Economics

With a new afterword by the author

PRANAB BARDHAN

ASSESSING THE ECONOMIC

RISEOFCHINAAND INDIA

AWAKENING GIANTS,

FEET OF CLAY

PRINCETON UNIVERSITY PRESS PRINCETON ANO OXFORD

Chapter?

Poverty and Inequality: How Is the Growth Shared?

In both China and India there is a general sense that much of the benefit of economic growth has been concentrated, that large masses of people are still very poor, and that the process of global integration of these two countries, for all the hype in the financial press, has left many people behind. In this chapter we discuss this issue with the help of the available empirical data and come to conclusions that are somewhat more ambiguous than are suggested in the popular discussion.

First, take the case of absolute poverty, the scourge of both countries for at least the past two centuries. There is a large literature on the intricacies' of poverty measurement: how to define a poverty line, if at all possible, and how it is related to local nutritional norms or the consumption basket of the poor; a head count of people below such a line does not capture the intensity of poverty among the different poor groups; there are large and increasing gaps between estimates of consumption expenditure derived from alternative sources such as national income accounts and household survey data (both of which have problems and are not quite comparable); what is an appropriate price index to use to deflate the monetary expenditure of the poor, and so on. For this chapter we shall largely ignore all this and draw on some crude head count poverty estimates on the basis of household survey data that are easily available?

'See. for example, Srinivasan and Bardhan ( 1974), Lipton and Ravallion (1995), and Deaton and Kozel (2005),

•’Note that for both countries the last year for which the survey data are available is

POVERTY AND INEQUALITY 91

Source: World Bank data

TABLE 6Poverty Measures for $1 a Day Per Capita (In 2005 PPP)

Country 1981 1984 1987	1990	1993	1996	1999	2002	2005	2008
Percentage of Population
China 73.5 52.9 38.0	44.0	37.7	23.7	24.1	19.1	9.2	7.4
India 42.2 37.8 36.2	33.3	31.5	29.2	27.8	26.8	23.6	20.7
Number of People (in Millions)
China 730.4 548.6 412.4	499.4	444.4	288.6	299.5	244.6	119.9	98.0
India 302.4 290.0 296.8	291.0	292.4	286.9	250.0	287.1	269.0	246.5

Take, for example, the admittedly crude’ poverty line of $1 a day4 per capita (at 2005 purchasing power parity). With this poverty line, according to World Bank estimates, the proportion of people below it in China fell from 73.5 percent in 1981 to as low as 7.4 percent in 2008 (see table 6).

Those who do not want to use the World Bank estimates will get a qualitatively similar decline in poverty from Chinese official data, as in table 7.

Either way, this is a dramatic decline in a relatively short period. A part of this decline may not be real, as there may be some overestimate of poverty in 1981 since there are grounds to believe that the official price deflator used for this estimation may not have been adequate for rural areas before 1985. But even a somewhat smaller decline still means nearly a half billion people lifted above the

2005, which means that they do not include the possible rise in poverty in the more recent years of sharply rising prices of food, fuel, and metals or the economic downturn.

' A poverty line like $1 per capita per day is an arbitrary cutoff with very little link to any estimate of the basic needs of the poor. Also, PPP calculations are not sensitive to the differential subsistence needs of the poor in different countries and cultural contexts. We still use this here because we do not have access to any alternative time series of poverty data that would allow us to compare across the two countries. We should, however, note that the qualitative results about the decline over time stated in the text will not change even if one uses the national estimates using local poverty lines in China and India, as we show later.

«The World Bank now recommends the use of a $ 1.25 per capita per day (at 2005 prices) poverty line. The $1 line, however, is closer to the official Indian poverty line. At rhe $1.25 poverty line, rhe percentage of people below it in 2005 was 42 percent in India and 16 percent in China.

92 CHAPTER 7

TABLE 7

Percentage of People Below a Poverty Line, from Chinese Official Data

1. 1984 1987 1990 1993 1996 1999 2002 2005
2. 8 24.1 16.8 22.2 20.0 9.8 7.6 7.3 5.2

Source: Ra valtion and Chen (2007).

Note: Poverty line is defined here as 850 yuan per person per year in rural areas and 1,200 yuan in urban areas.

TABLE 8

Percentage of People Below a Poverty Line, from Indian Official Data

1983_____________1987-1988______________1993-1994_______________2004-2005

45.2 39.3 35.8 27.5

Source: National Sample Survey

Note: The official poverty tine is supposed to be defined as supporting the consumption of 2,400 calories per person per day in rural areas and 2,100 calories in urban areas.

poverty line in less than a quarter century—an unparalleled achievement in world history.

As shown in table 6, the decline in the percentage of people below the $1 poverty line (in 2005 purchasing power parity) in India was from 42.2 percent in 1981 to 20.7 percent in 2008, a large decline but not a dramatic one. Table 8 gives the corresponding percentages for selected years with the Indian official poverty line.

