Why China's dollar pile has to shrink (relatively soon)

James Fallows has a piece in the February, 2008 Atlantic on what he calls “The $1.4 Trillion Question” – why China continues to accumulate $1 billion a day in relatively low-return American assets (mostly Treasury bills), why this can’t go on forever, and what it could mean if this pattern of investment ends abruptly rather than slowly. On the whole, it’s a good introduction, with some useful background on the people responsible for making the central government’s investment decisions. (The point that one of the two key figures, unlike his counterparts almost anywhere else, has never invested for himself, or even bought a house, is a nice touch.) I think the article overdoes its emphasis on a lack of transparency in China – the way in which sub-prime mortgages were re-packaged as “AAA” securities has made clear that the American financial markets China has been investing in aren’t always that transparent, either – but that’s a matter of tone and emphasis. What the article is missing, I think, are two important pieces of demographic and historical perspective, which help illustrate the pressures on the government. Fallows spends a fair amount of time on changes in China’s mood that may be real but are hard to get a handle on -- e.g. greater awareness among the population that their investments in the US are not earning much money (and some high profile ones have been outright losers, like the widely-publicized investment in the Blackstone Group) and that this is money that could be used to better things at home – and speculations about how much the government wants to, or can, continue resisting those popular desires in the interests of keeping inflation low, etc. I think the big story is more structural than that.

First the demography. Here the key point is one of the great under-played China stories : the rapid aging of the Chinese population. For roughly 30 years now, China has had compulsory birth control of various sorts, and (as most people reading this probably know) its birth rates declined at a rate that has very few historical parallels. So while the number of young people entering the work force every year has remained quite high until recently (China had so many births in the 1950s and 1960s that even with them having relatively few children per couple when they grew up, birth rates per 1,000 population stayed high into the late 1980s), the percentage of children in the population became quite low. Meanwhile, because Chinese death rates were very high before the Revolution, and stayed pretty high into the mid-1960s, there were also relatively few old people. So what economists call the “dependency ratio” – the ratio of people in the labor force to people whom workers need to support – has been extremely favorable for China over the last couple of decades:it's now at about 2 workers per non-worker, versus about 1:1 for the U.S. But that is now changing pretty quickly (thanks mostly to public health improvements under Mao)and China will soon have a fairly old population; by 2030, it will have as high a percentage of old people as countries like Italy and Germany today, whose pension problems, etc., you read about periodically. [Some of the best work on this is by my UC Irvine colleague Wang Feng and Andrew Mason at the University of Hawaii – their paper in a newly published Cambridge Press book – China’s Great Economic Transformation, edited by Loren Brandt and Thomas Rawski, is well worth a look, though the book won't be available for a couple more months.] China's dependency ratio will probably reach today's global average by 2020, and the current U.S. level of 1:1 by 2030.

A country with a higher ratio of dependents to workers –like a family in similar circumstances -- simply cannot save at the same rate as a country with relatively few dependents, no matter what the government may want to do and how many provisions it has to siphon the dollars China’s exports earn out of the economy and into a massive national savings account. And since China also has plenty of investment needs , as Fallows emphasizes – for schools, hospitals, sewers, you name it – it is likely to start spending down its dollar hoard before too long, no matter what happens in US-Chinese negotiations. Its true that both sides recognize the dangers of this happening too fast – leading to a run on the dollar and the collapse of China’s biggest market –but the pressures for it to at least start happening soon are even stronger than Fallows lets on.

That brings us to the history. China, like Japan and Taiwan before it, differs from Europe and the US in having undergone very substantial industrialization before its countryside began to empty out. (Japan’s rural population kept rising in absolute terms until World War II; China’s until roughly 1998.) Thus they were quite industrial before they were heavily urban, in part because they had lots of industry in the countryside. (Think of China’s Township and Village Enterprises.) Even today, China has a lower percentage of its population in cities than Britain had in 1840. There are all sorts of reasons for this – and anyone who becomes a loyal reader of my posts will eventually hear about them ad nauseam; but it is likely that in China, as in Japan, this will end with a period of extremely rapid urbanization. This rapid urbanization is now really getting underway (you ain’t seen nothin’ yet!), as rural industrial job creation slows to a crawl (as it now has) and the rural urban income gap becomes so large that even with many barriers to migration remaining, many more people will pick up and leave. So far, China’s urbanization rate pretty closely tracks Japan’s, with a 50 year lag – and beginning in the mid-1950s, Japan went from about 35% urban to about 70% urban in less than 20 years. Most people think China is poised to do the same – which will require China’s cities to grow by roughly the total population of the US and Mexico combined by 2030.

And here’s the rub. The Chinese government has worked very hard to avoid creating the kinds of slums that ring Mexico City, Manila, Cairo, etc . In fact, this has been one of the few real continuities in policy between pre- and post-1978, though the tools used to insure this -- outright prohibition of migration, guaranteeing land allocations, encouragement of rural industry, phasing out land taxes, various local policies that deny rural migrants access to urban services, etc. – have been an ever-changing mix. To a great extent they’ve been successful in meeting this goal: certainly there are grim communities in Chinese cities, but the numbers of people lacking access to electricity hook-ups, running water (of whatever quality), etc., is quite low by “third world” standards. This matters, among other things, for social and political stability. Maintaining this record as urbanization accelerates will require huge amounts of investment.

Meanwhile, even though the number of new job-seekers entering the labor force each year is now declining, China can’t really afford to see job creation slow down, because there is still a lot of labor to be absorbed. To go back to the Japan comparison, when Japan’s phase of very rapid urbanization began in the 1950s, its unemployment rate was around 2%, so even though people newly arrived in the cities faced crowding and other ills, they all had jobs. Nobody knows for sure what China’s urban unemployment rate is, but 15% seems like a plausible ball-park estimate. So job growth has to keep going, and presumably, most of that growth has to be making things and providing services for people in China. And that means a lot of the money now abroad has to come home – no matter how much, or little, resentment grows over China subsidizing U.S. over-consumption, or American backlash against Chinese ownership of U.S. assets. Nonetheless, Fallows has the main point right -- whether this happens smoothly or abruptly, and on what timetable, has enormous implications.