US PPPs by income group: Lowest Income Group

Figure 6.8 GDP per capita and construction expenditure per capita in 2005 $US PPPs by income group: Second-Lowest Income Group

Figure 6.9 GDP per capita and construction expenditure per capita in 2005 $US PPPs by income group: Second-Highest Income Group

Figure 6.10 GDP per capita and construction expenditure per capita in 2005 $US PPPs by income group: Highest Income Group

Generally, as the variance is greater, there is much less sign of a relationship between GDP per capita and construction per capita in the lower income groups (Figures 6.7 and 6.8). Bearing in mind that previous authors have emphasized the importance of national income in predicting construction expenditure, this is surprising. The somewhat better relationship between per capita GDP and construction in the higher income groups (Figures 6.9 and 6.10) may be a result of more geographical homogeneity.

A third grouping of countries was examined – by population size – and a statistical analysis of the results of this exercise and the other two exercises is set out in Table 6.4.

Table 6.4 Statistical tests by region, national income and population size

Construction per head as a function of GDP per head	Number of countries	R2	Mean	Standard error
By geographic regions
Africa	30	0.79	72	185.1
The Americas	13	0.93	518	405.9
Asia	42	0.82	361	1,567.2
Europe	41	0.73	1,528	1,009.8
By GDP per capita
Lowest-income countries	25	0.15	115	99.6
Second lowest	25	0.23	478	543.8
Mid-range	25	0.18	950	424.2
Second highest	25	0.63	2,111	677.9
Highest-income countries	28	0.53	4,556	2,483.7
By population size
Smallest by population	25	0.80	1,989	2,073.6
Second smallest	25	0.84	949	805.2
Mid-range	25	0.83	611	504.7
Second largest	25	0.93	533	377.8
Largest by population	28	0.75	983	768.9
All countries in survey	128	0.75	440	1,301.6

The table gives the mean and standard errors by geographic region, GDP per capita and population. The regions are shown to be relatively homogenous, as are the groups by population size; while income level in a country is a poor predictor of likely expenditure on construction, regional location and population size are relatively good predictors. In all cases, the standard errors are relatively high compared to the means, leaving wide margins for error in predicting individual country values.

Conclusion

This chapter reviews and comments on some of the main issues concerning the measurement, presentation and comparison of construction activity nationally and internationally. It uses ICP 2005 data to assess the relationships between national income and construction expenditure and concludes that previous theories are, at best, not proven and that more work is needed on defining, measuring and normalizing data on construction activity.

Consistent approaches are required as to what is included in or excluded from construction activity and how variables should be measured and presented. This needs to take account of the data requirements of statisticians, policy makers, international bodies, industry, researchers and others. It is an international issue and needs to be addressed at an international level; construction is too important a sector of the economy to be measured so poorly.

The current ISIC breakdown of construction activity is not particularly helpful to any user groups. It requires distinctions to be made among residential, nonresidential and civil engineering work and between a few subtypes of civil engineering work. It does not distinguish between construction investment (new work and improvements) or construction consumption (repair and maintenance) or among publicly sponsored, privately sponsored and mixed-funded work. Detailed breakdowns of construction activity could also address the different providers of construction output: construction contractors, the informal sector, households and so forth, although it needs to be acknowledged that providers often overlap.

Gross construction output is often a more useful measure of construction activity than construction value added, although the latter has its uses. Gross output data is needed to measure the size and capacity of the sector and its growth rate; value-added measures are needed for measuring construction productivity and the share of on-site activities in gross domestic product. In developed countries, contractors’ output can be viewed as broadly representative of construction value; in less developed countries, there is more likely to be greater activity by noncontractors, mainly households.

Construction prices tend to change over time relative to prices in the wider economy both nationally and internationally. Measures of construction activity, therefore, are subject to both quantity and price changes; in these circumstances, construction values are not necessarily good proxies for construction volumes. Indeed, the patterns of national and construction-sector development reported by Turin, Bon and others may be, at least partly, an outcome of relative price trends and may not necessarily reflect real activity. Using the best data available, we cannot confirm Turin and Bon’s hypotheses. If regional location and population size are better predictors of construction output per capita than national income, it must be accepted that the basic data as it is currently defined and collected is not as helpful as it might be.

In discussing the use of exchange rates or construction PPPs, a consensus is emerging that exchange rates are poor deflators, or normalizers, in international comparisons of construction activity. Although there are problems with PPPs in general and construction PPPs in particular, there is a rationale for using PPPs in preference to exchange rates for converting construction output and GDP. Furthermore, as PPPs are refined, they can be expected to improve over time, and it is important that their development is encouraged.

Because of the issues and difficulties raised in this chapter, construction activity data – national and international – can be misinformed and misreported. Governmental statistical agencies and industry commentators require improvements in the official definitions of construction and the way data is presented. There is a need for better information on the various measures of construction activity.