II

It is often claimed, both in the media and academia, that it was global integration that brought down the extreme poverty that had afflicted the two countries over many decades. While expansion of exports of labor-intensive manufactures probably lifted many people out of poverty in China in the past decades (not in India, where exports are still mainly skill- and capital-intensive), the more important reason for the dramatic decline of poverty over the past three decades may actually lie elsewhere. Table 6 suggests that more than

POVERTY AND INEQUALITY 93

half of the total decline in the numbers of poor people (below the poverty line of $1 aday per capita) in China between 1981 and 2008 had already happened by the mid-1980s, before the big strides in foreign trade and investment in China in the 1990s and later. Much of the extreme poverty was concentrated in rural areas, and its large decline in the first half of the 1980s is perhaps mainly a result of the , spurt in agricultural growth following decollectivization and one of J history’s most egalitarian land redistributions, with every rural family getting an equal piece of land (subject to differences in family size and regional average), in addition to an upward readjustment of farm procurement prices—these are mostly internal factors that had very little to do with global integration.5 To settle the issue of how much of the poverty decline is due to globalization and how much due to largely domestic factors, one needs causal models, which have not been econometrically tested much in the literature. In an attempt to econometrically measure, on the basis of a time series of province-level data, the effects of different factors on rural poverty reduction in China, Fan, Zhang, and Zhang (2004) show that since the mid-1980s domestic public investment (particularly in education, agricultural research and development, roads, and other rural infrastructure) has been the dominant factor both in growth and in rural poverty reduction, much more than economic reform. With a new provincial panel dataset Montalvo and Ravallion (2010)

'Ín the 1980s there was some trade expansion; for example, the export to GDP ratio went up from about 7 percent in 1981 to 12 percent in 1989. But labor-intensive manufactures were still not very important in Chinese exports; in the first half of the 1980s minerals and other natural resource-intensive products formed a substantial fraction of exports. In 1985, for example, the largest single export was petroleum. The mean tariff rate declined only slightly in the 1980s, from 31.9 percent in 1980-1983 to 29.2 percent in 1988-1990. The big decline in tariff rates started only around 1992; the range of commodities regulated by quotas and import licenses actually expanded over the 1980s. In any case, the proportion of the labor force in manufacturing in this period was small, so the large poverty decline in the first half of the 1980s is unlikely to be attributable to manufacturing exports. It is also worth noting that, after the sharp drop between 1981 and 1987, the poverty percentage went up or remained the same between 1987 and 1993. This indicates that by 1987 the agricultural spurt had worked itself out and the effect of labor-intensive manufactures was still weak. It was only after 1993 that the poverty percentage again started to decline; labor-intensive exports may have played a significant role in this decline, although even in this periini one should not minimize the effect of largely domestic factors such as easier migration from rural areas and higher agricultural procurement prices.

94 CHAPTER 7

‘There is some discrepancy between the NFHS data and those reported by the National Nutrition Monitoring Bureau (NNMB), not in the levels of malnutrition in 2005-2006, but in the trends over time. One should note that the NNMB surveys have a smaller sample size and are limited to ten states.

confirm econometrically that the poverty reduction in China has been mainly due to agricultural growth.

(In India, NSS data suggest that the rate of decline in poverty d not accelerate in 1993-2005, the period of intensive opening Íthe economy, compared to the 1970s and 1980s. This may be nnected with the fact that agricultural output (and TFP) grew at a slower rate in the past decade compared to the earlier decade (see table 1 in chapter 2). This may be largely on account of the decline [in public investment in rural infrastructure (such as irrigation, roads, and prevention of soil erosion), which has little to do with global- I ization. NSS data also suggest that there has been a decline in the / rate of growth of real wages in the period 1993-2005 compared to the previous decade, 1983—1993. We should also recognize that private consumer expenditure data of the NSS that are used in poverty estimates (as with the Chinese household survey data) do not capture the declining access to environmental resources (such as forests, fisheries, grazing lands, and water both for drinking and irrigation) on which the daily lives and livelihoods of the poor depend.

Global integration does not seem to have helped some of the other nonincome indicators such as those of health. The National Family Health Survey (NFHS) data show that some of India’s health indicators are worse than those of Bangladesh (in maternal mortality, infant mortality, child immunization rates, etc.), and even those of sub-Saharan Africa (in the percentage of underweight children), in spite of much higher growth rates in India than in those other countries. The percentage of underweight children (younger than age three) is 46 in India, and about 30 percent on average in sub- Saharan Africa (8 percent in China). Take the case of Gujarat, one of the richest, highest-growth, and highest-reform states in India: the percentage of underweight children, which was already high (higher than sub-Saharan Africa), went up between NFHS 2 (1998-1999) and NFHS 3 (2005-2006)?

POVERTY AND INEQUALITY 95

Some disaggregated studies7 across districts in India have also found trade liberalization actually slowing down the decline in rural poverty. Such results may indicate the difficulty of displaced farmers and workers in adjusting to new activities and sectors on account of various constraints (for example, in getting credit or information or infrastructural facilities such as power and roads, high incidence of school dropouts, and labor market rigidities), even when new opportunities are opened up by globalization. This is in line with textbooks in international economics that emphasize that productmarket liberalization need not be an improvement when there are severe distortions in input markets (such as those of credit, labor, electricity, land, etc. in the case of India).

India’s pace of poverty reduction has been less than China’s, not just because growth has been faster in China but also because the same 1 percent growth rate reduces (or is associated with reduction in) poverty in India by much less. This so-called growth elasticity of poverty reduction is much higher in China than in India; this may have something to do8 with the differential inequalities of opportunity in the two countries. We do not have good measures of inequality of opportunity,’ but in a poor agrarian economy such inequality is largely reflected in that of land and education. Contrary to common perception, these inequalities are much higher in India than in China.

As noted in chapter 3, the Gini coefficient of inequality of land distribution (in terms of operational holdings) in rural India was 0.62 in 2002; the corresponding figure in China was 0.49 in 2002. India’s educational inequality is among the worst in the world: according to a table in the World Development Report 2006, published by the World Bank, the Gini coefficient of the distribution of adult schooling years in the population, a crude measure of educational inequality, was 0.56 in India in 1998-2000, which is not just higher

'For example, Topalova (2007). In an unpublished comment, T. N. Srinivasan has raised some doubts about the methods in this study.

’Other important factors that may be operative here include those that restrict labor- intensive industrialization in India, such as those discussed in chapter 2.

’Most of the income or consumption inequality estimates that are usually cited are those of inequality of outcome, not of opportunity.

96 CHAPTER 7

than 0.37 in China in 2000, but even higher than almost all Latin American countries (for example, Brazil: 0.39), and some African countries. Because of the educational disparity among households, a study of the determinants of rural poverty in China and India by Borooah, Gustafsson, and Shi (2006) finds that education has a much bigger impact on poverty in India than in China. Apart from land and education, the other important ingredient of inequality of opportunity is due to ethnicity. In China, ethnic minorities make up only about 9 percent of the population. In India, with a much more heterogeneous society, about one-third of the population is in socially disadvantaged minority categories (such as Muslims, “dalits,” and tribal people), and thus minority status is a more important determinant of poverty in India than in China.

Comparing across states in India, as Datt and Ravallion (2002) point out, the growth elasticity of poverty reduction depends on the initial distribution of land and human capital. Purfield (2006) indicates that in the period 1977-2001 this elasticity was quite low in high-growth states such as Maharashtra and Karnataka, and high in states such as Kerala and West Bengal. Comparing across states and over time, Topalova (2008) estimates that “inclusiveness" of growth (measured by her as the difference between the growth in consumption of the bottom 30 percent of the population and the top 30 percent) depends significantly and positively on the share of the population that has completed primary and particularly secondary education, apart from financial development, flexible labor markets, and infrastructure in the state. Similarly, comparing across provinces in China, Ravallion and Chen (2007) find that growth had more poverty-reducing impact in provinces that initially had less inequality.

Ill

What about the link between market reform and inequality? At least two major problems beset the analyst in this matter. One is that so many other changes have taken place in the past quarter century of

POVERTY AND INEQUALITY 97

reform, it is difficult to disentangle the effects of reform from those of other ongoing changes (such as technological progress—often information-based or skill-biased—which usually changes the income distribution in favor of the better-off skilled labor groups, demographic changes, or macroeconomic policies). The second is that in both countries there are reasons to suspect that economic inequality (or its rise) could be underestimated (though not necessarily) in view of a widely noted fact facing household surveys (in many countries) of large (and increasing) nonresponse by the rich households. In addition there is reason to believe that the income data may not have adequately covered the rising importance of family-run businesses in China, which results in understating the rise in inequality. It is also difficult to compare inequality in China10 and India, as most of the inequality data that are cited in this context usually are for income inequality in China and consumption expenditure inequality in India (as the Indian NSS does not collect income data).11 Consumption expenditure data do show a rise in expenditure inequality in both countries in the past decade or so. But, as we have suggested, this rise may be an underestimate, and there is very little analysis as yet to show that this rise is primarily due to economic reform.

Even if economic reform were to be causally linked with higher growth, the link between growth and inequality is not always clear. In China, as Chaudhuri and Ravallion (2006) show, the periods of rapid growth did not necessarily bring more rapid increases in income inequality; the periods of falling inequality (1981-1985 and 1995-1998) had among the highest growth rates in average household income. In both countries, periods of high agricultural growth may have reduced overall inequality, since the agricultural sector usually has lower inequality than the nonagricultural sector. The

According to an estimate by Lin et al. (2008), that, unlike most other estimates, takes into account cost of living differences between rural and urban areas and across provinces, the national Gini coefficient of income inequality in China increased from 0.29 in 1990 to 0.39 in 2004, as mentioned in chapter 1.

”The National Council of Applied Economic Research data, which occasionally provide income data, suggest that (after correction for rural-urban price variation) the Gini coefficient for inequality of Indian income was 0.535 in 2004-2005, much higher than that in China.

98 CHAPTER 7

recent decline in agricultural growth rates may have had some influence in the rising inequality in both countries, which may not have much direct connection with market reform as such.

For the urban sector in India, figure 9 plots changes, over the period 1983-2004, in the real wage rate for urban full-time employees by level of education, and shows a faster rate of increase in the wage rate for those with higher education. According to the estimates by the Asian Development Bank (2007), the Gini coefficient of average real wages of urban full-time employees in India increased from 0.38 in 1983 to 0.47 in 2004. This increase in wage inequality is consistent with the skill-intensity of Indian economic growth (in which reforms may have played some role) and the looming talent shortage that the corporate sector complains about.

In urban China also the rate of return to college (and graduate) education compared to, say, high school education has more than doubled since the early 1990s. On the basis of household survey data in six provinces, Zhang et al. (2005) report that college graduates earned 12.2 percent more than senior high school graduates in 1988, but 37.3 percent more in 2001. In both China and India, here again it is difficult to separate the effect of skill-biased technological progjess from that of economic reform.

Beyond the discussion of wage inequality, while real wage rates have increased in both countries (much more in China than in India), it is quite likely that with market reform general workers may have experienced much more "churning” in the industrial labor market, particularly in China; but protests by organized workers against layoffs are more subdued in China. Strikes are not permitted, and the monopoly Party union has hardly any say in layoffs by employers. China laid off about thirty million workers from state- and collective-owned urban manufacturing enterprises in just five years, 1995-2000; this is unparalleled in the world history of the industrial working class, and unthinkable in India.

The official unemployment estimates are not comparable between the two countries. According to a somewhat more comparable estimate on the basis of household survey data by Giles, Park, and Zhang (2005), the unemployment rate of urban permanent

POVERTY AND INEQUALITY 99

residents in China was 11.1 percent in 2002. In India, according to NSS data, the urban unemployment rate was about 6 percent in 2004-2005. (In both estimates the reference period is the last week.) The much higher unemployment rate in urban China may be due to those massive layoffs from public enterprises in the second half of the 1990s. From a sample survey in thirteen cities. Appleton et al. (2002) find that of all the retrenched workers since 1992, 53 percent still remained unemployed in early 2000. Since then, however, the urban unemployment rate apparently has declined significantly, at least until the recent economic downturn, when by all accounts unemployment has risen steeply, particularly in globally exposed coastal China.

Since reform arguably has been more “urban-biased” in India and (at least since the 1990s) in China, one may look at the urban- rural disparity. The ratio of urban to rural mean income (or consumption) is higher in China than India, and it is rising in India, but the increase in disparity in China is attenuated when one takes into account the rural-urban differences in cost of living.12 Lin et al.

'•‘Secular er al. (2007) estimate the urban-rural income ratio to he 2.1, after adjustinc for spatial price differences. Another statistical problem is that the rural-urban definitions of households in China are somewhat uncertain because (a) rural areas are increasingly beine reclassified as urban in official data, and (b) the location of households by residence is different from that by registration (in view of the restrictions of the hulcnu svstem).

100 CHAPTER 7

1. decompose national inequality in China (measured by the Theil index), after taking into account cost of living differences, into inequality within each of the rural and urban sectors and inequality between the rural and urban sectors. It seems that both intraurban and urban-rural disparities contribute to recent increases in income inequality. The contribution of intraurban inequality to total inequality rose from 15.7 percent in 1990 to 34 percent in 2004; that of intrarural inequality fell from 72.2 percent to 35.6 percent; and that of urban-rural inequality rose from 12.1 percent to 30.4 percent. It should be noted that the urban population in China also gained disproportionately from the privatization (at highly subsidized prices) and marketization of housing since the mid-1990s.
2. we discuss in the next chapter, in China the urban-rural disparity in social services increased in the postreform period, with a near-collapse of the rural social services, whereas in urban areas social services, though weaker than before, still serve the majority of the resident (and working) urban population. The decline of rural health services in China (along with the one-child policy'5) may also have had some effect on gender equity in life chances. Male to female ratio in children (younger than six years) is very high at about 1.22 in China14 (1.09 in India). But one should add that female literacy and labor participation rates (above 70 percent in urban China, 24 percent in urban India) being substantially higher in China, women in China have had the opportunity to contribute to economic growth much more than in India.

Regional disparity in income (or consumption) is greater in China than in India. But over the past two decades, China’s backward regions have grown at rates almost comparable to its advanced regions, and regional earning disparities may be narrowing (though not yet per capita income disparities). Initiatives such as the Western Development Program supported large infrastructure building

”This policy may also have resulted in some underreporting of girls, apart from inducing gender-specific abortions.

”Unlike in India, where the gender imbalance in the population is greater in some of the better-off regions in the northwest, in China it is greater in some of the worse-off regions (such as the southeast).

POVERTY AND INEQUALITY 101

1. remote regions; and the mobility-restricting effects of the hukou system may be weakening. In any case, Benjaminet al. (2008) have shown on the basis of analyzing the survey data collected by the Research Center for Rural Economy that the contribution of interprovincial inequality in total inequality in China is smaller than is usually thought. Decomposing total national income inequality into between-province and within-province components after adjusting for cost of living differences, Lin et al. (2008) estimate that the former contributed only 11 percent.
2. India the poorer states (largely concentrated in the central and eastern regions) have grown much more slowly than richer states (mostly in the west and the south), so relative inequality has increased.15 In general, reform has advanced more smoothly in the west and the south of India, and better reform implementation in a state may have gone hand-in-hand with better initial infrastructure. Of course, with the removal of industrial licensing, which ostensibly used to give some weight to regional backwardness, private capital will move more to states where policies are business-friendly and infrastructure is better.
3. China provinces with more global exposure and higher growth did not have the larger increase in inequality. As Benjaminet al. (2008) show, while the Gini coefficient of income in coastal China increased from 0.35 in 1991 to 0.39 in 2000, it increased from 0.39 to 0.48 in the interior provinces.16 In the coastal provinces, more rapid job growth in the nonstate sector helped reduce the urban-rural income differential there.

But within-village inequality often rises in areas (particularly in backward areas) where poorer households are less able to participate in the booming nonfarm growth (possibly on account of deficiencies in skill and access to credit and marketing networks). For example the TVEs have been concentrated more in the eastern and coastal

”Decomposition of inequality of per capita consumption across states suggests, however, that the between-states inequality is still a very small proportion of the total inequality.

,,’The estimates by Lin et al. (2008) show that the results are somewhat modified in the comparison of the coastal and western regions if one adjusts for cost of living differences and extends the period to 2004-

102 CHAPTER 7

regions.17 Although India has had much less rural industrialization, in general rural industries have thrived more in regions that were already better off.

Finally, let us comment on one aspect of dynamic inequality on which there is very little comparative evidence or analysis, and yet in the long run it may be more important than the static inequality data that are usually cited. This relates to intergenerational mobility, which is likely to be much higher in China than in India. Khor and Pencavel (2006) found (subject to many qualifications they themselves cite for their estimates) that the degree of income mobility in urban China in the first half of the 1990s was much higher than that in the United States, and it increased between 1990 and 1995.'8 Contrast that with the usual statements about static inequality, which usually point the other way. Ding and Wang (2008) in a decomposition of the measurement of absolute income mobility in China for the period 1989-2000 show that a considerable part of the mobility is due to high growth. We don’t know of any comparable study of mobility in India. One sociological study of occupational mobility by Kumar, Heath, and Heath (2002) on the basis of socioeconomic data (for urban and rural males) collected over the period 1971-1996 in connection with the National Election Survey finds that while there has been some change in the occupational distribution (associated with the decline of agriculture) and some new opportunities for social advancement have opened up, there has been no systematic additional weakening of the links between the class positions of fathers and their sons. In terms of caste mobility over two generations, they find some modest evidence of an improvement in the chances of dalits (one of the lowest caste groups, confined earlier mostly to menial jobs) of getting into the salaried class, but not for other socially disadvantaged groups.

