Click here for search results

Corruption and Trade Tariffs, or a Case for Uniform Tariffs (WPS 2216)

HomeDataPapersProgramsDiagnosticsIn the News

by R. Gatti (1999)

A highly diversified trade tariff menu may fuel bribe-taking behavior. Setting trade tariff rates at a uniform level limits public officials' ability to extract bribes from importers.

By explicitly accounting for the interaction between importers and corrupt customs officials, Gatti argues that setting trade tariff rates at a uniform level limits public officials' ability to extract bribes from importers.

If the government's main objective is to raise revenues at the minimum cost to welfare, optimally-set tariff rates will be inversely proportional to the elasticity of demand for imports. So they will generally differ across goods.

Such a menu of tariff rates endows customs officials with the opportunity to extract rent from importers. If officials have enough discretionary power, they might threaten to misclassify goods into more heavily taxed categories unless importers pay them a bribe. Because of the bribe, the effective tariff rate for the importing firm increases, so demand for the good decreases.

The resulting drop in import demand implies an efficiency loss as well as lower government revenues, compared with the optimal taxation benchmark without corruption.

A similar argument applies when customs officials offer to classify goods into low-tariff categories in exchange for a bribe.

Setting trade tariffs at a uniform level eliminates officials' opportunities to extract rents. Thus, when corruption is pervasive, a uniform tariff can deliver more government revenues and welfare than the optimally set (Ramsey) tariff benchmark.

The empirical evidence confirms that these considerations are relevant to policymaking, since a robust association between the standard deviation of trade tariffs—a measure of the diversification of tariff menus—and corruption emerges across countries.


Comment on this paper (view posting guidelines)

Download article
(72 kb)


Back to top 520

Permanent URL for this page: