Microsoft Word EcRefChina Oct06. doc

social 
assistance, 
whose main component is an “urban living standard guarantee”,
120
 
with a 
corresponding, although much less generous, program in rural areas. Although the take-up 
rate for social assistance is rather low today, about a quarter of the number of eligible 
individuals (Social Policy Research Centre, 2002, p. 11), it may increase by receding 
stigmatization of living off benefits, in particular if unemployment reaches high levels for 
prolonged periods.
117
As in most developed countries today, the benefits are (or at least will be) financed by payroll taxes 
on firms (two-thirds of the cost) and employees (one-third).
118
The rules require that the benefit level is set between local so-called “minimum living standard” 
(
dibao 
standard) and the minimum wage (or possibly a fraction of this). According to the OECD, the 
benefit level is usually 120-150 percent of the minimum-standard in the province in question.
119
The OECD (2005a) reports that 7.4 million people received unemployment benefits at the end of 2003. 
120
The benefit level in this system is 25-30 percent of the average income in the district, or province, 
where the individual lives (Tang, 1999).
 
Today this system provides income support for about 20 million 
people in urban areas, and about 5 million individuals in rural areas (OECD, 2005a, p. 85). 

---

60
Like other social insurance systems in China, the emerging mandatory 
pension system 
mainly covers urban citizens (about 163 million participants in 2004).
121
It accounts for 
about 80 percent of total social insurance spending, and it is responsible for 28 out of 
the 40 percentage points of aggregate social insurance fees. It is a two-layer system.
122
The first layer consists of a strongly redistributive paygo system, amended by a (so far 
modest) collective buffer fund – the National Social Security Fund, established in 
2000. Although the role of the collective buffer fund remains unclear, one purpose 
seems to be to help poor regions finance their pensions, hence compensating them for 
the limited geographical pooling. 
A second layer is supposed to consist of a fully funded, actuarial system with 
individual accounts, based on earlier paid contributions, hence a system without 
intended (ex ante) redistribution.
123
One rationale for such a tier might be that a 
combination of paygo and funded elements provides better risk pooling than only one 
of these elements in isolation, since the risk characteristics between them differ. 
Another rationale might be to boost the development of a broad, domestic capital 
market. While these two rationales for funded elements in the mandatory pension 
system are highly relevant for China, two other usual arguments – to raise the 
aggregate national saving rate and to provide a higher return on mandatory pension 
saving – may seem less relevant. After all, the aggregate saving rate is already 
exceptionally high (about 50 percent of GDP), and the fast growth rate of aggregate 
earnings makes the return on mandatory saving in a paygo system quite high as 
compared to the return on a funded system.
124
However, since a fully funded system 
takes decades to build up, and both the saving rate and the growth rate are likely to fall 
121
According to a survey by the Chinese Research Center on Aging (2002), more than 70 percent of the urban 
elderly were covered by the new pension arrangements, while the corresponding figure in rural areas was 3 
percent. 
122
See, for instance, OECD (2005a, pp. 187-190); and Wang, (2006, p. 10). 
123
While the first layer promises a pension annuity amounting to 20 percent of the