• this chapter, after describing the decline in poverty (dramatic in China, solid in India), we showed that the relationship between
• ln the inequality among rural households reported in Khan (2004), more than half was attributable to inequality in rural nonagricultural income (both wage and business income). Rural industrialization also drains skilled labor from backward villages, and their remittances go primarily to better-off households in those villages.
• course, looking only at income mobility in urban China cannot capture whatever influence the mobility restrictions of the hukou system still had.

POVERTY AND INEQUALITY 103

poverty reduction and market reform or globalization is more ambiguous than is usually claimed. The impact of growth on poverty reduction is weaker in India than in China, probably on account of initial conditions, including larger inequality (of opportunity) in India, owing to inequalities of land, education, and social status. The link between economic reform and inequality is also ambiguous and difficult to disentangle from the effects of other ongoing changes. There is some limited evidence of significant intergenerational mobility in China, but not much in India.

Chapters

The Social Sector:

The Relevance of the Socialist Legacy

• some basic indicators of the social sector China is clearly more advanced than India, but this is one area where the Chinese differential advance was achieved in the socialist period and, if anything, the period of market reform has eroded some of this advance. We shall focus on two parts of this sector, health and education. Even those who are primarily interested in economic growth will agree that these two basic ingredients of human capital in a population are crucial not just for current well-being but for long-term development and worker productivity as well. In both of these spheres there are some egregious failures of both market and government in most countries, but they have been particularly acute in China and India. We shall discuss the problems in the larger context of the political and economic structure in the two countries and show that the structural deficiencies from which they arise are similar as well as different in the two cases.
• most broad aggregative measures of health outcomes China’s performance has been much better than India’s, and it has been so for several decades. For example, life expectancy at birth now in India is what it used to be in China in the early 1970s, before the onset of reform. In infant mortality, by 1975 China had achieved a rate that India had not reached even in 2000. To this we may add, to repeat from chapter 7, that of India’s children under the age of three,

THE SOCIAL SECTOR 105

as many as 46 percent are underweight,1 compared to China’s 8 percent. The mortality rate of children under the age of five in India is more than twice that in China. Maternal mortality rate (per hundred thousand live births) in India has been almost ten times that in China.2 In India, malnutrition among women (particularly evident in the very large incidence of anemia, already among the worst in the world) increased between 1998-1999 and 2005-2006, according to the data from National Family Health Survey, in a period of high economic growth. This increase is quite substantial even in high-growth states such as Gujarat. More than 1.5 million Indian children die every year due to waterborne diseases such as diarrhea and dysentery.

There are, of course, some differences in initial conditions between the two countries. India being in general nearer the tropics than most of China is, one would expect a larger incidence of certain diseases in India, and conditions of vector control may be more difficult, other things being equal. According to World Health Organization estimates for 1998, the burden of infectious and parasitic diseases (measured in terms of disability-adjusted life years per capita) is seven times as high in India as in China. This may be partly the result of differences in physical and climatic conditions, but only partly, as it is also partly an outcome of relative policy deficiency. Socialist China had a much more vigorous policy for public health and sanitation than India, and also a larger army of paramedics pressed into basic public-health service in the villages (which took care of easily treated diseases and injuries, with some regions having a system of referrals to higher-level medical services for more difficult cases). By the mid-1970s China had a rudimentary system of medical insurance (called “cooperative medical scheme”) that covered the overwhelming majority of rural people, something that did not exist in India. Also, the Chinese government showed an

’ In weight-for-age measures, “underweight” refers to weights at least two standard deviations below the median.

•'The percentage of births attended by skilled health personnel in China has been twice that in India.

106 CHAPTER 8

ability to mobilize campaigns for preventive health care and against public-health threats that were impressive by the standards of most developing countries.

In contrast, India after Independence has never had a system of public health and sanitation anywhere on that scale. There has been no systematic planning and delivery of public-health services (as opposed to curative medical services) or sustained large-scale disease control. As Dasgupta (2005) points out, in India “there is strong capacity for dealing with (disease) outbreaks when they occur, but not to prevent them from occurring. Impressive capacity also exists for conducting intensive campaigns, but not for sustaining these gains on a continuing basis after the campaign. This is illustrated by the near-eradication of malaria through highly-organized efforts in the 1950s, and its resurgence when attention shifted to other priorities such as family planning.”

This public-health and preventive care situation is not entirely unconnected to the very different political-economy factors in India and in China in the early socialist decades. With the advance of antibiotics, the elite in India felt less threatened than in the past by the spread of communicable diseases among the poor, leading to a policy deemphasis on environmental hygiene, and the elite succeeded in diverting public fonds to high-end curative treatment in big urban hospitals, away from rudimentary but effective and widespread health services in the villages of the kind China used to have.

