Be that as it may, now the government has begun to vacate private good production and supply with a vengeance. Even where education and health are concerned, the government provides only about a quarter of the total supply.

The rationale, as always these days, is the systemic efficiency that the private  sector - read  profit  motive - induces. Basically, goes the argument,  the  private  sector  can  produce  pretty  much everything more efficiently  because,  well,  it  has  to. So if efficiency is an important requirement, which  it is from the investors' point of view, why not leave it to the private sector?

But now comes a paper* by Hanming Fang and Peter Norman which argues exactly the opposite. It shows, with the use of some high-level mathematics, that the "public provision of private goods may be justified on pure efficiency grounds in an environment where individuals consume both public and  private goods." Perhaps the 11th Plan can be informed by the argument contained in this paper.

And through what route do the authors reach this heretical conclusion? "The government's involvement in the provision of private goods provides it with information about individuals' private good purchases that facilitates more efficient revenue extraction for the provision of public goods."

Of course, some prior conditions have to be satisfied. The most important of  these  is  that  there are no leakages in the system, of whatever kind. Thus, "the no-arbitrage  restriction we require for the joint provision mechanism to  improve upon the benchmark outcome in our model also accords with  many publicly-provided private  goods we see in reality because in-state college tuition, public health insurance, and public schools are all commodities that are difficult to resell."

The authors also say that their  model provides a basis for optimal taxation. It does so by recognising that "there is efficiency loss from separating revenue and expenditure problems in public finance�" This claim is perhaps open to serious challenge.

The authors make it clear that they are not talking about redistribution. For  example, they do not have highly restrictive benchmarks for outcomes. This, they say, enables them to explain "why some, but not all, private goods are publicly provided."

In other words, if you don't insist on a very degree of stringency, it will probably not matter who produces what because the public sector can do it as efficiently as the private sector. This may not be intuitively obvious but I suppose the Delhi Metro is a good example as indeed is the All India Institute of Medical Sciences, not to mention the IITs.

Perhaps  the most important finding, if one may use such a term for what is a  purely theoretical paper, is that it makes no sense to separate spending decisions from revenue-raising decisions because such  a  separation "generates efficiency losses."

Most people  will disagree with the conclusions that the authors have used because  visual evidence  and  experience  both point  in  the  opposite direction. But the immediate importance of this paper may not lie in its practicability but in the way it could force us to rethink what has become, for all  practical purposes a dogma, namely, that where the private sector is  concerned, we can effortlessly assume the existence of systemic efficiency.

*Toward an Efficiency Rationale for the Public Provision of Private Goods, NBER Working Paper No. 13827, February 2008

Budget 2008: Complete coverage