In many countries, the construction sector is one of the largest components of the economy. Industrial strategies and policy decisions depend on consis- tent and reliable measures, and decisions to invest in infrastructure and the built environment remain essential for economic and social development.

China Perspectives 46 | march-april 2003 Varia

A “Race to the Bottom” Globalisation and China’s labour standards

Anita Chan

Édition électronique URL : http://journals.openedition.org/chinaperspectives/259 DOI : 10.4000/chinaperspectives.259 ISSN : 1996-4617

Éditeur Centre d'étude français sur la Chine contemporaine

Édition imprimée Date de publication : 15 avril 2003 ISSN : 2070-3449

Ce document vous est offert par Cornell University

Référence électronique Anita Chan, « A “Race to the Bottom” », China Perspectives [En ligne], 46 | march-april 2003, mis en ligne le 24 novembre 2006, consulté le 23 octobre 2020. URL : http://journals.openedition.org/ chinaperspectives/259 ; DOI : https://doi.org/10.4000/chinaperspectives.259

Ce document a été généré automatiquement le 23 octobre 2020.

http://journals.openedition.org

http://journals.openedition.org/chinaperspectives/259

A “Race to the Bottom” Globalisation and China’s labour standards

Anita Chan

1 The Chinese government rejoiced on the occasion of gaining World Trade Organisation

membership in November 2001. There was an expectation in Peking that once the

country became integrated into the world economy, it would be on the right track to

attain economic prosperity. There might be some bumps along the way: some

industries and agriculture would suffer, affecting employment, but as a whole, it was

predicted, China would gain. Employment has been a major concern in China, and the

government’s best sell was that foreign investment would increase and the labour-

intensive manufacturing sector would gain: according to one estimate, 2.8 million

additional jobs in textiles and 2.6 million jobs in the garment trade, as the constraints

of quotas for garments and textiles end 1.

2 As predicted, foreign investment has been flowing into China in the past year at the

expense of its South-East Asian neighbours and the tiger economies of Hong Kong,

Taiwan, Korea and even Japan. Hong Kong and Taiwan have been the nurturers of

Chinese export industries for more than a decade, only to discover now that some of

their own industries are being “hollowed out” 2. As one observer, William Greider,

describes it, China is “sucking away” jobs. “Globalisation”, he writes, “is entering a

fateful new stage, in which the competitive perils intensify for the low-wage developing

countries. … In the ‘race to the bottom’, China is defining the bottom” 3.

3 In other words, though employment in the low-wage industries in China may be

expanding, the wages of the workers in these industries are not rising, and for many of

them have been falling. What within the Chinese system allows it to lead in this race to

the bottom in labour standards?4 First, let us examine the empirical evidence showing

that, when compared with other developing export-oriented countries, wages in China

are very low relative to the cost of living.

Chinese wages in comparative perspective

4 There is a popular image that the global divide in competition in world trade is

between the developed and underdeveloped countries. I would like to argue here that

increasingly the competition, particularly in the labour-intensive industries, is largely

A “Race to the Bottom”

China Perspectives, 46 | march-april 2003

among countries in the developing world. The intense competition in wages among

these countries is well illustrated by Figure 1, which shows the minimum legal wage in

a number of countries around the world as of 1999.

5 This shows an enormous gap between the minimum wage in the United States and

those of developing countries in Asia and Central America—with the US minimum wage

at least twenty times higher. Faced by cheap labour abroad in this era of global

production, labour-intensive industries are basically finished in the US and in most

other high-wage nations. Such goods continue to be produced there only to a marginal

extent by using illegal immigrants5 and home workers 6, “sweated” at a wage well below

the legal minimum wage.

6 Competition in these labour-intensive industries lies instead today among the countries

of the third world. All of the minimum legal wages in the developing countries in

Figure 1 hovered around US$30-50 a month. This is equivalent in China to 240-400 yuan

a month. The legal minimum wage in Shenzhen, the Chinese city with the highest

minimum wage, was equivalent to only US$42. China has set its minimum wage

standards very low, to the point that it is even competitive with Vietnam and

Cambodia, two countries where the cost of living is lower than in China. In Mexico, El

Salvador and Nicaragua, the wage levels are slightly higher than Asian wages, but this

competitive disadvantage is largely cancelled out by the proximity of Central America

to the American market.

7 When China first instituted a minimum legal wage system in the early 1990s, it had the

good intention of protecting workers in the export sector. But soon the function of the

minimum wage changed character. It simply became the amount that employers

reported to the government they had paid their workers in the labour-intensive export

industries; they rarely pay assembly-line workers above the monthly legal minimum

wage. The great majority of the workers in this sector are migrant workers from the

countryside, and as can be seen from Figure 2, they are not sharing in the standard of

living of the urban population.

8 It can be noted from this table that the setting of a minimum wage level is extremely

decentralised in China. In most Western countries there is only one nationwide

minimum wage, but in China there are hundreds. Each city or even a district in a city

can set its own minimum wage based on a formula provided by the central government.