But in the past quarter century of economic reform there has been a sea change in public-health policy in China. With decollectivization in 1978-1979 the rural health services collapsed. The paramedics who used to be paid in work points at the production brigade and team levels now lacked a systematic method of compensation. Soon the total number of paramedics fell to less than a quarter of what it had been in the 1970s. By the mid-1980s the “cooperative medical scheme" covered less than 10 percent of the rural population (most of it in the better-off coastal areas). In general, with the collapse of local public finances, particularly in remote rural areas, fewer resources were devoted to public health. There was a decline even in curative services; rhe total number of hospital beds per thousand ru-

THE SOCIAL SECTOR 107

ral residents in 2003 was about half of what it had been twenty years earlier. Yet those twenty years saw phenomenal economic growth in China. While the basic indicators of public health kept on improving (with rising income, nutrition, education and transportation), the pace of improvement was slower than before, and was particularly bad for rural girls. For example, between 1981 and 2000, while the infant mortality rate for boys decreased from 40 per thousand to 25.8, that for girls decreased much less, from 38.1 to 36.7. Preventive care declined as government funds for disease control and prevention decreased from 0.11 percent of GDP in 1978 to 0.04 percent by 1993 (and it has decreased further in subsequent years). The incidence of some contagious diseases—such as hepatitis, schistosomiasis, and pulmonary TB, in addition to HIV—has increased.

In this period China essentially moved from one of the most impressive basic public-health coverage systems in the world to in effect a privatized (or user charge-financed) system,’ particularly in rural areas. In the cities, formal-sector employees have some form of health insurance, but there too premiums and fees paid by patients have increased considerably over time. Private health insurance has started, but in 2003 it covered less than 6 percent of the population. Obviously, the poor have had to bear the brunt of all this, as even in the cities most of them are in the uncovered part of the population, migrants and informal-sector workers. Yip and Mahal (2008) point out that 76 percent of the lowest-income quintile of urban individuals do not have health insurance; the corresponding percentage in the lowest-income quintile of rural individuals is 80 percent. This implies that many sick people do not seek medical care, largely on account of financial hardship. Yip and Mahal cite data that in 2003 nearly half of those reporting an illness did not seek outpatient care. Those who did spent an inordinate proportion of their income on health care; according to their data, the poorest-quintile individuals in rural areas spend as much as 27 percent of their income on health care, and in the poorest urban quintile it is 11 percent.

’This was the case even before the recent appearance of openly for-profit hospitals and clinics. It is reported that 72 percent of clinics in China are now for-profit.

108 CHAPTER 8

This large change in the public funding basis of health services in China is linked with a systemic problem relating to decentralized development. As discussed in chapter 2, China is a glowing example of industrialization under decentralization. But one side effect of economic decentralization is acute regional inequality. Coastal China surged ahead, and local governments there flush with profits from the TVEs under their control could buttress social services as their funding from communes disintegrated all over the country. But the interior or agriculture-dominant provinces and remote areas, where TVEs were few and profitable ones even fewer, were largely left to their own devices in funding social services. There was a highly inadequate block grant from upper levels of government, and medical facilities were encouraged to generate as much revenue as they could (often from drug sales, test fees, and unnecessary procedures),4 the surplus from which they were allowed to keep or pay the staff in the form of bonuses. Informal payments from patients (sometimes called “red packages”)5 and kickbacks from drug wholesalers have been rampant.

Then the fiscal reforms of 1994 centralized revenue collection and allocation, and many local areas were left with unfunded mandates for basic social services including education and health care. The fiscal reforms of more recent years clamped down on some of the arbitrary fees and taxes that many local governments had imposed on the local population, leaving them more financially strapped. Increasing regional disparity in provision of health care is indicated by the fact that in 1985 the total number of technical medical personnel per thousand people was somewhat lower in cities than in the surrounding counties; twenty years later, it was more than twice in cities than in counties. It is also no coincidence that, as Yip and Mahal (2008) estimate, the crude measures of interprovincial inequality (such as the coefficient of variation) in aggregative health

*For example, the percentage rate of births by cesarean section (a more expensive procedure) has multiplied several times since the onset of market reforms.

Mt is reported that more than 70 percent of hospital patients in China have to make these informal payments; the reported figures in India are much smaller, though not insignificant.

THE SOCIAL SECTOR 109

outcomes such as life expectancy at birth or infant mortality increased in the twenty years following 1980.

While China moved away from its egalitarian and impressive basic health-care service of the socialist period, India’s health-care service remained dismal and inegalitarian throughout. Only about 15 percent of the people in India have any health insurance (primarily through their employers), and the share of out-of-pocket spending in total spending on health care exceeds 70 percent, which is higher6 than in China (though it has increased faster in the latter country). Appearances to the contrary, health care in India is predominantly private and largely unregulated. Household survey data suggest that 85 percent of all visits for health care in rural areas, even by the poorest people, are to private practitioners; about 60 percent of all hospitalizations in India are in private hospitals. While the poor quality of service in public clinics and hospitals (and absenteeism by nurses and doctors) often drive patients to private doctors’ (some of them quacks or crooks) in India, in China the high fees charged in public-health clinics (and their concentration on revenue-generating activities) in effect turn them into for-profit private providers. In India, unlike China, at least the public facilities receive the bulk of their revenue from government subsidies and they provide their (often paltry and poor-quality) services at low cost to those who are too poor to afford more expensive private care (although rampant corruption renders the public service provided not entirely free). China, however, unlike India, has had a long tradition of training and licensing village doctors. The Indian system is more top-down: Although health-care delivery is largely in the jurisdiction of the state governments, most of their budget is spent on staff salaries and they are thus much more dependent on the central government both for general health-care plans and priorities and for nonsalary items of public-health expenditure. Also, as a part

“More than 90 percent of per capita spending on health care is private, not just in socially backward states such as Uttar Pradesh but also in states such as Kerala, which has good public-health facilities.

:There has been very little attempt to train them or help them qualify for minimum licensing arrangements. Professional medical associations have not generally been helpful in this respect.

