中国减税的经济效应评估——基于所得税分享改革“准自然试验”

The Effect of Tax Changes in China: Evidence from the Income Tax Sharing Reform as a Quasi-experiment

在由高速增长阶段转向高质量增长阶段后,减税能否提振宏观经济,助推高质量发展,引起了社会各界的广泛关注。基于这一背景,本文采用规模以上工业企业数据评估了中国税率波动的经济效应。利用2002年所得税分享改革"准自然实验",本文采用理论和断点回归模型两个策略证实了,企业所得税由国、地税局分别征管的二元体制,导致改革前成立旧企业的实际有效所得税率约比改革后成立的新企业低1.99%。进一步的检验显示,旧企业劳均增加值增速约比新企业高6.43%。简单核算表明,税率每下降1%,企业劳均增加值增速约提高3.2%。这显示减税的作用值得期待。

Summary： Since China＇s economic growth started slowing in 2012, ways to boost the economy and promote high quality development have gained more and more attention from all sectors of society. The central government has focused on building a powerful new driving force for economic growth by accelerating supply-side structural reforms. Lowering costs is an important part of these reforms. As taxes account for a significant share of enterprise costs in China, many people have called for tax cuts. Others have questioned the scope of the tax cuts and whether they will have the desired impact. This dispute reached a climax when the Trump Administration introduced its plans for tax cuts. The key to resolving this debate and choosing the appropriate policies is to precisely evaluate the effect of tax cuts in China. The recent literature focuses on the evaluation of value-added tax （VAT） reform but has yet to come to a concrete answer. China started to reform its VAT system in six industries in the northeastern region in 2004. Reform gradually spread to the whole country and to the manufacturing industry in 2009. The VAT reform allows firms to deduct fixed assets from their taxation base. Some empirical studies use China＇s VAT reform as an identification strategy to explore the effect of tax incentives but face two limitations. First, many of the studies may be biased as the timing, industries and locations of VAT reform are endogenous. Second, most studies focus on the tax incentives for investment, employment and export. These are partial effects and do not account for the aggregate effect on firms＇ output or efficiency. While the aggregate effect of tax reduction on the economy is drawing more attention （ see, for example, Romer ＆ Romer, 2010, who use time series data to measure the U.S. tax multiplier） , the macroeconomic effects of tax changes in China have been ignored. This paper uses the 2002 Income Tax Sharing Reform to identify the aggregate effect of tax reform on firms＇ value-added per capita. The reform assigns responsibility for the taxation of enterprises to two different tax agencies according to a firm＇s establishment date： the National Tax Bureau for enterprises registered on or after January 1, 2002, and the Local Tax Bureau otherwise. Local tax bureaus share the same interests as the local government （of which they are a part） , while the National Tax Bureau is supervised by the central government and is therefore more independent from local governments. Theoretically, enterprises established before January 1, 2002 are expected to enjoy a lower tax burden, as local governments and local tax bureaus should set lower tax rates to keep locally registered companies from moving to other regions. Using a sharp regression discontinuity design based on the above-scale industrial firm-level data from 2003 to 2007, which are maintained by the National Bureau of Statistics of China, this paper finds support for the hypothesis that the effective tax rate for enterprises set by local tax bureaus is 1.99% lower than the rate set by the National Tax Bureau. Further analysis concludes that the growth rate of value-added per capita for firms established before January 1, 2002 is 6.43% higher than for firms established after December 31, 2001. This finding implies that a one-percentage-point tax rate decrease would raise firms＇ value-added per capita growth rate by 3.5% on average. This paper makes two major contributions. First, unlike other studies of the partial effects of tax incentives on enterprises, this paper studies the overall impact of tax cuts on firms＇ value-added growth rates and finds a positive correlation between the two factors; that is, tax cuts increase companies＇ growth rates. Because firms are the most important drivers of economic growth and value-added statistics are a key element of GDP, this paper provides strong support for the argument that tax cuts promote high quality economic growth. Second, using a sharp regression discontinuity design to analyze the 2002 Income Tax Sharing Reform, this paper arrives at a much more reliable estimate, as it avoids the limits imposed by selection bias in studies that investigate VAT reforms. This is different from all other research on the effect of China＇s tax changes.