Chinese Law Prof Blog

Editor: Donald C. Clarke
George Washington University Law School

Tuesday, March 5, 2013

Ghost Cities: What Do They Mean?

First ghost weddings; now ghost cities.

Ghost cities in China have been in the news a lot lately. The term refers to places like Ordos in Inner Mongolia, where a vast expanse of residential and office buildings lies uninhabited, the product of somebody’s miscalculation as to where people would like to live and work.

Ghost cities have often been Exhibit A in various arguments about China’s economic problems, in particular the argument that the system massively misallocates investment. But wherever there is conventional wisdom, there’s going to be contrarian pushback because it’s always fun to challenge conventional wisdom. Nevertheless, I don’t find two recent examples terribly convincing.

First, the contrarians have been talking up a recent report by Jonathan Anderson of Emerging Advisors Group provocatively entitled “Hurray for Ghost Cities.” I don’t have a copy of the report, but according to a description of it at the Wall Street Journal’s China Real Time Report, it is less than a resounding hurray. In fact, the argument (again, as reported) seems to be only that ghost cities aren’t quite as bad as you think, because if the money hadn’t been invested in ghost cities, it might have been invested in creating overcapacity in industry, which would have been much worse for the economy:

His point is that by investing in “ghost cities” to underpin growth, China saved itself from even more unwise overinvestment in areas that could have done lasting damage to the economy, such as manufacturing.

[text omitted]

They’ve certainly been a black hole, he says, but a hole that has emptied largely into the equally dark vaults of China’s state-owned banks, where bad debts can remain buried for a long time.

“Lesson learned: If you’re going to waste capital best to waste it completely, where it will do the least damage to everyone else,” writes Mr. Anderson.

Or, to put it another way, he offers: “Why truly crap investment projects help ‘save’ China.”

As George Will would say, in a paragraph all its own: "Well."

Anderson’s analysis in no way undermines the idea that ghost cities represent a massive misallocation of investment. If the money had been spent on digging holes and then filling them up again, Anderson could equally well have written “Hurray for Holes in the Ground.” Anyone who cites this piece as support for the contrarian case hasn’t read beyond the headline.

Second, in a recent column in the New York Times, Bill Bishop cites Tom Miller’s book “China’s Urban Billion” as being dismissive of the “ghost cities” problem. Here’s what Miller writes, as quoted by Bishop:

The truth of the matter is that China is not building too many apartments, and a handful of empty urban districts are not evidence of a giant property bubble. Chinese property investment may be inefficient, but it is sustained by a huge, growing and sustainable demand for new housing. …

China’s current modern housing stock, defined as homes with individual bathrooms and kitchens, is around 150 million units. But 200 million migrant workers currently live in dormitories or slum housing. If one believes that the urban poor deserve to live in proper flats, the corollary is that Chinese cities actually have a significant shortage of housing – somewhere in the region of 70 million units. China is not building too many new apartments; it is building too few.

My first comment is about what Bill Bishop says, with which (if I understand him correctly) I must respectfully disagree. I can’t see how what Miller says counts as dismissive of the ghost cities problem. Chinese cities may well have a significant shortage of housing. But the fact that people may want housing in Beijing or Shanghai or any one of hundreds of other cities does not mean that they want it in Ordos. Of course it is silly to argue that overinvestment in Ordos means there must be a housing bubble everywhere in China, but it would be equally silly to argue that underinvestment in some places means that the phenomenon of ghost cities can be dismissed as unimportant. To be clear, I don’t read Miller as making that argument in this excerpt. That’s why I disagree with Bishop’s citing Miller in support of the contrarian thesis.

My second comment is on what Miller says here. The argument that there is overinvestment in housing, as I understand it, does not address the issue of who deserves what. It addresses the issue of what kind of return one can expect from the investment. If it costs $10 million to build a block of 100 “proper flats” for the urban poor, the project won’t even break even unless the urban poor have $100,000 per family to spend on flats. If they don’t, then we have to ask where the money will come from. Economic analysis cannot, of course, answer the political question of whether such housing should be subsidized in some way, but it’s useful for thinking about the question of whether it will need to be subsidized. Merely asserting that poor people will need housing, and deserve it, doesn’t help us answer the question of whether housing developers will make back their investment, and therefore whether the banks from which they borrow will or will not get stuck with a big pile of non-performing loans.

If the housing is intended for a class of people who can afford it, then that is of course a different story. But Miller’s argument that China has too little housing is premised on the existence of precisely that group that can’t currently afford a decent flat.

I should add that I can’t agree with the way Miller loads the emotional dice here: apparently, if you think there’s overinvestment in housing in China, you’re not only wrong on the economics, but you’re also a bad person who thinks that the urban poor don’t deserve to live in proper flats. I don’t see this as a useful way to start a discussion of a complex issue.

Commentary | Permalink

TrackBack URL for this entry:

Listed below are links to weblogs that reference Ghost Cities: What Do They Mean?:


Post a comment