110 CHAPTER 8

• f reorientation of those priorities, the fraction of total health-care spending on control of communicable diseases has declined over the years, leading to a reemergence of some of those diseases.
• both countries doctors often overmedicate and refer patients for unnecessary diagnostic tests and procedures, driving up healthcare costs in general. This is part of a general market failure in health care, in which the decider (the doctor) is not the purchaser (the patient), and may have a financial interest in dispensing medicines or performing tests. In poor countries with little information and education, the problem is exacerbated as the patients themselves sometimes show a preference for unnecessary antibiotics and steroids, which the quacks oblige. In both countries the more important problem is a governance failure. The public-health delivery system is afflicted with poor provider incentives coupled with little accountability to patients.8

First, medical personnel are often paid a fixed salary independent of the number of patients or the number of patient visits, so they have no economic incentive to serve patients. In India, doctors in public hospitals are allowed to practice in private as well, which biases their financial incentives toward focusing their efforts on serving their private patients.’ In China, some of doctors’ nonfixed salary comes from commissions on drug sales, with effects on overprescription. Second, there is little monitoring or punishment for laxity in service in either country. Third, the poor have very little organized “voice” in sanctioning errant providers in either country. In the otherwise vibrant democracy of India, in most areas local democracy is not strong enough to keep public service providers accountable to the local citizens. Periodic elections provide a rather blunt instrument for keeping public officials in check, and in any case the electoral agenda is full of different pressing issues, of which poor health service is only one. Besides, politicians find it easier to claim credit for inaugurating a big hospital or installing new equipment there

1. For an elaboration of these issues in the context of India, see Hammer, Aiyar, and Samji (2007).
2. China, public hospital doctors are allowed to do contract work for other hospitals on their days off.

THE SOCIAL SECTOR 111

I

for regular maintenance of services or public sanitation and * control. In China, the channels of local accountability are weaker. In both countries, local social groups and nongovem- »1 organizations (NGOs) provide some accountability pressure ík focalized pockets (more in India than in China).

II

ÿ education, the contours of the basic story in both countries are

* Wedar to those in health care in many respects. Socialist China had ® active policy of spreading literacy and basic skills among the gen- «t? population,10 to an extent unmatched by India. In 1982, even iferr the chaotic years of the Cultural Revolution, two-thirds of the «Wt population in China was literate or semiliterate, at a time when ,®> India more than half the population was illiterate. The average wars of schooling in the Indian population older than age fifteen in

was less than what China had already achieved by the begin- I esr® of the 1980s. Of course, relative spending in India used to favor fetẃer education somewhat (just as health-care spending favored urfem hospitals and curative medicine). In 1982, less than 1 percent

1. ctf adults in China had any college education at all, while India had tferee times as many college graduates per capita as China. The Cul- scsal Revolution, which can be described as, among other things, an «prode of one of the largest destructions of human capital in his- sorv, wreaked havoc particularly on higher education. (In the area of health, the catastrophic disruption was the Great Famine in the early 1960s). In the reform period China has tried to make amends by dramatically expanding attainment particularly in higher education. By ?004, 6.7 percent of adults had some tertiary education. Of general secondary school graduates, only 4.6 percent enrolled in a regular institution of higher education in 1980; by 2000 that had increased ,
2. course, education did not significantly increase income; Communist Party membership mattered more for improving income. In the past two decades the rate of return to education has increased, but there is evidence that returns to Communist Party membership 1 not diminish.

112 CHAPTERS

1. 73.2 percent. India also accelerated enrollment in higher education, but not as rapidly. As in the health area, decentralized financing of education in China made things difficult for local governments, particularly in rural and remote areas, to keep up with the mandated educational expansion, and the requisite increased school fees made things difficult for the poor for many years. Thus educational disparities increased in the reform period, and differential schooling has now become a major source of widening class division in China.
2. More than half of adult women are illiterate in India; in China, ¡only around 10 percent are illiterate. For all its brilliant scientists and software engineers, India is the largest single-country contributor to the pool of illiterate people (and school dropouts) in the world. Only about half of India's children have completed eight years of education by age fifteen." One-fifth of India’s children in the five to fourteen age group are not in school. Youth illiteracy has been a special problem in India for many years (it is greater even than that in many African countries). In the twenty-five to twenty- nine age group, for example, China has only negligible illiteracy left, whereas in India about 30 percent are still illiterate. Household survey data in India suggest that the secondary school attendance rate at ages sixteen and seventeen is 50 percent higher for the richest expenditure group compared to the poorest, and three-quarters of the difference results from higher attendance in private schools by the richer children. Lower attendance of poorer children is partly due to supply-side problems (distance to school, teacher absenteeism, lack of remedial programs for poorly performing students and especially for girls, lack of safety in the journey to and from school, and lack of toilets in schools), in addition to the usual demand-side problems (poverty that makes teenagers drop out in order to earn something for the family or, in the case of girls, to take care of household chores when the mother is working, and social discrimination that works against girls’ education). In fact, India's educational inequality is among the worst in the world, as mentioned in chapter 7.
3. for quality of education, one dismal indicator noted by Pratham, a large education NGO, is that even in fifth grade some 35 percent of the children cannot read or write

