Microsoft Word EcRefChina Oct06. doc

between urban and rural areas
fell in the early 
1980s in connection with the agricultural reforms, the relative gap seems today to be 
approximately back at the pre-reform level, if the different price trends in urban and rural 
areas are properly taken into account (UNDP, 2005
; 
Ravallion and Chen, 2006). The 
generally accepted assessment among observers seems to be that in 2003, per capita 
disposable income in urban areas was more than three times (3.3 times) higher than per 
capita net income (an approximation for disposable income) in rural areas
72
– four times 
higher if the provision of human services were included in the calculations (UNDP, 
2005).
73
An important background factor is, of course, that labor productivity is about 5 
times as high in manufacture and services as in agriculture (an average for industry and 
services).
74
70
See UNDP (2005, Figure 2.5 and Table 2.2). Ravallion and Chen (2006, Table 10) argue that the Gini 
coefficient is somewhat lower after adjustment for cost of living differences between urban and rural 
areas: 0.40 rather than 0.45 (in 2002) without such adjustment.
71
For instance, Kanbur and Zhang (2005) report that the Gini coefficient for the distribution of average income 
across provinces
rose from 0.29 in 1978 to 0.37 in 2000.
72
Around 1990, the ratio seems to have been just above 2. 
73
Indeed, it seems that about two-fifths of income inequality in the country as a whole in 2002 could be 
accounted for by the urban-rural income gap, decomposing the Theil inequality index (UNDP, 2005).
74
Kuijs and Wang (2005) assert that labor productivity is about 3.5 times as high in services as in 
agriculture, and about 9 times as high in industry. Holz (2005b, Fig. 1) presents even higher figures for 
industry (14 times higher). 

---

41
In general, an increase in the dispersion of the distribution of income in the early phases 
of economic modernization is not unusual – and is often described in terms of movements 
along the upward-sloping section of the “Kuznetz curve”. This is often assumed to reflect 
the distributional consequences of the emergence and expansion of a high-productivity, 
modern sector of firms alongside traditional low-productivity production units.
75
When 
trying to explain the rise in income inequality in China, we have to add that the 
modernization process took place simultaneously with a shift to an economic system that 
relies, much more than the previous one, on economic incentives. It should, however, be 
added that the increased reliance on economic incentives in agriculture in the late 1970s 
and early 1980s, for a while actually contributed to 
reduce 
the relative income gap in the 
country – not only between urban and rural areas but also across households in the 
country as a whole. This illustrates the obvious but important point that the relation 
between economic growth and income inequality depends critically on the sector 
composition of aggregate economic growth. Similarly, the distributional consequences of 
increased reliance on economic incentives depend on which population groups actually 
experience such an increase. 
Specific policy measures have also contributed to increase the income gaps during the 
reform period. For instance, income differences across provinces were accentuated by the 
opening up of the country to foreign trade and investment in the mid-1980s, since this 
released the (potential) absolute and comparative advantage of coastal (eastern) provinces 
with relatively good locations for such activities. The differential effects of these policies 
across provinces were accentuated by the