ABSTRACT
This article examines the effects of anti-corruption efforts on firm innovation in China. Based on the number of invention patent applications filed by firms, we find that firms located in areas with intense anti-corruption efforts are more innovative. We control for a broad set of firm and province characteristics, including industry-year fixed effects, and demonstrate the robustness of the results using an instrumental variable analysis, alternative proxies for anti-corruption intensity and innovation and a fixed-effect Poisson regression model. A heterogeneity analysis shows that the innovation-enhancing effects of anti-corruption efforts on the likelihood to apply for invention patents is greater for young companies and those without political connections, while no significant difference when conditional on having invention patent applications. Moreover, anti-corruption efforts have no significantly different effects between state-owned and private enterprises. Further analysis shows that anti-corruption efforts significantly increase investments in research and development and innovation efficiency, leading to enhanced productivity and firm growth in the following years.
Disclosure statement
No potential conflict of interest was reported by the author(s).
Data availability statement
The data of this study are available on request.
Notes
1 After 2017, the standards for calculating investigated and prosecuted cases of official crimes in the China Prosecutors Yearbook changed, making data after 2017 incompatible with earlier data. Therefore, when using relevant data, our sample includes listed companies from 2008 to 2017.
2 Invention Patent: This type of patent involves new technical solutions to products or methods, typically representing significant innovations in a particular technical field. Utility Model Patent: These patents usually pertain to new and practical technical solutions related to the shape or structure of a product. Design Patent: This category focuses on the novelty and uniqueness of the appearance of a product, covering aspects such as shape and pattern.
3 The regression results using utility model and design patents are consistent with the benchmark results. Due to space limitations, the results are not displayed in .
4 The mean values of the proportion of R&D to total assets (R&D1) and sales revenue (R&D2) are 1.03% and 1.84%, respectively.
5 We estimate TFP using the Levinsohn and Petrin (Citation2003) model, which incorporates labour, fixed assets and materials as inputs in production function, estimated annually for each industry.