The actionability test of development research

I have blogged on multiple occasions illustrating how what appear to be logically perfect solutions turn out far less effective in reality. This highlights an important lesson in development policy making - there is  most often a world of difference between the assumed outcomes of a technically ideal policy and the realised outcomes of its real-world implementation.  

This is the implementation deficit. The difference is a combination of state capacity weakness, behavioural change problems, political economy difficulties, and a variety of entrenched systemic constraints as well as failings of the agents implementing them. 

The history of development policy making is littered with implementation deficits. Good and technically robust ideas implemented with good intentions but failing to yield the desired objectives. 

The problem is even more pervasive in development economics research. In fact, it's rare to find a paper that points to a significantly efficacious idea or insight which is also implementable (or actionable) in the median development setting. 

In this context, I am drawn to this description of an example of cash transfer programs and food rations by Yogendra Yadav from an interview of Jean Dreze by Ashok Kotwal, 

In response to Kotwal’s advocacy of cash transfer over subsidised food to the poor, Dreze drew a distinction between an economist who advises the government and an economist who advises the poor. The policy advisor must think of the possible benefits while assuming that his or her suggestions will be implemented fully and in the right spirit. The advisor to the poor must focus on the likely consequences of a policy, how would a policy be implemented, on the ground. Dreze said that on paper, direct cash transfer can be the most economical and efficient way to help the poor, but food-grain delivery through ration shops is their best real-life option.

This is what Dreze says in that interview,

On the question of the PDS vs. cash transfers, I think there are at least two possible sources of differences in our perspective. One is that I have been very influenced by numerous conversations with poor people in states such as Jharkhand, Bihar, and Chhattisgarh. I have been influenced by their fear of kind being replaced with cash and I think they have some very valid arguments. For example, they don’t trust the government to index the cash transfers – not just before the elections but also afterwards. They are worried about what will happen to local food prices if the system of procurement and distribution is dismantled. They are worried that cash can be more easily misused than food, because food can only be consumed in small quantities over time while cash is easily spent in one go.

Another possible difference relates to where we place ourselves and who we are advising. Economists, to the extent that they get involved in policy debates, think of themselves largely as government advisors. So, for example, this point about inflation and whether the government is going to index the transfers, if you are positioning yourself as a government advisor it is not much of an issue because indexation can be part of your advice to start using cash transfers. But if you are advising poor people rather than the government, things might look quite different. If poor people ask me, after hearing that the government may introduce cash transfers, whether they can trust the government to index the transfers, I would have to say no, there is no guarantee of it at all.

For me the bottom line is that the poor are getting something very important from the PDS. It has improved a great deal in recent years and it is now a very substantial form of support for people who are on the margin of subsistence. They have no reason to go along with something else unless there is a real guarantee that it is something demonstrably better, and as of now I do not see that... So I would not say that I oppose cash transfers, but I do resist them in the context of the PDS, at least for now... It is easy to do a blackboard calculation showing that the government could save thousands of crores by replacing kind with cash. But there is a key difference in practice – the PDS is in place and the cash payment system is not. When you consider giving up something that is in place and means a lot to poor people, you really have to think hard about the practicalities of the transition.

In other words, policy economists should see themselves as advisors to people instead of governments. Their objective should be the analysis of the world as it is, instead of an imaginary idealised world, with prescriptions relevant for the former.