政府回应是否被定价：来自城投债信用利差的证据

Abstract

Improving government responsiveness requires long-term effort, creating a tension between immediate costs and delayed incentives that may discourage local governments. Based on massive public messages from the Local Leadership Message Board (LLMB), this paper uses natural language processing models to construct measures of government responsiveness and examines whether capital markets provide immediate incentives for improvement. The results show that better government responsiveness significantly reduces the credit spreads of chengtou bonds, indicating that local governments' efforts are recognized by capital markets and translated into lower financing costs. Mechanism analysis suggests that responsiveness lowers credit spreads through enhancing local governments' organizational reputation and fostering a clean institutional environment. Further analysis shows that the suppressing effect is more pronounced for messages complaining about government agencies, those involving complex public demands, and in regions without central inspections or with weaker audit oversight. This paper demonstrates that capital markets can provide immediate incentives for local governments that actively respond to public appeals, which is conducive to improving government governance capacity.