China is the largest emerging market in the world. It has experienced spectacular growth over the last three decades, and is currently in the upper end of middle-income countries. Entrepreneurial activity and the entry of new firms are crucial for spurring growth at this stage of development and for avoiding middle-income traps (Zilibotti 2017). Corruption – being associated with special privileges for connected companies and preventing a level playing field for new entrants – can be an important impediment to growth for economies at this stage of development.
For this reason, widespread corruption in China has been considered a serious threat to sustained economic growth. In this context, President Xi Jinping’s administration, which viewed corruption as a threat to the Communist Party’s survival, initiated an anti-corruption campaign on 8 November 2012 – only 19 days after taking power. Xi’s anti-corruption drive has been considered more far-reaching and long-lasting than any previous attempts.
The effects of corruption on entrepreneurial activity
In a recent paper co-authored with Guanmin Liao (Central University of Finance and Economics) and Jiaxing You (Xiamen University), we exploit the anti-corruption campaign as a negative shock to the effectiveness of corruption, which should have benefitted small entrepreneurial firms in more corrupt industries to a larger extent. We show that extensive corruption in China may indeed have hampered the process of development and that the anti-corruption campaign has been an effective step forward in favouring an efficient allocation of resources and entrepreneurial entry and ultimately sustained growth.
We explore the effects of competing with highly corrupted incumbents on small entrepreneurial firms’ performance and entry patterns using a large-scale proprietary dataset. Our data provide comprehensive information on a sample of public and private firms, which is largely representative of the distribution of firms in the Chinese economy across geographic regions, industries, and size classes. This allows our analysis to have direct implications for the effects of corruption on the economy’s allocational efficiency.
In addition, we observe firms’ efforts to obtain political favours. An item on all Chinese firms’ profit and loss accounts – entertainment expenses – is highly correlated with the grease money firms spend to get better government services and the protection-money firms spend to lower tax payments (Cai et al. 2009). Entertainment expenses are also often discussed by news media as associated with corruption, and have been widely used in existing literature to measure potential corruption (e.g. Griffin et al. 2016, Lin et al. 2016). We can thus use the average entertainment expenses of the largest companies in an industry to measure the extent of corruption that entrepreneurial firms are likely to face in the industry in which they operate. We also explore the robustness of our results to the use of other proxies for political connections.
We find that corruption has significant negative effects on the performance of entrepreneurial firms. Small firms are less profitable and have lower total factor productivity when they compete with large industry peers spending more on entertainment expenses. This appears to be the case because, under these circumstances, small firms invest less, have lower growth of sales, and face higher financing costs than the median firm in the industry. Importantly, in all our tests, the negative spillover effect of large firms’ entertainment expenses is muted after the start of the anti-corruption campaign. This indicates that a negative shock to the effectiveness of corruption benefits entrepreneurial activity.
All results are obtained by controlling for firm-level entertainment expenses and including interactions of province and time effects. Thus, our findings cannot be interpreted as being driven by provincial shocks. Furthermore, we show that large firms’ entertainment expenses are unlikely to capture other large firms’ characteristics, such as size and leverage, which may in turn be correlated with small firms’ performance.
We also show that entrepreneurial firms in provinces in which large firms have high entertainment expenses have weaker performance. In these tests, we are able to absorb industry level shocks by including interactions of industry- and time-fixed effects, demonstrating that industry shocks do not drive our findings.
Finally, we provide evidence that large firms’ corruption efforts hamper an efficient allocation of resources. Estimating a model based on Bai et al. (2016), we find that labour (capital) is less likely to be allocated to firms with high marginal productivity of labour (capital) if these firms operate in high entertainment expenses industries. Also in this case, the allocation of capital and labour improves after the start of the anti-corruption campaign.
Corruption also appears to have an effect on the geographical distribution of entrepreneurial activity. Not surprisingly, since corruption lowers their profits, the proportion of young firms is lower if large firms in the same province and industry have high average entertainment expenses during the previous year. This is the case not only for young firms in general, but also for young firms with high productivity.
Overall, our findings highlight the negative spillover effects of corruption, and have broader implications for corporate governance literature, which has been dissecting the effects of political connections.
As shown by a large literature in corporate finance, firms in both developed and emerging economies attempt to obtain political favours, such as lenient taxation, relaxed regulatory oversight, and generous financing, by hiring politicians to their boards and other posts, through more or less legal lobbying practices, providing financial support to alternative political factions, or paying bribes. A number of papers have shown that these behaviours benefit firm shareholders in a variety of countries (e.g. Faccio 2006, Amore and Bennedsen 2013, Borisov et al. 2016, Zeume 2017). However, evidence on the extent to which these behaviours cause costs for society is scarce.
We document a mechanism through which corruption may stifle economic growth, i.e. by increasing the rents of a few incumbent firms, corruption may stifle entrepreneurial activity and decrease the ability of small entrepreneurial firms to grow and compete with the incumbents.
–Mariassunta Giannetti, Xiaoyun Yu