Microsoft Word EcRefChina Oct06. doc

The 
Economist
, Sept. 11, 2004, p.55 (based on calculations by the National Bureau of Statistics in China, and 
Hu Angang, Tsinghua University).
81
According to McKinsey Global Institute (2006, Exhibit 3.18), less than 2 percent of the households 
accounted for more than 60 percent of liquid financial assets in 2003.
82
Of course, if the imputed capital value of land-lease contracts were included in household wealth, the 
distribution of wealth would seem more even. Indeed, with this definition of wealth, the Gini coefficient for the 
distribution of wealth would be only slightly above the Gini coefficient for income (UNDP, 2005, pp. 31-37, 
based on a study of household surveys by the Chinese Academy of Social Sciences, 2002).

---

43
wealth uneven. The 
way
in which the housing stock was privatized also contributed, to 
begin with, to making the distribution of wealth more uneven, since housing was often 
sold at prices below market values to groups of citizens with income and wealth above the 
average – another example of “asset stripping”. However, in recent years, as the 
ownership of housing has gradually broadened, the distribution of housing wealth has 
tended to become somewhat more even (UNDP, 2005). 
It is not self-evident that the income and wealth gaps will continue to widen in the 
upcoming decades. One reason is that the modernization of production is likely to spread 
across firms and across regions, so that China may begin to slide along the downward-
sloping (rather than the upward-sloping) section of the Kuznetz curve. Indeed, already 
today, there is a tendency for the modernization process to spread to a number of interior 
provinces (Agarwala, 2002), although it is unlikely that 
all
provinces will take part in this 
process to the same extent.
83
The future development of the overall distribution of income 
in the country depends also crucially on what happens to productivity in agriculture and 
the relative price of agriculture products – in the same way as these factors dominated the 
development of the distribution of income in the early 1980s. Moreover, further 
expansion of Chinese exports of labor-intensives products may, in itself, favor workers 
with modest skills, although the highly elastic supply of such labor, and perhaps also the 
absence of traditional labor unions, tends to dampen the positive effects on wages for low-
skilled workers.
84
But what will actually happen to the distribution of income in China 
during the coming decades depends, of course, also on what type of economic and social 
policies that will actually be pursued in the country – in the same way as the development 
of the distribution of income in recent decades partly is a result of specific policy 
measures during that period. 
Let us turn from China’s social development in terms of income and wealth and their 
distribution to the provision of 
human services
, such as education and health care. China 
made considerable achievements in these fields already in the pre-reform period 1950-
83
On the basis of their study, Brun et al. (2002) argue that the spatial spillovers have been confined to 
coastal regions. By contrast, Luo (2005) argues, on the basis of provincial panel data from 1978 to 1999, 
that there have been large spillover effects on the entire Chinese economy due to the rapid expansion of 
the coastal provinces, but that these spillover effects have not been strong enough to prevent a widening 
of regional (relative and absolute) per capita income gaps.
84
For a survey of the rather ambiguous results in the literature on the distributional consequences of 
increased openness to international trade, see, for instance, Wei and Wu (2003).

---

44
1978, as reflected in the huge fall in illiteracy and nearly a doubling of life expectancy.
85
Broad education and health indicators have continued to improve during the reform 
period. However, the pace has slowed down considerably. Indeed, China seems to have 
lost some of its lead among developing countries in East Asia in these fields. In particular, 
China does not rank very high today among developing countries in terms of public-sector 
spending on education and particularly not in health care (UNESCO, 2005; World 
Developing Indicators, 2005). 
In the case of primary school enrollment, China today seems to be on level with other 
countries in East Asia with about the same per capita income. However, in spite of 
considerable progress in education during the last half-century, only a modest fraction of 
today’s adult population has received more than elementary schooling.
86
Education is also 
very unevenly distributed across geographical areas and population groups.
87
Therefore, 
there is a strong relationship between children’s socio-economic background and their 
education.
88
Girls in remote rural areas lag behind the most, especially those who belong 
to minority nationalities.
89
These distributional problems in the field of education are largely the result of the 
relatively small tax financing of education in China, and a correspondingly large share of 
non-government (“non-budget”) financing, including payments from various 
organizations and out-of-pocket money from households. Non-government financing 
seems to account for 32-46 percent of the total spending on education,
90
and according to 
Zhang and Kanbur (2005), private out-of-pocket money has increased from about 2 to 
85
The illiteracy rate fell from 80 percent in 1949 to about 32 percent by 1980 (Zhang and Kanbur, 2005). Life 
expectancy from birth increased from 35 years in 1945 to 68 years in 1981 (China Human Development 
Report, 2005).
86
For instance, in 2000, 16.5 percent of the population aged 25 or older had high school training, 4.3 percent 
college level education, 1.4 percent bachelor’s degree and 0.1 percent an advanced degree (master’s or doctor’s 
degree); see, Holz (2005b, Table 4). No more than about one percent of a relevant cohort is currently in 
research training; see UNESCO (2005); EU (2005); Freeman (2005).
87
Adult illiteracy is about twice as high in rural as in urban areas.
Moreover, for the country as a whole, 
illiteracy among adults is reported to be 2.6 times higher for females than for males (UNDP, 2005, p. 3). 
According to a World Bank report (2001), about half the boys and nearly all the girls in some of China’s 
poorest villages do not even attend school, and they are not likely to ever achieve literacy. 
88
Indeed, according to the World Bank (2002, pp. 42-43), the share of children from poor families who 
complete basic education fell from 68 percent in 1988 to 53 percent in 1995 in non-poor regions, and 
from 54 percent to 42 percent in poor regions. The population census in 2000 also shows a slower 
improvement in educational attainment during the 1990s for ethnic minorities than for the Han majority, 
which constitutes 92 % of the population.
89
For a summary of the socioeconomic status of ethnic minorities in China, see West (2004). 
90
The highest figure, 45 percent, is reported by Chow and Shen (2006).

