The Chinese growth miracle:

[V]isitors to Chongqing marvel at the soaring skyscrapers and modern infrastructure built during Bo’s tenure there. But do they know that Bo’s administration borrowed the equivalent of more than 50% of local GDP to finance the construction binge, and that a large portion of the debt will go unpaid? Unfortunately, Bo’s case is not the exception in China, but the rule...

Inflating local growth numbers is so endemic that reported provincial GDP growth data, when added up, are always higher than the national growth data, a mathematical impossibility. And, even when they do not doctor the numbers, local officials can game the system in another way.

Because of their relatively short tenure in one position before promotion (less than three years, on average, for local mayors), Chinese officials are under enormous pressure to demonstrate their ability to produce economic results quickly. One sure way of doing so is to use financial leverage, typically by selling land or using land as collateral to borrow large sums of money from often-obliging state-owned banks, to finance massive infrastructure projects, as Bo did in Chongqing.

The result is promotion for such officials, because they have delivered quick GDP growth. But the economic and social costs are very high. Local governments are saddled with a mountain of debt and wasted investments, banks accumulate risky loans, and farmers lose their land.

From what I gather, government debt at the local level is excessively high, and the role of government owned banks is widely unappreciated.