Does Economic Growth Reduce Corruption? Theory and Evidence from Vietnam∗

Government corruption is more prevalent in poor countries than in rich countries. This paper uses cross-industry heterogeneity in growth rates within Vietnam to test empirically whether growth leads to lower corruption. We find that it does. We begin by developing a model of government officials’ choice of how much bribe money to extract from firms that is based on the notion of inter-regional tax competition, and consider how officials’ choices change as the economy grows. We show that economic growth is predicted to decrease the rate of bribe extraction under plausible assumptions, with the benefit to officials of demanding a given share of revenue as bribes outweighed by the increased risk that firms will move elsewhere. This effect is dampened if firms are less mobile. Our empirical analysis uses survey data collected from over 13,000 Vietnamese firms between 2006 and 2010 and an instrumental variables strategy based on industry growth in other provinces. We find, first, that firm growth indeed causes a decrease in bribe extraction. Second, this pattern is particularly true for firms with strong land rights and those with operations in multiple provinces, consistent with these firms being more mobile. Our results suggest that as poor countries grow, corruption could subside “on its own,” and they demonstrate one type of positive feedback between economic growth and good institutions. ∗Contact information: Bai: [email protected]; Jayachandran: [email protected]; Malesky: [email protected]; Olken: [email protected]. We thank Lori Beaman, Raymond Fisman, Chang-Tai Hsieh, Supreet Kaur, Neil McCulloch, Andrei Shleifer, Matthew Stephenson, Eric Verhoogen, and Ekaterina Zhuravskaya for helpful comments.