Police reform — what might work

Posted in development, economics, governance, institutions by jrahman on May 27, 2012

I last wrote about police reforms over two years ago.  In the past two years, we have had a number of fairly high profile crimes that remain unresolved.  Meanwhile, RAB continues to be a serial human rights abuser, with abduction replacing ‘crossfires’ as its preferred method of ‘law enforecement’.  And amidst all the noise about the collapsing law and order, I haven’t heard anyone seriously discuss police reforms. 

Which is a shame, because a recent research paper by MIT’s Abhijit Banerjee and colleagues show that there are some reasonably low hanging fruits available to any government serious about reforming the police.

Abhijit Banerjee and his partner Esther Dufflo are best known for leading randmised experiments in development economics.  Their book, Poor Economics, is a breakthrough study in the field (here is the website) — maybe I’ll write about it some other time.

In 2007, they did a random experiment with the Rajasthan police.  They chose 162 police stations — about fifth of the state’s thanas and policemen — and tried a set of interventions.  Three interventions — decoy visits, a freeze on transfers, and in-service training — improved police effectiveness.  In contrast, two other interventions — placing communityobservers in the police stations and weekly duty rotations with a guaranteed day off — turned out to be ineffective. 

This work has pretty clear relevance for Bangladesh.  Rajasthan is not exactly a high performing, shining, rising part of India.  Bangladeshis police is governed by the same institutional origins and dysfunctions as that of Rajasthan.  So, reforms that can work in Rajasthan should also work in Bangladesh.

Decoy visit is a straightforward reform initiative.  Interesting to note that the two other successful interventions in Rajasthan tackled frequent transfers and the lack of training, two factors identified in the ICG paper that I blogged about in April 2010.  Transfer freeze is relatively straightforward to implement — presumably all that is needed is the political will in Dhaka.  Training might be costly, but is it more costly than maintaining the RAB? 

What about the policies that didn’t work?

Banerjee et al argue that the successful interventions required decisive action at the top and the local officials just fell in line, while the unsuccessful interventions required concerted actions at the local level.  Why were the local officers not co-operative?  Simple, they had no ‘skin in the game’.  Why would the local inspector care for the constable’s day off? 

I am sure this won’t be the last paper on the subject.  Sadly, I am still waiting for anything coming out of Bangladesh on this very important matter.


One Response

Subscribe to comments with RSS.

  1. সাতকাহন « Mukti said, on August 30, 2012 at 12:22 pm

    […] – Finally, they’re talking about police reform.  (See this and this). […]

Comments are closed.

%d bloggers like this: