Chinese Economic Development

Chris Bramall

• Publisher: Routledge
• Year: 2008
• Print ISBN: 9780415373487
• eBook ISBN: 9781134190508
• Topic: Business
• Subtopic: Business General
• Index: Business

This book outlines and analyzes the economic development of China between 1949 and 2007. Rather than being narrowly economic, the book addresses many of the broader aspects of development, including literacy, morality, demographics and the environment.

The distinctive features of this book are its sweep and that it does not shy away from controversial issues. For example, there is no question that aspects of Maoism were disastrous but Bramall argues that there was another side to the whole programme. More recently, the current system of government has presided over three decades of very rapid economic growth. However, the author shows that this growth has come at a price. Bramall makes it clear that unless radical change takes place, Chinese growth will not be sustainable.

This large, comprehensive text is relevant to all those studying the economic history of China as well as its contemporary economy. It is also useful more generally for students and researchers in the fields of international and development economics.

Part 1
Starting points

1 Measuring development

Before considering China’s developmental record over the last half century, we need to answer a fundamental question: what do we mean by a successful developmental record? The answer is contested terrain, not just in the case of China but for all countries. However, there are two main issues: which indicator should we use, and what should a country’s record be compared with?

As far as indicators are concerned, the approach favoured by economists emphasizes measures of affluence such as GDP per head. It is therefore axiomatic in some circles that Maoist China ‘failed’ because its GDP per person grew only slowly, whereas China since 1978 has ‘succeeded’ because there has been very rapid growth. In the wider development community, however, the approach pioneered by Amartya Sen—that policy should aim to expand the capabilities or freedom of the population, and that an increase in the supply of goods is only one factor in that calculus—has won much favour. Again this is of great relevance to assessing China, because its record under Mao was much better in terms of life expectancy than in terms of GDP per head. More recently, the focus has shifted to more subjective measures of welfare, an approach in which persons are asked about their perceived level of happiness. The relevant question to be asked here is therefore whether the happiness of China’s people has risen since the 1980s. This approach is thus unusual, because it relies on direct measures of development. Earlier approaches, whether focused on GDP per head (the opulence approach) or on life expectancy (the capability approach), sought to measure development by looking at proxy indicators, but one can see the attraction (at least in principle) of looking at well-being in a more direct way.

The second issue in assessing development centres on the appropriate comparison to make. If we believe that China’s performance is best assessed using (say) life expectancy, how do we decide whether China has done well or badly in any given time period? This subject is much less explored in development textbooks than the subject of development indicators, and this is undoubtedly because the issue is especially difficult. One answer is to rely on temporal comparisons; how has a country performed in a given period relative to its performance during a previous time period? For example, we can assess the developmental record of Maoist China by comparing it against China’s record in the 1930s. These sorts of historical comparisons are attractive because they normalize for country-specific factors. However, the historical approach runs headlong into the problem of how to normalize for differences in technology or for the performance of the world economy. This is one reason why many economists prefer to rely more on international comparisons. We might, for example, compare China’s record during the 1990s with that of another large poor country such as India during the same decade. However, this approach is also problematic, because of differences in country size, differences in the degree of religious or ethnic fragmentation, differences in the availability of natural resources, etc. A third possibility is to compare actual performance against potential; we might deem a country successful if it fulfilled its potential, even if that potential was low. Self-evidently, however, the biggest problem with this sort of methodology is that it is hard to construct a plausible counterfactual against which actual performance can be compared. A fourth issue is making comparisons is that of how much weight to give to short-run fluctuations as opposed to long-run trends, and how much significance should be attached to policies which expand the potential of the economy even if their short-run effects are small (or perhaps even negative). This is one of the key issues when it comes to assessing Maoism. Even if we conclude that China’s record was poor between 1949 and 1976 (for example, many believe that actual consumption levels fell short of potential), can we still argue that Maoism was a developmental success because it laid the foundations for the growth of the 1980s and 1990s?

Development indicators

Much of the discussion on how to measure development during the last twenty years has focused on the respective merits of opulence and capability indicators. The opulence approach, first developed during the Second World War as a way of making systematic international comparisons, has focused on trends in GDP (or GNI) per capita. The second, or human development approach is often associated with the work of Amartya Sen (1983, 1985, 1999; Hawthorn 1987), Mahbub ul Haq (1995) and the United Nations Development Programme (UNDP) (1990). In the human development approach, close attention is paid to trends in mortality and education.¹

In one sense, this debate about opulence versus capability is philosophical, because in most instances capability and opulence indicators tell the same sort of story: most obviously, the OECD countries are developed in terms of both opulence and capability. However, there may be a real policy dilemma here for the governments of developing countries if opulence and capability measures diverge. For the governments of developing countries, the issue is about which strategy to follow; spending on basic health care may lead to big improvements in longevity, but if this delays industrialization is it a wise strategy? The only sensible way to resolve these dilemmas is via a democratic form of government. The populations of developing countries must decide for themselves which strategy to pursue.

Donors with limited aid budgets face in some respects an even more difficult decision in that they have to decide which country is most needy. At the most basic level, this is about whether opulence is a better measure of development than capability. Is a country with a high level of life expectancy but a low level of per capita GDP more in need than a country with low life expectancy but comparatively high GDP per head?²

The opulence approach

The most commonly used opulence criterion is GDP per capita. Unless a country is a large net recipient of income from abroad (e.g. property income; migrant remittances), GNI and GDP measures produce very similar results and therefore it matters little which one is used for measuring development. That is certainly so for China. I therefore use the two terms interchangeably.

GDP can be measured in one of three ways: as the sum of all expenditure (investment, consumption, net exports and government spending); as the sum of domestic incomes (wages, profits, rents and dividends); and as the sum of the value of all types of goods produced in the economy. In principle, these three should give the same figure, but in practice the data on expenditure tend to be more reliable than those on income and production; the former are distorted by tax evasion, and the latter by under-reporting of production in family businesses.³ GDP per head provides an attractive way to assess development. It is value-free in the sense that all types of marketed goods and services are included. International comparisons can easily be made by converting the value of national GDPs into a common currency (usually the US dollar) using current exchange rates. Furthermore, data on GDP for most countries are now available. Calculated originally by national statistical offices, these figures have been brought together by the World Bank going back to 1960, and these have been published in the Bank’s annual World Development Report (more detailed data are published in its World Development Indicators). More recently, thanks to the work of Angus Maddison, consistent data on GDP per head have been compiled for most countries going back to 1500 (Maddison 2006b: 294). We therefore now have a vast amount of data which allow far more systematic analysis of development trends going back all the way to the Renaissance.

Nevertheless, GDP is problematic as a measure of development even on its own terms.⁴ For one thing, there are big measurement difficulties. Most of the pre-1945 data are not very reliable by any standard, partly because of data collection problems, and partly because national income accounting was not developed until after the Second World War. The work done by Maddison is therefore (well-informed) speculation for many countries

There are also big conceptual problems. The GDP approach assumes that an extra dollar earned by a rich woman increases national well-being by as much as an extra dollar earned by a poor man, an assumption with which most would take issue. Second, a range of outputs which typically do not have a well-defined market price (externalities) are difficult to incorporate into GDP: environmental damage is one example. Some non-marketed output is usually included in estimates of GDP. For example, the value of farm products which are produced and consumed by farm households without entering the market is estimated using the market price multiplied by the volume of production (itself calculated as the yield of rice in any given year multiplied by the area sown). The main problem from a national accounting point of view is how to deal with the value of housework, leisure and urban disamenities. As all three are large, the results are very sensitive to the method of valuation.⁵ Third, many would disagree with the notion that certain types of goods and services (such as crime, drugs, telephone marketing and advertising) are as valuable to society as education or health care products. For this reason, the Soviet Union used a much narrower measure of opulence (material product) which excluded a range of ‘social bads’ to assess its developmental progress. China adopted the same approach after the Revolution, and this continued until the early 1990s, when it reverted to using the GDP approach.⁶ The difference between the two measures is quite considerable. In 1978, for example, Chinese GDP (calculated retrospectively) was 359 billion yuan whereas its net material product or national income (guomin shouru) was only 301 billion yuan (SSB 1990a: 4–5).⁷

At least as problematic is the question of which set of relative prices to use to value output. In practice GDP is measured using a fixed set of market prices, or constant prices. For example, we might value Chinese GDP in 1952 by using data on the volume of goods produced in 1952 multiplied by the prices of 2007, and compare it with the value of GDP in 2007 (calculated again using 2007 prices but this time applied to the volume of goods produced in 2007). This approach thus factors out the impact of inflation, because the same prices are used to value output in both 1952 and 2007. Although this approach is elegant and makes sense, it does presuppose that the relative prices which existed in 2007 provide a true measure of the value of goods. This is by no means obviously correct. Even where markets do exist, they invariably function badly because of a range of imperfections (such as monopoly power and imperfect information); only in economics textbooks are markets in equilibrium. Furthermore, relative prices can fluctuate dramatically from year to year. For example, the price of oil was quite low in the late 1980s compared with other goods; by contrast, it was much higher in 2006. Should we therefore use the relative prices of 2006 or those of the late 1980s to value output? Unfortunately there is no easy answer to this question, and the set of prices used can make a considerable difference to estimates of GDP growth.

Comparisons of GDP across countries are also difficult. Aside from a lack of consistency in the way production is measured, the main problem is that the exchange rates used to convert GDP from national currencies into dollars are heavily distorted by capital movements, speculation and by barriers to trade in goods and services. As a result, actual exchange rates are not equilibrium rates. But the problem is not merely one of market imperfections. Rather, price distortions arise primarily because of the existence of non-tradable goods. The prices of labour-intensive non-tradables in poor countries are typically much lower than in rich countries; labour productivity differences between countries in these sectors are relatively small (because little capital is employed), yet wage differentials are very large because national wages are driven by trends in the high wage—high productivity sector. In principle, the exchange rate should adjust to reflect these price differences, but the exchange rate cannot adjust to reflect the implicit competitive advantage of poor countries if the goods are not traded.⁸

A way to circumvent this problem is in effect to revalue production in all countries using world or international prices. This is the purchasing power parity (PPP) approach.⁹ It makes a very big difference to international comparisons of GDP per head, as Table 1.1 shows. The GDP per capita of poor countries is increased, whereas that of relatively rich but high-price countries—Switzerland is a good example—is cut.

These sorts of adjustments are not enough to make either China or Ethiopia rich, but they do lead to a significant narrowing of the gap between rich and poor countries. In China’s case, per capita GDP increases by a factor of nearly 2.5. The adjustment also has the effect of making China the second-largest economy in the world in absolute terms, which emphasizes the fact that China is increasingly becoming a global rival to the USA.

Nevertheless, the PPP approach remains an inexact science, because of the pricing problem previously discussed. We can all agree that the prices used by poor countries should not be their own prices, and intuitively the case for using US prices to value production in all countries makes sense because the US economy is the largest in the world. However, significant problems remain. World market prices are themselves heavily distorted, and the same is true of those which prevail across America. In any case, it is not obvious that US consumer preferences—which play a key role in determining US prices—should be used to value production in China or indeed in any other country.

The capability approach

The main criticism of the opulence approach is that it takes a materialistic view of living standards, focusing exclusively on the ownership and consumption of goods. Much of the criticism derives from the work of Amartya Sen (Nobel Laureate in 1998). Sen’s argument is that goods are only a means towards achieving happiness, rather than an end in themselves. We also need to consider other factors, of which probably the most important are health and education. In other words, a person cannot make full use of his or her ownership of commodities without good health and without an education (for example, participation in a democracy depends on education, as does avoidance of illness). From this, Sen infers that a much better indicator of living standards is life expectancy at birth (measured in years). Life expectancy is obviously influenced by the ownership and consumption of goods but it also depends upon health and upon education. In some sense, therefore, life expectancy is a broader measure of well-being than commodity ownership and consumption of goods because it is influenced by other factors; for an especially clear discussion of the case for using mortality data, see Sen (1998). Of course a person may choose not to live as long as is possible, perhaps by following an unhealthy lifestyle. Accordingly, our real interest is in what Sen calls capabilities (the range of choices available) rather than what he calls functionings (the actual choices made). More precisely:

Table 1.1 Estimates of GDP per person using national and US prices in 2005 ($US)

A person’s ‘capability’ refers to the alternative combinations of functionings that are feasible for her to achieve… While the combination of a person’s functionings reflects her actual achievements, the capability set represents the freedom to achieve: the alternative functioning combinations from which this person can choose. (Sen 1999: 75; original emphasis)

Thus for Sen the aim of development should be to expand the freedom to achieve, or what he calls the capability set. We should focus much more on ends (capabilities) than on means (the possession of goods) in measuring development. Sen and others have also argued that capability measures tend to do a much better job of incorporating information on inequalities in well-being. It is rare to find examples of countries where high average life expectancy coincides with high levels of human poverty, but extensive income poverty is commonplace in countries with medium or high average GDP per capita; oil-exporting countries are often in this category. In other words, because the distribution of life expectancy is more equal than that of income, life expectancy provides a broader measure of development than GDP per head.

In practice, of course, we cannot easily measure capabilities. We have to rely on data on functionings (actual achievements or choices made) instead. Nevertheless, for societies as a whole, we would expect to see a close correspondence between the two. Measured life expectancy (which is calculated from actual age-specific mortality rates) is likely to correspond to the capability for life. We can thus evaluate the record of developing countries by looking at the extent to which they have been able to improve average life expectancy at birth over time.

To be sure, the record in terms of opulence and ...

Table of contents

• Cover Page
• Title Page
• Copyright Page
• List of boxes
• List of figures
• List of tables
• Glossary
• Introduction
• PART 1 Starting points
• PART 2 The transition to socialism, 1949–1963
• PART 3 The late Maoist era, 1963–1978
• PART 4 Market socialism, 1978–1996
• PART 5 The transition to capitalism, 1996–2007
• Bibliography