Microsoft Word EcRefChina Oct06. doc

participated in programs financed mainly by out-of-pocket money. 

---

71
On the level of tertiary (college and university) education, the number of students is, of 
course, largely determined by a combination of incentives to acquire such education, and 
the strictness with which the slots are rationed in the admission process, which in turn 
depends on the capacity (resources) of the system of tertiary education. Today, the 
bottleneck seems to be in terms of resources, since there is fierce competition for the 
admission to tertiary education. Naturally, the situation may change in the future as a 
result of the gradual build up of the resources of tertiary education. If the (expected) 
economic return on tertiary education, rather than the resources at colleges and 
universities, would turn out to be the bottleneck (a low “college premium”), the problem 
could be mitigated by either allowing wider wage dispersion or reducing the progressivity 
of the tax system. An alternative would be to increase the per-student subsidies to 
university education, including cash payments to students for living expenditures – or a 
combination. But it is difficult (in fact impossible) to calibrate and differentiate 
educational subsidies across different types of education so that they provide the same 
relative incentives across skill groups and professions, as do larger wage differentials. 
From that point of view, wage differentials fulfill an important role even if the 
government’s financial involvement in higher education is considerable. 
However, to bring about broad-based recruitment of students at the tertiary level, 
international experiences suggest that it is important to provide student loans on 
reasonable terms, such as government loan guarantees – possibly with the amount of 
yearly amortization contingent on subsequent earnings. There is also a case for direct 
means-tested cash grants to students from low-income families, since they are likely to be 
particularly reluctant to incur debt when investing in human capital. Indeed, such targeted 
subsidies were quite usual in today’s developed countries as late as a few decades ago. Up 
to a certain number of university students, there is also a general efficiency argument for 
subsidies for tertiary (and not only primary and secondary) education because of various 
externalities of having an educated labor force. 
Although China today devotes only just over one (probably 1.3) percent of GDP to 
research and development (R&D), the rate of increase in such investments has been 
impressive during the last decade (the corresponding figure being 0.6 percent in 1996). 
Indeed, the new Five-Year Plan of 2006 has the explicit goal of raising the R&D spending 

---

72
share of GDP to 2 percent and beyond. It is also likely that R&D spending will be better 
and more extensively applied in the future, since the share of such spending financed by 
enterprises (rather than by the government) recently seems to have increased 
considerably.
139
This may help explain why R&D spending and innovation in China 
during the last decade have become more labor-intensive and less capital- and energy-
using than earlier (Jefferson, 2005; Jefferson, Su and Zhang, 2004).
140
There is, however, 
a large regional concentration of such spending, predominantly to eastern regions (about 
70 percent of total spending) and, indeed, to four large cities (Beijing, Guangdong, 
Jiangsu and Shanghai). This will contribute to preserve the large regional differences in 
per capita GDP. 
Clearly, non-mainland firms have contributed substantially to the technological and 
organizational progress in China in recent decades. It seems, however, that this has taken 
place through direct import of technology rather than through R&D activities by these 
firms in mainland China.
141
However, non-mainland firms seem to stimulate domestic 
R&D spending indirectly.
142
One mechanism may be that they contribute to competitive 
pressure, another that they simplify the access to foreign technology for domestic firms 
through various types of spillovers, for which there is clear evidence.
143
In spite of fast technological progress in some sectors and region, China has, of course, a 
long way to go with regard both to the production and the actual implementation of R&D.