The Latest Research Findings on Corruption

Let me put on a hat as a development economist today.

The November issue of the Quarterly Journal of Economics (or QJE for short), one of the top 5 academic journals in economics, is released today. It features two papers in the field of development economics.

One paper, written by Marianne Bertrand, Simeon Djankov, Rema Hanna, and Sendhil Mullainathan, analyzes corruption involved in the acquisition of driver's licenses in India. They find the following facts in a convincing way (there are other findings in the paper, but these are less convincing):

1. If a man in Delhi is more willing to obtain a license quickly, he will be more likely to obtain one without taking the licensing exam.

2. Such a man pays about 50 percent more fees over and above the official fee for licensing.

3. But he is no more likely to bribe bureaucrats at the transport office.

4. Instead he is more likely to hire an "agent", a professional who helps individuals obtain a license in return for a fee. Presumably, these "agents" pay bribes to bureaucrats on behalf of license applicants.

5. These agents can procure licenses irrespective of the applicant's driving ability. They don't even charge a higher fee for applicants who can't drive.

6. These agents are, however, less likely to procure licenses for applicants who don't have documents required for application (proof of age and area of residence).

Maybe these findings are obvious to Indians or people from developing countries with rampant corruption. For development economists, however, this is a "finding", especially the presence of "agents". We always believed that when corruption occurs, citizens directly pay bribes to bureaucrats. That's not the case, at least for driver's license applications in India.

Another lesson from this paper is that corruption is socially harmful. Some economists in the past argued that corruption is good because it speeds up the lengthy bureaucratic process. But the study shows that corruption increases the number of people with a driver's license who actually cannot drive.

Finally, finding 6 suggests that corruption to bend official rules is more difficult for rules that are easy for higher-up government officials to check.

For the other development economics paper published in the QJE this month, please wait.