This takes into account the cost of living in the locality, the prevailing wage, the rate of

inflation etc., and it is adjusted each year 7. The table shows that these minimum wages

have been rising every year, but when these increases are compared with the annual

consumer price indexes for each of these cities, it becomes evident that the rises in the

minimum legal wage only kept pace with inflation. In other words, even though the

Chinese economy is rapidly developing, in real terms the minimum wages have

remained level throughout the 1990s.

9 As can be seen in Figure 2, the cities in Guangdong province and other big cities along

the coast have the highest cost of living and consequently the highest minimum wages.

Elsewhere in China, the minimum legal wages are lower, which poses a threat to the

coastal region. The results of this threat, as I shall explain, are shown in Figure 2.

10 Legally, the minimum wage in each locality is not simply supposed to keep up with

inflation. According to the international standard employed by the Chinese

government, the minimum wage of a locality should be set within the range of 40% to

60% of the average wage in that locality 8. This Figure uses 40% as the cut-off point to

A “Race to the Bottom”

China Perspectives, 46 | march-april 2003

see whether the minimum wages officially set by various localities have reached this

standard in past years. 1993 was the only year in which all of these minimum wages

fulfilled the Chinese government’s own criterion of reaching at least 40% of the average

wages. Since 1993, in most localities, the minimum legal wages did not attain even this

40% mark, in direct violation of the national directive requiring that it do so. Instead,

with only a few exceptions (see the squares shaded in grey), the general trend in

minimum wages has been one of stagnation or steady decline when compared to the

incomes of urban residents. This results, for example, in the minimum legal wage in

Peking declining from 36.7% of Peking’s average wage in 1994 to only 27% in 1999; and

in Shenzhen city from 40% in 1993 to a bit under 24% in 1999. This means the income

gap between the regular urban population and the migrant workers kept on widening

in the 1990s 9.

Figure 1: Minimum wage standards of textile, garment and footwear workers in different countries, 1999 (in USD)

Source: Based on wage figures provided by U.S. Department of Labor, Bureau of International labor Affairs, Wages, Benefits, Poverty Line, amd Meeting Worker’s Needs in the Apparel and footwear Industries in Selected Countries, Washington D.C., Departement of Labor, 2000, p.I-51.

11 Another important conclusion that can be drawn from Figure 2 is that globalisation

scarcely leads to improved wage conditions for the workers who make goods for export

compared to the populace at large. Guangzhou and Shenzhen, the two cities that have

the highest average income in the country, and the first cities in China that the central

government allowed to woo foreign investment, have the lowest minimum wage to

average wage percentage. In these two cities it did not even reach 30%. The worst of all

the nine cities is Shenzhen, the most famous model in China of a special export zone.

The percentages in these two cities have been consistently the lowest of the nine cities

since 1997, dropping to a low of 23.8% in 1999. On the other hand, in the interior of

China the minimum legal wage in the city of Chongqing, which is the least linked with

the global economy, reached that mark of 40% in 1999. These figures reflect a very

worrying trend. As a region becomes more prosperous, it violates the national

guidelines and seeks to maintain its attractiveness to foreign capital by keeping its

minimum wage level low, in order to compete with other localities in China in selling

the labour of migrant workers. The benefits of globalisation in accordance with this

competitive logic have not, and will not, trickle down to those who make the products.

12 What is even worse—and is not revealed by this table—while on paper the local

governments comply with the central government’s decrees about raising minimum

wage levels annually to adjust to the average urban wage and inflation, in reality the

A “Race to the Bottom”

China Perspectives, 46 | march-april 2003

wages of the migrant industrial workers are often considerably lower than the official

standards. For one thing, the minimum wage, set by the month, does not reveal the

illegally long hours worked by migrant workers to attain that minimum. According to a

survey I conducted in China’s footwear industry, the average workday there amounts to

about 11 hours each day, often with no days off—that is, about an 80-hour workweek.

Nor do the official statistics take into consideration the staggering amount of wages

owed but not paid to migrant workers. Of the 51,000 cases of workers’ complaints

lodged by letters and by personal visits to the Shenzhen authorities during 2001, 43%

related to unpaid wages 10. One Chinese newspaper article described the non-payment

of wages as having become a “custom” in Guangdong 11. while another described it as

an “incurable disease” 12. When the illegally long work hours and these unpaid wages

are taken into account, a sizeable proportion of the workers are making considerably

less than the minimum legal wage.

13 In short, as China develops, the benefits have not trickled down to the assembly-line

workers from largely rural backgrounds who make the exported goods. Indeed, their

situation has even worsened since the Asian financial crisis of 1997-98; the downturn

intensified competition with South-East Asian labour, which had become much cheaper

in the wake of currency devaluations. Among the reasons why China is attractive to

foreign investors, who have been rushing into China at the expense of its Asian

neighbours, is that local Chinese authorities have been able to hold down wages by

turning a blind eye to violations of China’s own labour regulations and laws. The

central government normally does not intervene.

14 Competition within China between different regions exacerbates the problem of low

wages. And the central government has intervened in a way that encourages even

lower pay. Though migrant workers’ wages in Guangdong province are very low, the

central government has been worried that Guangdong is pricing itself out of the

international market. The government therefore has started to encourage foreign

capital to move inland, to places where the pay is even lower. The owner of an

Australian toy company who sources some of her merchandise from China noted

enthusiastically to me this past year that she is now contracting toy production at

factories further north and away from big cities. The products, she said, are just as

good and cheaper.

Mexico: China’s main competitor

15 The geographic race to the bottom that operates within China also operates in an

international context. In the 1990s China’s main competitor for the American garment

market was Mexico, on the other side of the globe 13. Since the signing of the North

American Free Trade Agreement (NAFTA), Mexico has gained a large number of new

clothing factories. Today China and Mexico are competing neck and neck for the

American market, each supplying around 15% of all apparel imports to the US. Mexico

enjoys two substantial advantages over China: it is next door to the US (and hence can

meet a faster turnover rate for orders) and it enjoys an absence of quota restrictions

due to NAFTA. As a result, Asian investors who serve as the subcontractors for the

name-brand Western multinationals—and these are particularly South Koreans and

Taiwanese—became increasingly active there in the 1990s, even moving apparel

production out of Asia to Mexico. Along the US-Mexican border assembly plants called

maquiladoras have mushroomed, employing about a million migrant workers in various

A “Race to the Bottom”

China Perspectives, 46 | march-april 2003

labour-intensive industries. This number is still small compared to the 12 million in

Guangdong province alone, but it represents a 150% increase in Mexico since 1990 14.

16 As in China, expansion in employment does not mean rising wages for Mexican migrant

workers. The minimum legal wages there are almost double that of Shenzhen, and this

produces pressures on Mexican wage trends. In the manufacturing sector, real wages

dropped by 20% during the 1990s. According to the International Labour Organisation’s

estimate, the migrant workers’ wages in Mexico’s apparel industry shed 28% of their

purchasing power in the period between 1994 and 1999.

17 Despite this drop in real wages, the maquiladoras recently have been losing ground. As

trade barriers continue to fall due to the WTO, the middleman firms from Taiwan and

South Korea have begun shifting production back to Asia, particularly China. The

numbers of maquiladoras swelled from 120 in the 1970s to 3,700 in 2000, but have

dropped by 500 factories since then 15. Pressures are therefore tightening on Mexican

enterprises to more vigorously compete with China’s long working hours and bargain-

basement wages. This also explains why Mexico was the last country to sign a trade

agreement with China, delaying China’s entry into the WTO. Mexico knew that when

the trade barriers are removed, it would have much to lose. But the international

pressure was too great for Mexico to stand its ground.

China’s pass system—the drive to the bottom

18 There are numerous reasons why Chinese wages can be kept so competitive compared

to other countries. First, it has an almost inexhaustible supply of cheap labour from the

countryside. Second, the decentralisation and deregulation in wage-setting under

China’s economic reforms has enabled local governments to turn a blind eye to labour

exploitation. Third, there is no autonomous union movement in sight in the foreseeable

future to fight to preserve wage levels, and the Chinese government is intent on

making sure that none is allowed to arise.

19 There is also a fourth fundamental reason—China’s so-called hukou system, or

household registration system, which prevents an uncontrolled rural-to-urban influx of

population. This works in similar ways to the pass system under South Africa’s former

system of apartheid. To be sure, the two systems differ markedly from each other in

origin and ideology. The South African pass system was intertwined with a history of

racism, colonialism and the development of South African capitalism, all of which

favoured the control of movement of African people to provide greater political

security and enhanced efficiency in the use of black labour. The ideology on which the

system was based was white supremacy, and apartheid was the cornerstone of the

state-building project of the South African white ruling elite after World War II.

20 The hukou system in China has a very different history. It was established after the

Communist Party came to power in 1949. To ensure that the planned economy met the

basic needs of the urban population, a rationing system was instituted in the 1950s,

which in turn required the registration of people. As ration coupons could only be used

in the locality where they were issued, this automatically restricted the geographical

mobility of all people, not just peasants 16. To reside in a different locality, one needed a

special temporary certificate.

21 This system of passes has been retained to the present day. The constraints this system

places on the geographical mobility of migrant workers, and the means by which this

can drive down wages and other labour standards today, is what is similar to what

prevailed in South Africa. In China it is there by default. It was in place before it had its

A “Race to the Bottom”

China Perspectives, 46 | march-april 2003

present function. It simply continued to be used, when it was found to work well under

a greatly changed economic system.

22 In this transitional period from socialism to capitalism, the temporary work permits

required under the hukou system act like sluice gates controlling the influx of labour

into urban centres. The hukou system helps to regulate the flow, letting in more

labourers when needed and driving them out when their number exceeds demand, or

when the number stretches local facilities to the limit. When workers lose their

working ability due to industrial accidents, or when they have become too “old” by the

age of about thirty to keep up with the break-neck work intensity, the pass system

enables cities to ship them back to the countryside—because without a job a migrant

has no right to stay in an urban area. This kind of labour flexibility cannot be as easily

imposed on the local urban population.

23 As more and more state factory workers were laid off in the nineties, some urban

governments placed tighter restrictions on job opportunities for the immigrants. One

category of jobs after another, especially in the service industries, was reserved only

for local residents. In 1993, 40,000 rural migrants in Shanghai were detained and

deported; but as the cities clamped down on opportunities, this rose in 1996 to 80,000,

and in 1997, 100,000 17. There are, however, regional differences in how strictly local

governments enforce the hukou system. In areas where the supply of labour does not

exceed demand by a large margin, the police and local government are considerably

more relaxed about the presence of migrants. This is the case in Chengdu in Sichuan

province and Fuzhou, Fujian 18.

Figure 2: Proportion of minimum wages to employees average wages in cities of China (1993-2000)

Sources: average wages are from various statistical yearbooks and minimum wages are from various sources, including newspapers and labor bureaus.

A “Race to the Bottom”

China Perspectives, 46 | march-april 2003

24 For local governments, allowing migrants to come in from the countryside can be

lucrative. Migrant workers generate local tax revenues by attracting companies that

want cheap labour, but because of the hukou system the local government has no

obligation to pay anything for the welfare of these temporary sojourners. They are not

eligible for any of the medical, housing or unemployment benefits available to the local

urban populace. Nor are the workers from the countryside allowed by China’s pass

system to bring their families with them, and thus the urban government has no

additional educational expenses to meet either.

25 Despite this system of permits, the enormous bureaucratic edifice that was erected to

control the influx of migrants has not been able to stem the flow, just as had occurred

in South Africa. It is impossible to estimate the exact number of Chinese peasants

surging out of poor regions in search of jobs, but a range of between 50 to 80 million is

often cited. In the week immediately after the Chinese New Year, when migrant

workers who have gone home for the festival return to the cities, bringing with them

relatives and friends, the effect on transport is dramatic. For instance, in a matter of

days Guangzhou, the largest city in South China, suddenly has to handle several million

migrant workers descending upon it in trains and buses. In early 2002, before the

Chinese New Year, the Guangdong provincial government, in the hope of dampening

this vast simultaneous inflow, announced that factories should not recruit new

migrants at that time of year; but 5.2 million migrants nonetheless poured in after the

New Year, a quarter of a million more than the year before 19.

26 This large volume of people looking for low-end jobs drives down wages and working

conditions and allows these migrants to be exploited by employers, who can pay them

the lowest possible wages. New arrivals, in particular, desperate to recoup the amount

they have invested in transport expenses and in applying for the array of necessary

documents and certificates before leaving home, will take any job available.

27 Here is the case of one migrant reported in a Chinese newspaper. The young migrant

was informed by a friend that if he went to Shenzhen he would find a job. But he was

advised that before he left he had to apply for a number of documents. These included a

“border region pass” (at 120 yuan, taking six months), a personal identity card (another

80 yuan, taking one month), an unmarried status certificate (60 yuan, valid for one

year) and a certificate to prove that he was not born out of quota (45 yuan, valid for one

year), all of these totalling 305 yuan 20. To put this into perspective, the minimum wage

in Shenzhen in 2000 was 547 yuan for a full month’s work, and this young man would

be lucky if he could enter a factory that would pay him as much as that minimum wage.

28 On arrival in Shenzhen, armed with all these documents, he thought he could become a

“legal” migrant worker and could begin working without a problem. But the factory

demanded 300 yuan as a deposit before it would give him the job. He then had to spend

40 yuan for a work permit, and another 300 yuan for a temporary residence permit. It

short, on arrival at his destination he had to spend another 640 yuan. In all, without

including transport costs, he had to spend almost twice as much as the monthly wage.

Most new migrants therefore are usually in debt after they first arrive in a city.

29 According to official statistics, each of the three or four million migrants in the

Shenzhen Economic Zone on average spends 600 yuan a year on certificates 21. Migrant

workers have to carry these documents with them at all times or, if caught without

them, may be held in detention. To possess all of the necessary certificates, one needs

A “Race to the Bottom”

China Perspectives, 46 | march-april 2003

to have proof of a job, and so there is a nervous race to secure one. The deposit that this

particular migrant needed to pay to the employer is symptomatic of the desperate

situation of most migrant workers. Paying a substantial deposit has become a common

practice at the foreign-funded factories. At first sight, the practice seems paradoxical.

Instead of the employer paying a worker for the work he or she performs, the worker

first has to provide a payment to the employers as surety for the job. The deposit

obliges the worker to remain at the factory, or he or she forfeits it. To all intents and

purposes the worker is bonded labour 22.

30 Another practice used by many unscrupulous employers is not to pay a portion of the

wages every month, promising to pay the withheld portion at the end of the year. In

this situation, the longer a worker has worked, the more money he or she is owed by

the employer, and the more difficult it is for the worker to leave. This leaves the

worker vulnerable, scared to forfeit all of these unpaid wages when facing poor

treatment at the hands of managers.

31 Finally, and perhaps most effective of all, it is a widespread practice among employers

to take away the migrant workers’ documents. Without these, under China’s system of

permits, the workers could not look for another job even when the working conditions

in a factory are intolerable and they desperately want to quit.

32 Workers’ dormitories, usually located within the factory compound, extend

management control over workers’ lives beyond work hours. Movement into and out of

the factory compound can be monitored and controlled. Disciplining workers is easier

because there is near-total control over them. Especially in the factories in China

managed by Taiwanese and Koreans, where the discipline is so strict that the

management style can be described as militaristic. In some of the bigger factories that I

have visited, workers are even marched to and from meals and to and from dormitories

in tight military-style squads 23.

33 With migrant workers so controlled and cowed, physical abuse has become pervasive in

some of the factories owned and managed by Taiwanese, Koreans and Hong Kong

Chinese, and acute occupational health and safety problems are also commonplace. A

startlingly high incidence of severed limbs and fingers has been recorded. In Shenzhen

city alone, there were over 10,000 certified cases in 1999 among a migrant population of

some three to four million 24.

34 The system of permits needs an enforcement agent—in this case the police. Under the

hukou system, much as in apartheid-era South Africa, detention by the police if caught

without the necessary papers is an inherent part of the system. Their behaviour

towards migrant workers has become associated with corruption and abuses of power.

Detention is associated not only with fines and deportation from the city, but also with

mistreatment, physical violence and forced bribery. With so many migrants pouring in,

the arrests are essentially random. In much of Guangdong, people who seem to be of

rural origin are simply pulled off the streets and roughed up, sometimes for no

particular reason. Among ten young migrants whom I interviewed recently in

Shenzhen, five said they have been picked up by the police within the several months

they had been there, a few of them more than once; and nine out of the ten knew of a

friend or relative who had been detained.

35 Many migrants do not have all the right papers on them because they are not aware of

what they need. Others are too poor to buy them all. But oftentimes, through no fault

of their own, their documents are kept locked up by their employer; or they have left a

A “Race to the Bottom”

China Perspectives, 46 | march-april 2003

factory without being able to get their documents back because the employer did not

want them to leave. As a result of the latter, borrowing documents from friends and

purchasing forged ones off the street have become very common 25, as it was in South

Africa. According to one survey conducted by a government labour bureau in

Guangdong, 80% of foreign employers openly admitted that they did not care whether

the documents were fake or not, as this did not affect production 26, The infringement

of regulations being so widespread implies tacit approval has been granted by the local

authorities and police.

36 Yet this does not stop the police from detaining migrants arbitrarily. Police stations

consider this a profitable business, because bail, fines and forced bribes, also imposed

arbitrarily, can amount to several hundred yuan. Even neighbourhood committees that

have no power of detention get into the act. Some have been detaining migrant

workers and charging bail 27. The practice has become so out of hand in the past couple

of years that the central government in January 2002 issued a decree reducing the fee

for a temporary resident permit to 5 yuan a year nationwide, to enable migrants to

afford one and thus avoid detention. And in March 2002 the Guangdong provincial

government passed regulations emphasising that the detention of “vagabonds” should

be restricted to beggars and not applied to migrant workers who do not have the right

papers on them.28 But rather than obediently comply with the regulations, the

provincial police responded by declaring that they have done a good job in sheltering

beggars and vagabonds; and reaffirmed the necessity of rigorously implementing the

pass system, without mentioning that they were continuing to detain and abuse large

numbers of migrant workers 29. In a few months local-level governments and the police

came up with new fees to make up for the loss in revenues and private incomes.

Proclamations of new policies do not mean elimination of the hukou system. Those who

gain from the system are not going to desist so easily.

37 As can be seen, the Chinese hukou system and the pass system under apartheid in South

Africa generated quite similar outcomes. They produced a large, vulnerable underclass

living in constant insecurity, accompanied by daily discrimination, repression,

hardship and denial of their human dignity.

38 In light of these circumstances, it becomes possible to perceive how the Chinese hukou

system can keep wages down more easily than in Mexico. As already noted, in Mexico

the workers who produce for export are, as in China, largely migrants from the

countryside, and the majority similarly are female. But there is a major difference.

Almost all of the Chinese female migrant workers are single women in their late teens

or early twenties who, because of the household registration system, cannot bring their

families with them 30. Many factories make sure that only single women are recruited

by asking to see their officially issued identity certificates, which in keeping with the

Chinese state’s strict family-planning policy require that the marital and family

planning status of each woman is listed. Since the workers are poor single women

living in dormitories, management only needs to pay them enough for their individual

survival 31.

39 In Mexico the context is different. While most of the women workers in the

maquiladoras are migrants from poorer regions, many of them have come with their

families, since there is no pass system, and quite a number are single mothers. Very

often these women workers are the sole bread-winners. Since they live with their

families, a part of their waking hours has to be spent on “unproductive” chores (from

A “Race to the Bottom”

China Perspectives, 46 | march-april 2003

management’s vantage point): in commuting, in household tasks such as cooking,

taking care of the old and the young. No matter how ruthless, there is a limit to the

amount of overtime work that management can squeeze out of these Mexican workers

—fewer hours than with the young single women in dormitories in China.

40 There are also legal pressures in Mexico to pay workers a bit more so that they can

provide for part of their families’ livelihood. The Mexican Labour Law states that “The

minimum wage must be sufficient to satisfy the normal necessities of the head of the

family in the material, social and cultural order, and to provide for the obligatory

education of his children” 32. This article echoes Article 25 of the Universal Declaration

of Human Rights of the United Nations. It is similar to the concept of a living wage.

Although in reality the minimum wage levels set in Mexico are far below the standard

stipulated in the law, the notion of a wage that can provide for a family exists.

41 No such concept of a “living wage” exists in the Chinese discourse on wages, nor is it

stipulated in the Chinese Labour Law. Even the migrant workers’ protests do not centre

on how low the wages are, but instead revolve mainly around the issue of unpaid

wages. Only when workers have not been paid for several months, when the situation

becomes desperate, do the workers begin to protest. Local governments in Guangdong

province occasionally launch campaigns, especially just before Chinese New Year, to

collect unpaid wages or unpaid payments for overtime work 33. But these brief

campaigns to collect money owed to workers are only the tip of the iceberg 34. The

expectation of adequate labour standards is much lower in China than in Mexico.

42 In China, the official trade union is an arm of the Party-state. It has little autonomous

space to protect labour rights. In fact, because of the massive influx of foreign capital

and the rapid rate of industrialisation, the trade union’s efforts, even when undertaken

with good will, face a near-impossible task. A parallel situation used to exist in Mexico,

where the trade unions were affiliated with the government. But since the defeat in the

last election of the former ruling party, the Institutional Revolutionary Party, there is

now a fledgling independent trade union movement. This is challenging the old unions’

authoritarian and pro-management practices with the help of North American trade

unions and the anti-sweatshop movement. This anti-sweatshop movement is composed

of trade unions, NGOs, labour advocates, university students, human rights groups and

church groups. It grew rapidly in the 1990s and has become a force that can no longer

be ignored either by multinational corporations or governments.

43 Ripples of the anti-sweatshop movement have spread to Asia, and China. The concept

of corporate social responsibility is just beginning to circulate in China. The factories

there that are run by contactors from Taiwan, Hong Kong and South Korea are now

feeling pressure from the Western multinationals that they supply. The multinationals

have nervously begun to urge them to upgrade their labour conditions. But the

factories being monitored by the multinationals are just a small minority of the better

and bigger factories, among the thousands and thousands of factories that subcontract

production. The potential role of the state becomes important here. A willingness by

the Chinese government to enforce its own laws would be much more effective than

sporadic monitoring. But the Chinese government has not yet awakened to the growing

pressures emanating from the West for improved labour standards in the export

industries.

44 A few weak rays of hope have emerged in the past two years elsewhere in Asia. Three

countries, Cambodia, Vietnam and more recently Thailand, have expressed interest in

A “Race to the Bottom”

China Perspectives, 46 | march-april 2003

improving labour standards within their borders to attract foreign capital. Cambodia

has signed a bilateral agreement with Washington that accepts linking labour

standards to trade and has agreed to let the ILO monitor progress. The Vietnamese

government has publicly encouraged its factories to try to raise standards in order to

acquire the certifications issued by an American-based organisation that verifies labour

standards for Western corporations. The Thai government is currently engaged in talks

with this organisation about operating training programmes to upgrade labour

standards in Thailand. That is to say, three Asian countries are now taking a new

direction in their industrial development strategies. They are trying to attract foreign

investment and trade by raising labour standards instead of depressing them.

45 It is not clear whether Peking is aware of this new strategy adopted by its Asian

neighbours. But so far, China has not shown any signs of changing its policy of low

labour standards. The government has not publicly addressed the issue of corporate

social responsibility, unlike the Vietnamese government. The Guangdong provincial

government has tried to alleviate some of the most blatant abuses, but no fundamental

change in policy has been adopted. Let me quote here the director of the human rights

programme in Asia for Reebok:

Who enforces Chinese labour law? Nobody. If it were enforced China would be a

much better place for millions of people to work in. But it is ignored more than in

any other country I work in 35.

46 There have been a few reforms of the Chinese hukou system, but only to allow

successful people with considerable money or education to apply for an urban hukou.

