When I gave my driver’s test in US, it certainly was a different experience. Though, I got my licence in India by taking a driver’s test and without paying an agent, the test which I took in US was definitely more rigourous than the Indian one.
So when I saw this article in the QJE, I was very excited to see what the authors had to say about the process of obtaining driver’s license in India. In particular I was interested in what these guys had to say about corruption and its efficiency.
The paper takes an experimental approach to studying corruption. It is based on the project financed by IFC, New Delhi which involved a group of actors recruited to obtain a driver’s license. The experiment had built in incentive scheme to see how many of the participants obtain a driver’s license with the help of an agent and how many of the total particiapants obtained a license without giving a driver’s test. And as expected the numbers are not surprising but definitely embarrasing. They find that close to 71% of the license getters did not take the driver’s test.
Now is this evidence of corruption also an evidence for distorted incentives which result from understaffing or from a system outnumbered by test takers? Though, one is tempted to answer this question positively, it cannot be overlooked that corruption in such a case should have been only a time saving device. As th eauthors put it, if the agents are simply offering time saving devices, why does it also mean that in most cases they can easily bypass the RTO test?
Thus, payment to the agents for getting a license goes beyond greasing the wheels and entails an important social cost: more bad drivers and hence more accidents!


Bertrand M, Djankov S, Hanna R, & Mullainathan S, (2007), Obtaining Driver’s License in India: An Experimental Approach to Studying Corruption, The Quarterly Journal of Economics, November.