Microeconomics Of Local Government Debt: Contracts, Risks And Regulations

This thesis attempts to make clear the microeconomic incentives of the financing and investment activities of local governments and their financing vehicles(LGFV henceforth) in China in the framework of incentive and contract theory, so as to reach a deep understanding of the potential debt crisis of local government sector that has been noted in recent years, and further, to give some suggestions for the solution of the crisis. The choice of this analytical framework is based on our fundamental judgments on the methodology of policy analysis: to gain a deep understanding of the basic logic of any economic problem, one need to precisely grasp the key optimization problem faced by related interest groups, including their objectives, feasible strategies, and institutional environments faced by them. Although such problems are usually analyzed in a reduced form due to their complexity, but reliable explanations and solutions are made only if key incentive mechanisms are remained unchanged. To this end, incentive and contract theory is indeed a suitable framework.It is also argued that one can grasp those key incentive mechanisms without contract theory. These arguments are very popular in Chinese academic circle, such as: the promotion incentive to local government officials leads to over-spending of infrastructure investment; local governments’ incentive to promote further economic growth due to the “debt-overhang” problem, etc. But unless the formal model has been written down, we are not assured that the detailed environment that allows any of these mechanisms to work actually exists. It is also left unknown whether other competing mechanisms might be working together. Hence those popular arguments mentioned above are one-sided on some particular mechanism, and have failed to provide complete, hence reliable solutions. In contrast, incentive and contract theory does better in the completeness of analysis, since critical environments are built into the model through parameters and constraints, so as to determine properties of the solution.Chapter 1 is the introduction of this dissertation; here we have widely arranged government documents and data on local government debt and LGFV from the State Council, the State Auditing Administration, The Ministry of Finance, and National Bureau of Statistics. Based on these materials, we first summarize the basic status of local government debt, including its magnitude, composition on different dimensions, and spending areas. Second, we introduced the background of current local government debt crisis from multiple perspectives, such as the task-budget-contradiction between central and local governments, the “land finance” mode of local governments, the long-term task of national-wide urbanization, and the shock from global financial crisis in 2008. Finally, we listed main challenges of the current debt crisis, including regional disparities of debt crisis, growth rates of GDP, budgetary income and government fund income are slowing down, etc.Chapter 2 reviewed existing academic investigations on the problem of local government debt, we classified them into 3 categories. The first category includes discussions on the possible causes of the debt crisis, one direct cause is the stimulate package released by central government after the global financial crisis in 2008, other causes are not so direct but more fundamental, which includes local government officials’ promotion incentives, growth competition between local governments, and the problem of soft budget constraint. The second category investigates potential risks of local government debt, to this end, several risk pre-warning models are built. The last category gives many suggestions for government debt management and to prevent and defuse systematic risks. We have summarized them into five sub-categories: marketization of LGFV, lending through legal system, financing from private sector, optimization of fiscal system, and strengthen budgetary descipline.Chapter 3 is built on the model of Holmstrom and Tirole(1998, 2011) to investigate the financing capacity and risk management of LGFV in China. We featured LGFV’s with the high ratio of governmental assets in their capital structure. By introducing imperfect information disclosure, it is shown that the most preferable source of external financing is debt; and further, although the existence of government asset improves the financing ability of the LGFV, it reinforces LGFV’s incentive to over-borrowing, hence increases its exposure to liquidity risk. The term “over-borrowing” describes the situation that government asset not only increase government debt through leverage financing, but also increases the leverage itself. To our limited knowledge, this is the first investigation that models local governments’ over-borrowing propensity in a corporation finance framework.Based on chapter 3’s main result, we estimate the impact of government asset on debt leverage(or, liability-to-asset ratio) using micro-level data from 1230 LGFVs’ financial statements, and the coefficient estimated supports our arguments. Then we estimate the model within sub-samples divided both by bureaucratic level and geographical area. Using the fixed effect model, we find that the main result hold in all the three bureaucratic levels—province, city and county—and three geographical areas—the east, the middle and the west. Among these results, one striking finding is that the lower the bureaucratic level, the stronger the propensity to over-borrowing.Chapter 5 analyses the strategic interactive behavior between central and local governments in a game-theoretic framework, it argues that the over-borrowing problem is the result of collective behavior of the latter. Suppose central government has no prior partiality to one local government over another, and local governments are maximizing their GDP, we find that when the central government makes policy under the situation of no commitment, local governments will behaves in a sense of collective “moral hazard”, in which all local governments will hoard an identical level of liquidity that is too low, and choose an identical scale of investment that is too high; If possible, each one will maximize its risk correlation with the others a whole, as a result of being shocked simultaneously. This makes the central government has no choice but makes costly effort to remedy the debt crisis ex post. We suggest that the central government to limit the scale of debt within an upper-bound. The consistency of polices should be strengthened, and large-scale stimulating plans should also be prudently used.Chapter 6 deals with the optimal incentive mechanism problem of LGFV’s choice of operating strategy in a framework of incomplete contract. We classified LGFVs into 3 categories according to “Audit Report on the Local Government Debt in China”, which is published by the State Audit Administration. The first category includes those LGFVs should be liquidated, the second includes those should transform their businesses to private sector, and the third includes those should continue their current business. We have devised optimal incentive mechanism for each category by imposing a gross debt cap on their balance sheets. The optimal mechanism can be used to deal with two sided moral hazard problem, and further, it is highly adjustable to macroeconomic shocks both ex ante and ex post. The feasibility of tractable adjustments is due to two features of LGFV’s, one of which is the high ratio of governmental assets in their capital structure; another is the existence of creditors whose interests are congruent to the local governments. We believe that these results are suggestive to the corporate governance management of LGFV’s. At the end of this chapter, we also provide the conditions and practice of central government intervention when macroeconomic shocks are large.Chapter 7 discussed the problem of delegating government authority when local governments have over-borrowing propensity in the framework of Aghion and Tirole(1997). We distinguish the delegation of formal authority from real authority, the former is the power to decide established by rules and laws of an organization, while the latter is the power to make effective decisions. By introducing the propensity of over-borrowing of local governments into the model, it is found that higher the debt correlation coefficient makes central government tend to delegate more formal and real authority, and the propensity to over-borrowing is positively related to incentives to delegate formal authority. Furthermore, the delegation of formal authority fueled the incentives to delegate real authority further; meanwhile, central government reinforced the effort to monitor the local government debt.In chapter 8, we first surveyed applications of the soft budget constraint theory to economic reform and the problem of local government debt, and then we argued that the theory has certain logic contradictions and lacks explanatory power when applied to practices of economic reform and government borrowing. We classified those problems that were formerly conglomerated under the name of “soft budge constrain” in to four sub-problems, namely, dynamic commitment, transaction cost, paternalism, and competitive rent-seeking, and point out that each of them has an original analytic framework. Finally, we illustrate how former conclusions to economic reform and local government debt drawn from the soft budget constraint perspective can be adjusted to form better interpretations under their original frameworks.