Does corruption affect firm growth? New evidence from Mozambique

Abstract

This paper estimates the effect of corruption on firm growth; specifically, in the context of Mozambique. Using instrumental variable estimation to address potential problems of endogeneity and measurement error, the paper finds that corruption has a robust negative effect on sales and productivity growth. The effect is not constant but declines as firm performance drops and corruption increases. This means corruption is most damaging to high-performance firms, meaning firms with higher sales and productivity growth rates, and to firms with lower bribe rates. For low-performance firms and firms with higher bribe rates, the effect gradually approaches zero. The paper contributes to the literatures in two ways. First, it offers a more nuanced understanding of the relationship between corruption and firm growth by showing that the effect of corruption is conditional on firm-specific factors. Second, the paper extends the empirical research on corruption and firm growth to Southern Africa, which has previously been excluded from the literature.

Keywords:

• Mozambique
• corruption
• private sector
• firms
• growth

Acknowledgements

This research was carried out with the support of the World Bank. The author would like to thank Shireen Mahdi, Fiseha Haile, Hans Lynge Ottosen, John Andersen, and two anonymous reviewers for useful comments.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1 The most famous case is what has become known as the “secret debt” scandal (Hanlon, 2017). Between 2011 and 2013, then-President Armando Guebuza was presented with a proposal to create a tuna fishing fleet, a maritime security company, and a ship repair and maintenance company. Three companies were created and USD 2.2 billion was raised in finance by issuing government-backed bonds. Illegal fishing is a problem, and a national tuna company could have been profitable. Meanwhile, Mozambique could have established sovereignty over its territory and provided security to the emerging offshore gas industry. However, most of the money that was raised disappeared into the accounts of the domestic elite and of the international financial institutions who organised the loans.

2 Transparency International broadens this definition to include “abuse of entrusted power for private gain” (Transparency International, 2020).

3 Other early proponents of the argument include Leff (1964) and Leys (1965).

4 This section draws on Fisman and Svensson (2007). The appendix contains a formalised version of the argument.

5 Wald’s (1940) method was developed specifically to address errors-in-variables problems. Durbin (1954) showed that Wald’s method is a special case of IV estimation.

6 Note that Maputo City and Maputo Province were merged into Greater Maputo. Gaza, Inhambane, and Niassa were not included, which means inferences can be made only to the remaining seven provinces plus Maputo City.

7 The population of informal firms is unknown.

8 Breusch–Pagan tests suggest the possible use of (heteroskedasticity-consistent) robust standard errors