The controls over the unskilled migrant workers who work on the production lines and

construction sites, imposed by the pass system, remain the same. And the police seem

adamantly against any changes. The hukou pass system seems likely to remain in place

for the foreseeable future, and China will continue to dominate the world’s export

market, to the point that the new initiatives taken by Vietnam, Cambodia and Thailand

may possibly collapse under the weight of Chinese competition.

NOTES

1. Gongren ribao (Worker’s daily), November 19th 2001.

2. George Wehrfritz and Mahlon Meyer, with Hideko Takayama, “Trapped in a Chinese

Box”, Newsweek, February 18th 2002.

3. William Greider, “A New Giant Sucking Sound”, The Nation, December 31st 2001.

4. Labour standards, a term that was once used almost exclusively within labour and

government circles, has been making inroads into ordinary conversations, in political

speeches and in the mainstream press. It refers to wages, work hours, shopfloor

conditions, work intensity, and occupational safety and health hazards.

5. Kwong, Peter. 1998. Forbidden Workers: Illegal Chinese Immigrants and American Labor,

New York, New Press.

A “Race to the Bottom”

China Perspectives, 46 | march-april 2003

6. Ross, J.S. Robert, forthcoming. “Declining Labor Standards in the North American

Apparel Industry”, Chapter 10 in G. Kohler and E. J. Chaves, Globalization: Critical

Perspectives, New York, Nova Science.

7. “Wages: After the Labor Law”, China News Analysis, no. 1544, October 1st 1995.

8. See Notice Concerning Regulations of Enterprises’ Minimum Wages, issued by the Chinese

Ministry of Labour on November 24th 1993.

9. This trend parallels China’s increasing Gini coefficient in the 90s -- 0.42 in 1996 to

0.458 in 2000 (Zhongguo gaige bao (Chinese Reform News), September 11th 2001). A gini

coefficient of 0.4 is considered high internationally.

10. Zhongguo laodong baozhang bao (Chinese labour social insurance daily), February 19th

2002.

11. Gongren ribao, May 9th 2001.

12. Zhongguo laodong baozhang bao, February 19th 2002.

13. The following two paragraphs draw from information in a paper I have co-

authored: Robert J. S. Ross and Anita Chan, “From North-North to South-South: The

True Face of Global Competition”, Foreign Affairs, Vol. 81, No. 5 (September 2002), pp.

8-13.

14. The Economist, July 7th 2001, pp. 27-30.

15. Washington Post Foreign Service, June 20th 2002.

16. Tiejun Cheng and Mark Selden, “The Origins and Social Consequences of China’s

Hukou System”, China Quarterly, no. 139, September 1994, pp. 644-668.

17. Jonathan Unger, The Transformation of Rural China, Armonk, M.E. Sharpe, 2002, p.

128.

18. Observations during field work in these two cities in September and October 2002.

19. Clara Li “Migrant Workers out in Cold”, South China Morning Post, February 26th

2002.

20. Yangcheng wanbao (Guangzhou Evening News), March 30th 2001.

21. Ibid.

22. Anita Chan “Globalization, China’s Free (Read Bonded) Labour Market, and the Chinese Trade Union”, Asia Pacific Business Review, Vol. 6, No. 3 & 4, Spring/Summer

2000, pp. 260–81.

23. Anita Chan, “Regimented Workers in China’s Free Labour Market”, China

Perspectives, No. 9 (January-February 1997), pp. 12-16.

24. Gongren ribao, March 31st 1999.

25. Anita Chan, “The Culture of Survival: Lives of Migrant Workers Through the Prism

of Private Letters”, in Perry Link, Richard Madsen and Paul Pickowicz, eds., Popular

China: Unofficial Culture in a Globalizing Society, Boulder, Rowman & Littlefield, 2002,

pp. 163-188.

26. Nanfang ribao (Southern Daily, Guangzhou), March 29th 2002.

27. Nanfang gongbao, Dagongren (Southern Workers’ News, Migrants), April 5th 2002.

28. Nanfang ribao, March 29th 2002.

29. Yangcheng wanbao (Guangzhou Evening News), April 22nd 2002.

30. By the author, 1998.

31. By the author, 2002.

32. The Federal Labor Law of Mexico, Chapter VI, Article 90 (1970), amended in 1998.

1996 Translation Glenn McBride, www.mexicanlaws.com.

33. Nanfang ribao, March 13th 2000; Zhonghua gongshang shibao (Chinese Commercial

News), July 5th 1994.

A “Race to the Bottom”

China Perspectives, 46 | march-april 2003

34. Based on an interview in Guangzhou City with a reporter of a Guangdong provincial

newspaper in November 2001.

35. Associated Press Newswire (Hong Kong), May 29th 2002.

A “Race to the Bottom”

China Perspectives, 46 | march-april 2003

A “Race to the Bottom”

blog6

Get help from top-rated tutors in any subject.