中文摘要
地方辖区能够多大程度上自由调整其税收和支出？本文通过研究针对永久性外生收入变动的财政调整，提出一种检验市级政府实际财政自主程度的实证方法。基于一个辖区支出受到部分约束的税收竞争模型，我推导出一项可检验的预测：对于一个幅度较小的正向收入冲击，如果地方决策者面临较强（较弱）的财政约束，那么感知到的税基流动性越高，减税幅度就越大（越小）。我利用瑞士某州市际转移支付制度改革所产生的收入冲击实施这一检验，并以已划定用途的土地储备的可获得性作为税基流动性的代理变量。我发现，住宅用地可获得性越高，税率响应越强，但工业用地储备的结果在统计上不显著。结合理论来看，这表明地方辖区的实际财政自主程度较低。基于公共账户的常用财政分权指标可能高估地方政府的实际自主性。
Abstract
How freely can local jurisdictions change their taxes and spending? I propose an empirical test of the effective degree of municipal fiscal autonomy by studying fiscal adjustments to a permanent exogenous revenue shift. Based on a tax competition model where jurisdictions are partially expenditure constrained, I derive a testable prediction: tax cuts from a small positive revenue shock will be larger (smaller) with higher perceived tax base mobility, if the local policymaker is strongly (weakly) fiscally constrained. I apply this test using a revenue shock generated by a reform in an inter-municipal transfer system within a Swiss canton. tax base mobility is proxied by the availability of zoned land reserves. I find that higher residential land availability is associated with stronger tax rate responses, but I find no statistically significant results for industrial land reserves. In light of the theory, this suggests that the effective degree of fiscal autonomy of local jurisdictions is low. Usual indicators of fiscal decentralization based on public accounts might overestimate the actual autonomy of local governments.