Corruption in Procurement and Shadow Campaign Financing: Evidence from Russia

Abstract

Using a measure of tunneling out of Russian firms, constructed from banking transactions data as transfers of cash out of legitimate to fly-by-night firms, we show that firms that get public procurement contracts exhibit abnormally high tunneling rates within four weeks of regional elections. In contrast, firms with no public procurement revenue exhibit no tunneling around election cycles. Since politicians on the campaign trail need cash the most around election time, they are the most likely recipients of tunneled cash close to elections. Using variation in the quality of tax inspectors as a source of exogenous variation in tunneling, we document a causal relationship from tunneling around elections to obtaining public procurement contracts. Two placebo experiments-using placebo election dates and legitimate rather than shadow transfers-confirm the validity of our empirical strategy. Our estimates yield a locality-level measure of corruption in public procurement. Using this measure, we reject the efficient greasing" hypothesis by showing that, in more corrupt localities, public procurement contracts are allocated to less efficient firms, and therefore, corruption has negative welfare implications.