This week I returned from a trip to Shanghai and Nanchang; the latter is in Jiangxi province, and as a GMU aside, the food in Jiangxi was stunning.  I was in Nanchang to give some lectures at Jiangxi University of Finance and Economics, a very enjoyable experience.  
But I’m here to talk about China’s empty buildings: The ones everyone sees when they drive through China’s growing cities, the ones that show up in YouTube videos.  
Nanchang, a city of about 5 million, has an entire new central business district that looks fairly empty: Early in the evening, the 40-story apartment buildings are mostly dark dark dark.  
A sign of rampant speculation?  A sign of home-flipping, with apartments used like the tradable tokens in one of Vernon Smith’s bubble experiments?  Or perhaps the best available protection against inflation in a country with thin financial markets and where it’s tough to get more than 4% in a savings account?  
I’m guessing there’s something to all these stories, and the people I spoke with last week–business school students, finance professors, a few investment bankers–tended to agree.  
Then I heard a new story–new to me anyway, but let me know in the comments if this tale is an old one. It’s just based on a few anecdotes, so the sample size is small, but perhaps others out there have observed the same phenomenon. 
Here’s the story: Two different people told me that their families own a second partly furnished home that nobody lives in but that they fully plan to move in to at some point.  And in both cases, the reason for not moving in was the same: Family ties.  And again in both cases, they could tell stories of other families in the same situation.  So family rigidities might be a cause of ghost towns.  
In the first anecdote, it was important to stay close to the grandparents who provided childcare for a newborn girl.  In the second anecdote, a child was going to college in the central city and the parents didn’t want to part from the college-age child, so the family stayed together in the old central city apartment.  Both examples of family rigidities.  
Now, let’s glibly theorize about what might be going on if this really is a broader phenomenon (a big if).  How do family ties keep the family from moving to the new home?  One reason is logistics combined with the savings motive.  The new suburbs are nice but they’re far from the true city center–so why not wait until other people move in, until the shopping and maybe even the jobs have moved out there?  This is a twist on Peter Gordon’s general theory of urban sprawl.  If even one family member feels like staying it might be reasonable to stay.  In the meantime, the house can serve as a savings vehicle.  You can see how if most people decide to wait until the place is half-filled, that channel alone can create ghost towns.  Nobody wants to be the first mover.  
But here’s a more public choice oriented channel: Multigenerational familiesstill common in China through some combination of tradition and grandchild scarcity (and perhaps thin financial markets)–create veto points that are less common in North America.  If the grandparents don’t want to move–particularly when grandparents are the family’s net savers—that reduces the chance of the whole family moving.  The golden rule at work.  
When key players disagree on what to do next, the status quo will probably win out.  In the meantime, it’s reasonable to use the house as a savings vehicle.  Agreement will eventually come, through death, through renegotiation, through better amenities in the ghost towns, but that takes years.  And that means the ghost towns will take years to fill up.  
It turns out Nanchang’s ghost town–apparently designed in Singapore–wasn’t really empty at all.  Active retail had started to move in, the 40-story apartment building I toured had a modestly filled parking lot, and there was even a little foot traffic in the area at 9pm.  
Of course, even if the ghost town had been a purely irrational bubble the market would eventually clear at some price, so the fact that old apartments were eventually selling isn’t evidence against a past bubble.  But by talking with people, asking their stories, I learned about a channel that sounded less like liquidity and more like rigidity.  
Liquidity helps create bubbles.  Rigidity, I would claim, helps to tamp them down.  Customers who plan to eventually live in the apartment they’ve bought, customers who have already furnished it with their own family photos and knick knacks, are more likely using cool, rational, System 2 forms of thought, the kind of thinking that tamps down on bubbles.  
There surely have been bubbly tendencies in the Chinese housing market, but if family rigidities are important, those rigidities tend to reduce them–and help explain why these new cities are empty for so long.