---

45
about 13 percent of the total spending on education in recent decades. The latter consists 
largely of tuition fees, schoolbooks, transportation, and school uniforms.
91
In the case of health outcomes, China is relatively successful in terms of life expectancy, 
but less so in terms of infant mortality (UNICEF, 2005). There is, however, general 
agreement that the uneven distribution of health-care services across geographical areas 
and population groups is a major social problem.
92
Indeed, the consumption of health care 
seems to be at least as uneven as the distribution of income. Thus, the increased disparity 
of income in recent decades has carried over to the distribution of health care. The 
deficiencies in health care are also a significant source of economic risk for the individual, 
in particular among those with narrow financial margins. In some cases, a vicious circle of 
poor health and poor financial resources may therefore have emerged. 
The low level of tax-financing of health care is, of course, an important explanation for 
the uneven distribution of health-care services across population groups. The non-
government share of total expenditures on health care is reported to have increased from 
about 16 percent in 1980 to as much as 61 percent in 2001 – a considerable fraction paid 
in the form of out-of-pocket money by households (Zhang and Kanbur, 2005). The non-
government share is much larger in the countryside than in cities: 87 percent as compared 
to 44 percent. Indeed, the number of health personnel in rural areas is reported to have 
declined by around 12 percent between 1980 and 2001 (UNDP, 2005, p. 58).
One reason, 
in addition to the lack of government-financing, is that many barefoot doctors found it 
economically rewarding to move to urban areas or shift to other occupations, including 
agriculture.
91
Some families also finance additional education for their children, for instance in the evenings, on 
weekends and during vacation periods – partly to enhance the competitiveness of their children when 
applying to institutions for tertiary education. Some relatively affluent farmers also send their children to 
schools in urban areas.
92
Although about 57 percent of the current population live in rural areas today, according to Chow (2006b) 
only about 20 percent of medical resources have in recent years been allocated to these areas. According to a 
calculation by Pei (2006, p. 173) rural residents receive only one-third of the health care per person of the urban 
population. 
Only 27 percent of government subsidies to health-care provision are allotted to “township health centers”, 
even though these centers are supposed to cover 43 percent of the population (World Bank, 2003a). Moreover, 
while households in urban areas seem to spend 6-7 percent of their income on health care, the corresponding 
figure for households in rural areas is reported to be 2-3 percent (preliminary estimates by Lindelöw and 
Wagstaff, 2005). In both areas, however, poor individuals feel compelled to spend a larger share of their 
income than the non-poor on such services (8-9 percent of their income in urban and 6-7 percent in rural areas.)

---

46
The uneven availability and affordability of health-care resources across geographical 
areas is also reflected in various health indicators, such as life expectancy, infant 
mortality, maternity mortality rates and underweight among children in different areas of 
the country.
93
Indeed, in some especially poor rural areas in the west and north, the health 
status does not seem any better than in some of the most impoverished regions in the 
world. In the poorest Chinese villages, there is still a high incidence of various types of 
infectious and endemic diseases, including chronic worm infections (World Bank, 2001). 
At the same time, in certain other areas, such as in the Shanghai region, both health care 
and health status is approaching the standard of developed countries.
An important background factor for the huge regional and local variation in both 
education and health is, of course, that the per capita tax base varies dramatically across 
villages and municipalities.
94
The inter-government money flows sent down from the 
central authorities to lower government levels partially compensate poor local 
governments for their limited powers of taxation.
The compensation is quite modest, 
however. On their way down to local governments a considerable fraction of these money 
flows tend to be siphoned off – in more or less the same way as some Chinese rivers, 
which are full of water inland, turn out to be nearly dry when they reach the coast. As a 
result, transfers from the central government have not prevented an increasing gap in 
fiscal spending across geographical areas during the period of economic reform.
95
Indeed, 
in some poor areas, this has resulted in a combination of relatively high taxes and 
deficient human services. 
Summing up: when China gradually shifted to a new economic system and the old 
arrangements for education and (in particular) health care broke down, the authorities 
were slow in building up new arrangements. This is, of course, a parallel to the slowness 
93
For instance, in 1999, infant mortality was 37 per 1000 live births in rural areas, as compared to 11 per 
1000 in urban areas; maternity mortality rates were 61 versus 33; underweight for children below five 
years of age was 14 versus 3 percent; and stunting rates for children under five were 20 versus 2.5 
percent, respectively. While life expectancy is reported to be about 66 years in the least advanced 
provinces, it is 80 years in the most advanced ones (UNDP, 2005, p. 98). 
94
Local authorities finance most of the government spending on both education and health care. An 
important background factor is the administrative decentralization of financial responsibilities within the 
public sector initiated in the early 1980s – the principle of