T C A Srinivasa-Raghavan: Why public goods provision is so poor

OKONOMOS

You can take a donkey to the water but can you make it drink? Politicians and their parties can be punished by the voters if they think that public goods are not being supplied in sufficient quantities. But does that improve their supply? Not always.

Abhijit Banerjee, Lakshmi Iyer, and Rohini Somanathan (who now teaches at the Delhi School of Economics) have tried to find out what really causes the supply of public goods to increase. Sadly, in the end, they don't quite succeed.

They do, however, manage to debunk the notion that bottom-up collective action is the key to improving supply. "A large part of the variation in access to public goods seems to have nothing to do with the 'bottom-up' forces highlighted in these (collective action) models and instead reflect more 'top-down' interventions."

For example, ask the authors, some of the princely states in India, even though they had no reason to improve the health and educational attainments of their services, did precisely that. Even now, we find that hospitals and schools work well in the most unexpected of places. Why?

This is not to say that these collective action groups don't matter. But "empirical research on group characteristics and public good provision finds that while these characteristics do matter, the social composition of communities is able to explain only a fraction of the total variation in provision." And so on.

So what does matter? Personally, I would put it down to chance (perhaps something even as insignificant as a good district magistrate who kicks butt effectively). But economists like to look for "better" (generalisable), multivariate explanations.

There is also another problem they don't like to confront: change. Places which have good public goods provision when the survey is done could reach the bottom of the heap in a few years' time. Bihar is an example of this, as are UP and MP. And the reverse also happens, for instance, Tamil Nadu since the mid-1980s and earlier, Haryana in the 1970s under Bansi Lal. This also needs explanation.

The authors say "research on group characteristics and collective action suggests that there are many missing pieces to the public goods puzzle." The problem, however, may not be that these are hard to find. It is that these pieces may not exist at all, except as highly transient phenomena.

There is nice discussion of incentives for bureaucrats who, in the end, are the main instruments of delivery. If their incentives are not good enough or perverse, poor outcomes are virtually guaranteed.

The authors think that with the right incentives the bureaucrats may deliver. The problem with this approach is that it is, in a sense, tautological. With food in its mouth, even my dog stops begging.

I would have put it differently. The solution lies not in devising incentives for action but very strong disincentives for non-action, such as dismissal. Amend Article 311 of the Constitution and see how the bureaucracy gets galvanised. There is nothing like the fear of losing your job.

They also ask whether decentralisation to the local level would help. But they raise more questions than they answer, which suggests that there may not be a 'one-size-fits-all' solution to the problem of public goods supply.

There is also the issue of demand. My own view, sneered at by a very acerbic economist who has poured scorn on thousands of others, is encapsulated in a paper I wrote for the Rajiv Gandhi Foundation.

This is that the demand side also needs to be examined because Say's Law doesn't always work for education. And this is because of the opportunity costs implicit in attending school. Banerjee, Iyer and Somanthan also say something similar.

The authors are absolutely right when they say that as growth picks up and spreads, "we will need to shift our attention to quality differences, where both the theoretical and empirical literature is in its infancy. There are several challenges to delivering quality rather than access."

Thus, as they point out, you can see a school building. You may even see a teacher or two and students. But "judging whether the children are learning as much they should can be quite a challenge."

Having attended the Delhi School of Economics for my MA, I fully endorse that.

Public Action for Public Goods, Working Paper 12911, February 2007, http://www.nber.org/papers/w12911