Government discretion: how high a barrier to entry for Chinese private enterprises?

Abstract

A robust literature has demonstrated that de jure regulatory barriers to entry deter firm formation. This paper identifies the discretion available to government officials in tollbooth positions to establish barriers to entry as a de facto policy barrier which is an important source of uncertainty for Chinese entrepreneurs. Defining discretion as the share of provincial government revenue deriving from discretionary charges levied by government departments, I estimate the impact of discretion on the formation of Chinese domestic private enterprises. I find that a 1% decrease in discretion is associated with an additional private enterprise for every 10,000 provincial residents.

Keywords:

• entrepreneurship
• regulation
• uncertainty
• barriers to entry
• China

Acknowledgements

Previous versions of this article benefitted from the valuable insights of Gordon Betcherman, Jose Galdo, Lisa Mills, Jeremy Paltiel, Yanling Wang, and an anonymous reviewer.

Notes

1. Two World Bank studies are notable exceptions. These are Doing Business in China: 2008, which assesses de jure barriers across provincial capitals, and Enterprise Surveys undertaken in 2005 and 2012, which survey firms on both de jure and de facto barriers (see http://enterprisesurveys.org). The academic literature investigating these data has focused on the behaviour of existing firms, rather than firm formation. See, for instance, Hallward‐Driemeier et al. (2006), Lin et al. (2010) and Lu et al. (2013).

2. With the important distinction that Holmes estimates the effects of sharp policy discontinuities at the borders between US states. Though discretion as constructed here is a property of Chinese provinces, it need not be the case that there is an abrupt discontinuity in the discretion available to local officials as one crosses provincial borders. The Holmes model is thus employed here as illustrative of firms’ decision-making process and cannot be used for estimation in a regression discontinuity design.

3. In English translation, this body is sometimes referred to as the Industrial and Commercial Bureau. I use SAIC throughout the paper for convenience.

4. That is to say, beyond the optimal level required to ensure regime support, according to the dynamic described by North et al. (2009). The gap between local officials’ optimum level of discretion and that of the central government may reflect both monitoring costs and local officials’ less encompassing interest in growth in the sense defined by Olson (1993).

5. For more on these reforms, see Yang (2004), Mertha (2005), Zou (2006) and Guo (2008).

6. Though not unique to China, the phenomenon of ‘guanxi’ has received extensive attention in the Chinese case. See, for example, Xin and Pearce (1996).

7. For operational definitions and sources for all variables, see Appendix 1.

8. Foreign-invested enterprises (FIEs) are also excluded from the analysis. Huang (2008) documents the favoured status of FIEs relative to the domestic private sector. Though a full account of this phenomenon is beyond the scope of this paper, one explanation is that FIEs, being barred from strategic sectors, less often compete directly with state-owned enterprises. They may also create technology spillovers and other positive externalities valued by provincial governments.

9. For a defence of the reliability of Chinese government statistics, see Zhou (2011, p. 862).

10. For descriptive statistics for these variables and all covariates, see Appendix 2.

11. A case could be made that lagged values of regressors should be used to estimate discretion’s effect on firm formation. Analysis using lagged values has been performed, and results do not differ significantly from those reported below. As taking lagged values requires dropping observations from an already limited dataset, I employ same-year values for all variables.