THE SOCIAL SECTOR 113

• China, private schools have grown rapidly in recent years, particularly for secondary education in urban areas. The share of total education funds for all levels coming from government budgets dropped from 84 percent in 1991 to 62 percent in 2004- Among the nonbudgeted items, tuition fees alone grew from 4 percent of educational expenditures in 1991 to about 19 percent by 2004. Inability to pay school fees became a major source of educational disadvantage for the poor. Girls are still at a disadvantage in number of years of school, although gender disparity in education seems to be decreasing.
• have already referred to the increased emphasis on higher education in both countries. But less than 10 percent of the eighteen to twenty-three age group enroll in higher education institutions in India (compared to more than 20 percent in China). Of course, even a microscopic minority that is highly educated in a large country is sizable in absolute number and can make a splash in the world markets—for example, it has been reported for some time that there are more IT workers in Bangalore now than in Silicon Valley in California. But the sustainability of this for India as a whole is in some doubt, particularly when the majority of higher education institutions in the country are currently dysfunctional. The student’s university performance as signal of quality is increasingly replaced by that in competitive examinations outside. The higher education institutions are strapped for government funds," and yet ways of mobilizing private resources for public institutions remain largely blocked. They are overregulated in most academic decisions and often hopelessly politicized, and low in research productivity (measured, for example, by top journal citation index) in science and technology. As figure 10 shows, in terms of the number of research articles published in top-quality international journals by resident researchers China has now outpaced India. In the 2008 Times Higher Education ranking of world universities in terms of academic reputation and performance, the list of the top one hundred universities included two from China and none from India.
• contrast, state financing for higher education more than doubled in China in just five years, between 1998 and 2003.

IH CHAPTERS

Both in research and development expenditure as a percentage of GDP and number of patents taken out, China is far ahead of India. It is projected that by the early 2010s, China will surpass the United States and Japan in applications for new patents (particularly in the areas of chemical engineering, digital computers, and telephone and data transmission).

Already a talent shortage is reported to have hit India’s capital goods industry, and even for the IT-related sector it is believed in some quarters that the reservoir of India’s technical and managerial skills may prove rather shallow in the near future.11 In any case, companies often have to put the engineering graduates they hire through long, intensive training programs to make them employable for their purposes.

In both countries there is now a renewed effort on the part of the government to press more resources into and improve the delivery of public education and health-care services. The Chinese

‘According to the McKinsey Global Institute, talent shortages are looming in both China and India. China is producing more engineering graduates than India, but the propor* tion of them who are. according to the study, “suitable” for professional jobs, particularly at the international standard, is somewhat higher in India than in China. In 2003 the total number of “suitable” young engineers (excluding civil and agricultural engineers but including those with IT and computer science degrees) was 160 thousand in China and I 30 thousand in India.

THE SOCIAL SECTOR 115

program seems more ambitious, in attempting to provide partially subsidized universal basic health care14 and a program of free universal and compulsory education for nine years, and the Chinese government has more budgetary resources to devote to this. There is a new county-level voluntary health insurance scheme afoot, known as the New Cooperative Medical Scheme. In particular, its aim is to insure rural residents against catastrophic health-care expenses. The National Rural Health Mission in India plans to more than double the government spending on health care, with the purpose specifically of increasing access to primary health care for the rural poor. The major initiative to universalize education in India particularly at the primary and lower secondary levels is the recently launched Sarva Shiksha Abhiyan (“education for all” campaign).

But in both countries the governance and accountability issues mentioned earlier will not be resolved easily. In India, the weakness of local democracy (in terms of autonomy, funds, and administrative personnel), coupled with a corrupt and inert bureaucracy, dissipates many a well-intentioned policy measure from above. In China, it remains to be seen how effective and adequate the actual implementation of the ambitious program will be and how much the current burden of most of the revenues of public hospitals being extracted from patients will be reduced. Over the past several years, the constant chanting of the "harmonious society” mantra by the central leadership has not always succeeded in reining in local officials from their hitherto single-minded, frantic (and lucrative) pursuit of income growth often at the expense of social welfare. In any case, the latter are not too enthusiastic about the large share of the proposed increase in funding to be matched from local government coffers. Without a fundamental change in the way rural hospitals and schools are funded, poorer areas will still have a great deal of difficulty in scraping together local financing, even with a somewhat larger financial commitment by the central and provincial governments (which so far has mainly covered the loss of local

H Recently the government announced that this policy will be accelerated as part of the domestic demand stimulus program, since health-care uncertainties motivate a significant part of precautionary saving, particularly in mral China.

116 CHAPTERS

revenue from the abolition of agricultural taxes and miscellaneous fees).

In both countries, the basic problem of equity and quality of social services remains when both the bureaucracy and the provider are bound to act out of self-interest if their incentive systems are not restructured, and when the intended beneficiaries are not well informed about what is best for them and often lack the “voice" or power to sanction, even when they are.

Chapters

Environment: The Alarming Signs

T he environment includes many different aspects of nature and the ecosystem that envelops human life, even though most public discussion currently tends to concentrate on issues around global warming. In this chapter we shall focus mainly on the local commons that affect the daily lives and livelihoods of many poor people in China and India. Local environmental resources include the quality of air and water as basic necessities of life as well as the conditions of the village commons such as forests, fisheries, agricultural land, and irrigation, which are particularly important for the rural poor. For comparative purposes we need some quantitative indexes for these different aspects of environment in the two countries. Table 9 provides some aggregate information regarding different indicators of environment, based on calculations by the Yale University Center for Environmental Law and Policy and the Columbia University Center for International Earth Science Information Network. These are, of course, necessarily crude measures, which through their methods of standardization and aggregation lose many important local particularities and nuances, but they give us a rough comparative picture.

Tire overall environment performance score for China in 2008 is 65.1 (with rank 105 out of 149 countries), somewhat better than that of India’s 60.3 (with rank 120). For comparison, the highest score among these countries is for Switzerland, 95.5, and the lowest is for Niger, 39.1. The scores for China and India are also