Recently I invited Danila Serra from SMU to give a talk on her research on corruption up here in Normatopia. She gave a great talk and there was a huge turnout.
She reported on Lab experiments designed to see if websites like “I Paid A Bribe” can deter what Danila calls extortionary corruption, and also if there were feasible refinements to the site that might perform better.
The first reports experiments showing that even if you let bribe demanders post false information about bribe activity, a reporting platform significantly reduces bribes. Among other things, the second paper reports experiments showing that while allowing bribe demanders to collude raises the size of bribes, the collusion breaks down fairly quickly over time.
It seems like reporting platforms really can reduce bribes.
At the end of the talk, I asked Danila if she favored eliminating completely this form of corruption. I got the feeling that she did.
But I think if bribes for services were eliminated in developing countries with no other reforms also being introduced, service delivery would actually become worse rather than better for citizens.
Civil servants might quit showing up to work. Processing times could skyrocket.
My reasoning is that many low level civil servants are not paid well, if at all. In fact, some civil service jobs are allocated by the job seekers buying the position! Collecting bribes is a functionally important incentive for showing up and doing any work.
So I think eliminating corruption, ceteris paribus, could well make things worse. And this is of course, just an example of the brutal theory of the second best which implies that in a world of multiple distortions, eliminating a single distortion can make things worse rather than better.
To me, absent major civil service reform, Danila’s extortionary corruption is often a necessary evil.