Abstract: China is a super large country with multi-layers governance structure, how to achieve the governance efficacy and efficiency is a key issue in state governance. County and township level governments are the nerve endings of the governance system, and their governance performance is vital for the overall effect of state governance. However, since the 1990s, the widespread fiscal distress of grassroots governments has resulted in a serious shortage of public service. In addition, in order to cope with new reform tasks, policy shelving and vague implementation have become typical phenomena of grassroots governance. State governance has fallen into the “last kilometer” dilemma. In order to relieve the fiscal distress of counties and townships, the central government not only actively promotes the reform of the fiscal management system, such as “County Directly Administrated by Province” and “Township Finance Supervised by County”, but also focuses on optimizing the organizational structure of grass-roots governments, such as large-scale townships consolidation, in an effort to cut government operating costs and fiscal expenditure. County-level government consolidation has been on the rise, especially after 2009, the frequency of consolidation has accelerated significantly.
Academia has conducted a series of studies on the characteristics, causes and consequences of grass-roots government consolidation. But there is rare empirical evidence related to the spending-cut effect of consolidation. Besides, the existing research makes use of large sample data to observe the average effect, which fails to identify the effect in a particular scenario. This paper conducts a comparative case study using two consolidation cases of “Wansheng district-Qijiang county” and “Shuangqiao district-Dazu county”, which were implemented synchronously in Chongqing Municipality in 2011. By adopting the synthetic control method to construct a counterfactual analysis framework, we find that the fiscal expenditure effect of the two mergers is completely opposite. After the merger of “Wansheng district-Qijiang county”, the fiscal expenditure is cut by 11% on average, but the merger of “Shuangqiao district-Dazu county” boosts the expansion of fiscal spending. The huge different effect is related to the initial characteristics of the consolidated administrative zones. There are huge gaps between Shuangqiao district and Dazu county, such as the development level and the population size. Therefore, the consolidation cost is too high, and it's also hard to obtain scale economies in public resources input and management. Contrarily, the size and the development level of Wansheng district and Qijiang county are similar, thus the internal friction cost after the consolidation is small.
Based on these findings, we believe that selecting appropriate consolidation targets is an important prerequisite to ensure the spending-cut effect. Firstly, it is inappropriate to select administrative districts with significant contrasts to consolidate, because it often requires additional fiscal resources to bridge the gap. The public pool effect is also likely to stimulate the free ride motivation of the weak district. Secondly, it is not suitable to consolidate administrative districts with large size disparities, because the scale economies can be very limited. From the past experience of county-level district consolidation, the central city and peripheral county consolidation is a common path, but this type of consolidation needs to consider the above two issues most. The central cities tend to be much smaller than peripheral counties, and their modernization level is usually significantly ahead of the latter. Our suggestion is that if the scale of central cities is too small, we can consolidate these small-scale central cities first, and then push forward the consolidation with the peripheral counties.

Key words: Government Consolidation, Fiscal Spending, Scale Economies, Synthetic Control Method

摘要： 优化政府组织结构是降低政府运行成本、增强国家治理效能的重要路径。本文选取重庆市2011年同步实施的“万盛区—綦江县”和“双桥区—大足县”两个合并案例进行比较研究。通过采用合成控制法构建反事实分析框架,发现两项合并的财政支出效应截然相反,“万盛区—綦江县”合并后五年平均削减了约11%的财政支出,但“双桥区—大足县”合并反而推升了财政支出规模。迥异的合并效果与被合并政区的初始特征有关,双桥区与大足县的发展水平及辖区规模均存在显著差异,致使合并面临较高的整合成本,合并后辖区规模的小幅增加也难以发挥公共资源投入和管理的规模经济,而万盛区与綦江县较小的发展差距及规模差距,使得合并后辖区内部摩擦成本小,同时能够较充分发挥规模经济优势。如何选取适宜的合并对象是未来推进县级政府合并需要考虑的一个重要问题。

关键词: 政府合并, 财政支出, 规模经济, 合成控制法