A2 An Analytical Model of Central and Local Governments
A2.1 The Soft Budget Constraint
It is well recognized that if local governments face soft budget constraints, they will have an incentive to over-spend, over-borrow, and/or pay insufficient attention to the quality of the investments that their borrowing finances. Such welfare deteriorating over-spending/borrowing can occur through the common pool mechanism. See, for example, Wildasin (1997, 2004), Goodspeed (2002), Akai and Sato (2005), and Boadway and Tremblay (2005) among others. That is, the standard result is that if the central government imposes soft budget constraints, inefficient too much investment should arise. On the other hand, Besfamille and Lockwood (2004) showed that hard budget constraints can be too hard and discourage investment that is socially efficient. Namely, they pointed out the possibility that the hard budget constraint over-incentives the soft budget constraint to provide effort by penalizing it too much for project failure, thus leading ultimately to the possibility that socially efficient projects may not be undertaken. Thus, welfare implications of soft budget constraint seem ambiguous. However, the conventional conjecture is that the soft budget is welfare deteriorating if public investment is too much, and vice versa.
We pay attention to the vertical externality of shared tax bases between the central and local governments in a real economy. Multileveled government normally means some commonality of tax base between central and local governments. As a result the tax base may overlap and shared tax bases create the common pool problem. It is now well recognized in the tax competition literature that such vertical externalities are likely to leave local taxes too high. This is because each local government unduly discounts the pressure on central government’s spending it creates by raising its own tax rate.
We develop a two-period intergovernmental financing model of two governments, the central government (or CG), the lower-level local government (or LG) in a small open economy, in order to explore how local public investment and wasteful spending may be stimulated under the soft-budget constraint. For simplicity, we consider the representative local government, and do not consider the free-riding and/or spillover effects within multi local governments. There are many papers to explore the horizontal and vertical externalities due to non-cooperative competition among multi local governments. See Wilson (1999) among others. In Japan many local governments often cooperate. The analytical results would be qualitatively almost the same even if we consider non-cooperative behavior of multi-local governments. Moreover, this is in particular a good approximation in Japan where many local governments behave cooperatively against the central government and their rent-seeking behavior may be summarized by the representative local government.
One contribution of this appendix is to show that the soft-budget outcome could occur even in the case of the representative local government where the central government intends to transfer between central and local governments to attain the optimal allocation of central and local public goods. This is a new result since the conventional literature on the soft budget normally assumes multi-local governments where the central government intends to transfer among local governments to attain the optimal allocation among local public goods. Moreover, the soft budget outcome may occur even if we assume away information asymmetry or cost sharing. In our framework rent seeking is crucial for the soft budget game. Another important contribution is that the soft budget is shown to be welfare deteriorating even if it may attain the first best level of public investment. Thus, our formulation captures important aspects of intergovernmental finance in Japan.