The latest issue of the Journal of Economic Perspectives has a good group of articles on issues in the Chinese economy; there’s a lot to talk about in there, but the piece on education by Hongbin Li, Prashant Loyalka, Scott Rozelle, and Binzhen Wu is particularly worth flagging. It touches on one of the hotter social debates in China over the past few years: whether the massive expansion of college education since 1999 has created an over-supply of graduates, or is just the beginning of the necessary transformation of the education system to meet the needs of a modern economy.

This debate is interesting not only because it is a very consequential one, but also because the two sides tends to use very different styles of argument. The case for the prosecution tends to rely more on close observation of current social phenomena in China (what you might call anecdotal evidence), while the case for the defense tends to rely more on economic theory. A good example of the argument for an education glut is a recent piece by Edoardo Campanella:

Education is never a bad thing in itself, but the move toward “mass universities” of the type that emerged in the West after World War II is occurring too fast. …

China, with a graduate unemployment rate of 16%, is producing more highly educated workers than the economy can absorb. The wage premium for workers with a bachelor’s degree has decreased by roughly 20% in recent years, and new graduates often must accept jobs – such as street cleaning – for which they are vastly overqualified.

As more Chinese students attend university, fewer are graduating from vocational schools, which teach the skills that the economy actually needs. In fact, the demand for qualified blue-collar employees is so high that in 2015 the country’s 23 million textile workers earned, on average, $645 per month – equal to the average college graduate.

Li et al. in the JEP piece note the same widely-reported factoids: that new graduates take a long time to find jobs, and their starting salaries are often of similar levels to manual laborers. But they counter with a combination of theoretical reasons not to be too concerned by these phenomena, and a more involved estimation of the financial returns to education:

In contrast to this common perception of too many college students, we believe that college expansion is a great policy achievement of China. If we assume that the demand for human capital is fixed in the short-run, then given the unprecedented increase in the supply of college graduates since 1999, it is not surprising that the return to college for young college graduates would decline for a time. However, in the long run, human capital investment can lead to investment in physical capital and skill-biased technological changes, which ultimately will increase the productivity of and return to human capital. In addition, regions and cities in developed nations that experience arguably exogenous shocks to the supply of human capital ultimately also experience increases in the productivity of skilled labor due to human capital spillovers. There is no obvious reason to expect that China’s case would be different in this respect.

Moreover, college expansion could well be a result of rising demand for human capital. Our analysis of data from China shows that the return to college education for the labor force as a whole has continued to rise despite the fast expansion of China’s colleges. In particular, the return for those with 5–20 years of work experience has risen from around 34 percent in 2000 to 41 percent in 2009. A possible reason is the rising demand for skilled workers driven by the influx of foreign direct investment and expansion of trade starting from the early 1990s. The high return to college education for experienced workers implies a high lifetime return (the 10-year lifespan return to college education for the year 2000 graduate cohort is as high as 42 percent), which explains why urban students flood into colleges in spite of the seemingly low short-term return.

My own impression is that the education-glut argument is more popular within China, perhaps because it can be more easily illustrated by tales of struggling new graduates. But the statistics that are usually used to support it seem questionable: if a recent college graduate is making the same wage in their first year of work as a migrant worker is making in their 20th, it’s not obvious that actually indicates the market is devaluing a university education. The proper measure is really the lifetime returns to education, and there seems little reason to doubt that today’s college graduates in China are going to have much higher lifetime incomes than today’s migrant workers without a degree. Perhaps the issue is that new graduates do not feel that the gap between themselves and manual workers is as wide as they expected it to be.

Li and his co-authors do point to some worrying evidence that the quality of higher education in China has in fact suffered as the number of students has massively expanded, an issue that Campanella also highlights. But while Campanella recommends making higher education much more restrictive and shunting most students into vocational education, Li and co. argue for decentralizing and deregulating higher education, so that universities are not mainly trying to meet government-set enrollment quotas but are instead competing to deliver a good educational experience.

A more serious problem than any over-supply of college graduates is likely to be the rather shocking under-provision of high school education for rural students, which the JEP article shows is weighing down the overall education level of China’s workforce.