Does efficient corruption pay?

Buying and selling a product or service involves a number of costs, including time spent searching for the best prices, negotiating for good discounts, researching product quality and writing contracts where applicable. Broadly, these are called the transaction costs of economic exchange, and part of the reason firms exist is to keep transaction costs at a minimum.

Recently, I came across an interesting paper by Fisman and Gatti which suggests that bribery also involves a transaction costs—it takes time to negotiate a bribe rate and the terms of the favor to be done, e.g. the number of regulations to be avoided. The interesting point here is that if a transaction cost is indeed present, then the greater the number of regulations a firm would like to avoid, the more time negotiations require. Further, we can expect the bribe amount to increase too with the number of regulations the firm wants to avoid. Now putting these results together gives us a couple of testable hypotheses:

1. If transaction costs are present in bribery, then the bribe paid and the time that firms spend dealing with government officials should be positively correlated; and,
2. The bigger the transaction cost, the stronger the positive relationship between the time spent by firms dealing with government officials and the bribe amount.

Using data from Enterprise Surveys, the Fisman and Gatti study does find a positive relationship between the bribe amount and the time spent by firms dealing with government officials. More interestingly, it finds that this positive relationship is weaker in countries where agents are better aware of a going bribe rate (a more efficient bribe market) and in countries with a more elaborate structure of rules or greater degree of legal formalism. Both these factors leave less room for negotiations, lowering the transaction cost. The immediate effect of this is more “efficient corruption” and therefore increased welfare.

The idea of “efficient corruption” and the sorts of institutions that will lead to more efficient corruption is fascinating from a research point of view. Nevertheless, my first thought on this is the immiserizing growth effect—a lower transaction cost of bribery can increase the incidence of bribery, reducing welfare in the long run. If this effect is strong enough, it could more than outweigh the positive effects of more “efficient corruption” discussed above. Is this immiserizing growth empirically valid? And if it is, at what level of legal formalism does the immiserizing growth effect kick in? These are important questions for future research.