Politicians and Firms: Evidence from Politician Rotations in China

The interaction between politicians and firms is a central issue in an economy. Prior studies highlight that firms with connections to the government officials enjoy the preferential treatments, such as access to finance and government concessions. Yet, it is less understood that whether and to what extent the governing style of local top politicians shapes corporate policies and the underlying mechanisms. This project proposes a method to explicitly quantify the impact of the top local politicians on corporate behaviors.

The method is based on the unique politician rotation system in China that generates a large variation in politician turnovers at different government levels (province, city, and county) and a census data of all above-scale manufacturing firms. A politician-firm matched panel data set is constructed to track a list of local politicians and corresponding firms governed by the politicians. Controlling for both observable and unobservable differences across firms, this study attempts to evaluate the similarity of corporate behaviors of firms operating in different regions but governed by the same politician at different time periods. This study will shed light on whether “politician fixed effects” matter for a wide range of corporate decisions, such as investments, employment, leverage, exporting, innovation and land transactions. For each corporate policy of interest, the model will generate a measure of the additional explanatory power of the “politician fixed effect” for each government-level of politicians.

The project will dig into the underlying channels of the politician effects. First, by examining the heterogeneity of the politician effects among SEOs affiliated to different levels of government, private firms and foreign firms, this project attempts to uncover how politician affect the corporate behavior through sending commands to SOEs and the resulting spillover effects. Second, this study will identify the heterogeneous effects of politicians according to the quality of regional institution. The project will explore whether politician effects will be exaggerated by weak property right protection and severe government intervention. Last, this project will try to link the politician effects to the observable politician characteristics. The hand-collected dataset on the detailed demographics (birthplace, education etc.) of politicians facilitates an analysis on whether early-life experience (such as exposure to violence) and foreign education matter for governing style.

Quantifying the magnitude of the politician effects on firm behavior, this proposed research uncovers the black box of the relationship between politicians and firms and has policy implications for the design of the political promotion system.