Abstract: Proponents of the PublicPrivate Partnerships (PPP) argue that PPP can decrease financial pressures, diversify investment risks, and improve resource allocation efficiency for local governments. In contrast, opponents argue that a large number of government guarantees and subsidies are often embedded in PPP, encouraging corruption and rentseeking. The reduction of fiscal pressure within the budget is only a financial illusion, which invariably accumulates systemic risks. This paper uses the bond issuance data of China during 20152020 and the information of PPP projects in various places to test whether or not the interest rates of local public bonds respond to the large scale implementation of PPP, with local public bonds as the treated and national debt and financial corporates' bonds as the control under a Difference in Differences setup. The empirical results show that the sharp increase in PPP projects around 20162017 raises the interest rates of local government bonds, indicating interest penalties from the financial market. Specifically, the risk premium is 6.729.7 basis points for local bonds with a 5year maturity, while the risk premium is 2141 basis points for 10year bonds, and the effects are heterogeneous across regions with more and fewer PPP projects. The results are robust to a battery of tests, including changing the control group, postponing PPP projects actual time to take effect on local public finance, and certain placebo tests. Finally, possible explanations to the empirical findings are discussed, and some feasible measures are proposed to promote the sound development of PPP.

Key words:  PPP, local fiscal risks, bond market, difference in differences

摘要： 政府与社会资本合作（PPP）的倡导者认为PPP能够为地方政府分担财政压力，分散投资风险，提高资源配置效率；其反对者则认为PPP背后隐藏着政府的隐性担保和补贴、腐败和寻租，预算内的财政压力减小只是一种财政幻觉，反倒无形中累积了系统性风险。本文利用中国2015—2020年间的债券发行数据以及PPP在各地的落地信息，以地方政府债券为实验组，以国债和金融债为控制组，构建双重差分模型，检验PPP集中落地是否增加了地方财政风险，进而抬升了地方政府债券的发行成本。实证结果表明，2016—2017年左右的PPP集中落地使得地方政府债券发行受到来自市场的利率惩罚，5年期地方债发行利率的风险溢价为67个基点~297个基点，10年期地方债发行利率的风险溢价为21个基点~41个基点，且这种效应在PPP项目较多和较少的地区之间存在异质性。更换控制组、推后PPP项目对地方财政的实际冲击时间、假设虚假的实验组等一系列稳健性检验证明这一发现具有可信性。此外，本文还讨论了这一实证结果背后的可能原因，并提出促进PPP良性发展的措施。

关键词: 政府与社会资本合作, 地方财政风险, 债券市场, 双重差分