China’s struggles to deal with its mounting burden of local government debt continue to make headlines, giving the dry topic of intergovernmental fiscal relations an unusual amount of relevance these days. The proximate cause of all this local government debt is obvious enough: when the world was falling apart in late 2008, the central government wanted to do a big stimulus, but didn’t want to pay for it. So instead it told local governments to spend whatever they needed on infrastructure to get growth going, and told banks to lend the local governments whatever they needed, and turned a blind eye to any irregularities in all this borrowing. But many scholars and analysts argue that the roots of the problem lie even deeper, that they are the result of a system where there is a fundamental mismatch between the huge burden of public services that local governments have to provide, and the scanty revenues the central government permits them to raise. It’s not just in infrastructure that Beijing pushes the burden down on local governments, the argument goes, but in schools and hospitals and so on. Exhibit 1 in this argument is usually some variant of this chart: From these figures, it is easy to get the impression that China’s central government is just rolling in money and not spending any of it–so no wonder local governments have to borrow so much just to make ends meet! Yet this impression is totally false. What is the central government doing with all this revenue? Simple: giving it back to local governments. Of the 6.45 trillion RMB in revenue that the central government collected in 2014, it sent 5.16 trillion back to local governments as transfers (it also spent 2.25 trillion itself, so it was running a deficit). These enormous transfers account for about 60% of total local government revenue, and are the main method by which money is redistributed from richer provinces to poorer ones. If these transfers are treated as what they really are–the main source of local government revenue–then the apparent mismatch disappears: So while it is fashionable these days to blame excessive centralization of revenues for the mess in Chinese local government finances, this clearly is not a sufficient explanation. Yet it is equally clearly the case that there are real problems in local government finances, and that many localities have to resort to extra-budgetary or extra-legal means to raise money, for infrastructure projects as well as other spending. So how to understand this? As Linda Chelan Li and Zhenjie Yang put the question in a recent article (available for free for a limited time):

The dominant view is that excessive centralization of revenues and decentralization of expenditure responsibilities have precipitated a fiscal crisis in many Chinese counties and townships, with dire consequences to local governance. The ‘gap’ argument emphasizes the relative ratio of central vis-a-vis local revenues and the impact of decentralization of expenditure, and slights the impacts of other parallel fiscal developments, in particular the large flows of central subsidies to local coffers since the 1994 tax sharing reform. As pointed out by the few sceptics, the presence of a large and growing central fiscal subsidy, of a comparable size to the centralized tax revenues, means that logically the latter cannot in itself constitute a sufficient condition for local fiscal difficulties. What then accounts for the difficulties, as localities are not blatantly short of monies?

The answer is that there is not one thing called local government in China: there are multiple levels of local government (at least three). Money does not just have to flow from the central government to local government, but between different levels of local governments. And it is easy to see that this flow might not always be smooth, or that “leakage” may happen along the way as different officials get their hands on the funds. This in fact is the key problem: the central government sends plenty of money to local governments in aggregate, but it does not end up in the right places. The lowest-level governments (counties) are where spending responsibilities are concentrated (see table below), but they are the furthest away from the flow of money from the top. And since responsibility for spending is often assigned without regard to where the revenues come from, mismatches abound. So the lowest-level governments do in fact have many fiscal obligations and often not enough resources to meet them. The better recent research on China recognizes this more complex reality. Li and Yang’s article demonstrates it through a case-study approach that documents how higher-level local governments undercut those below them, while the OECD’s urban policy review of China has a more data-driven presentation (see pages 189-212). The issue is not that the central government does not give enough money to local governments; it is that the money does not go to the right local governments. As the OECD notes, transfers are often made based on assessments of need that do not match up well with reality:

One of the limitations of the Chinese system is that the need for transfers is assessed mostly based on registered rather than actual population in a province. The problem is that the actual population is generally lower than registered population in low-income provinces, given that migrants remain registered in their home province, regardless of where they live. The government is set to henceforth include 15% of the difference between actual and registered population in the formula for determining transfers. This will partly take into account the cost of migrants to a province.

It’s understandable why the simpler account of a central-local fiscal imbalance dominates the discussion: “the central government has too much money” is a much better slogan than “the fiscal transfer system is suboptimal.” But the first the step in fixing the problem is coming to an accurate diagnosis.