Against conventional wisdom

JOHN HARRISS

Weakening Welfare: The Public Distribution of Food in India by Madhura Swaminathan; LeftWord Books, New Delhi; pages 140, Rs.275 (HC), Rs.95 (PB).

WRITING in The Hindu (February 23, 2000) the Harvard economists Nirupam Bajpai and Jeffrey Sachs argue that "rather than providing across-the-board subsidies on food which tend to get dissipated in corruption, administrative costs and lower prices for the wealthy (the Government of India should) target the aid to needy households". This is just one of many statements of a conventional wisdom which Madhura Swaminathan challenges, lucidly and with passion, in a book which is modest in length but lo ng in argument. The common prescription which derives from neo-liberal economics is of course that subsidies are mostly inimical to proper economic order, and should be reduced, if not eliminated. In practice, in the structural adjustment programmes whic h have been implemented in so many countries, following this orthodoxy, the reduction or elimination of food price subsidies has been a favourite target. As Swaminathan points out, referring to evidence from Mexico, Sri Lanka, Zambia, Jamaica and Tunisia , reductions in these subsidies may have helped to bring down public expenditure - but at very great human cost, as food consumption among the poorest people has been reduced and undernutrition has increased. In the 1990s, India has gone quite some way d own this road. Swaminathan challenges not only this policy, but also the view which is more widely held, by economists who are otherwise critical of India's structural adjustment programme - as well as by the likes of Bajpai and Sachs, who are enthusiast ic about it - that food subsidies should be 'properly targeted' to the most vulnerable households. While recognising the faults and the limitations of India's existing Public Distribution System (PDS), Madhura Swaminathan's case is that the system needs to be extended as part of a process of reform, not reduced, and that it makes far more sense in Indian conditions to base it on the principle of universality rather than on a targeted approach.

The starting point of the argument is with the marshalling of evidence which shows that, massive though the numbers are of those in India who are 'poor' according to conventional consumption criteria (and it helps to be reminded that those who are poor i n this sense in India outnumber the total populations of the United States and Canada by a good margin), there are nonetheless very many more who, by widely accepted standards, are vulnerable to food deprivation - probably as much as 70 per cent of the p opulation as a whole. In these circumstances, can it possibly make sense to attempt to target food subsidies to the 'most poor', especially when it is so difficult, practically, to identify these households with any confidence? Useful though the economis ts' conception of poverty may be analytically, it does not convert at all easily into practical means for identifying at-risk households. Are not errors of exclusion from the system of public provision of some food security much more serious, in these ci rcumstances, than (likely) errors of inclusion which occur when those who are not vulnerable are still included within the system of provision? Is it not more efficient administratively to claw back from the rich through additional taxation - anathema th ough this common-sensical argument may be to neo-liberals - rather than attempting to exclude richer people from the system?

In fact, the Government of India has started in the 1990s to target food subsidies, even though the level of these subsidies as a share of gross domestic product (GDP) has neither changed very much over time nor has it been high (at an average of 0.31 pe r cent of GDP over the period 1966-97) by comparison with other countries (in Sri Lanka, for example, equivalent to 1.3 per cent of GDP even after the introduction of targeting). Yet the central issue prices in the PDS have been sharply increased (to com pensate for increased procurement prices), and the differential between open market and PDS prices very considerably reduced. There has been a fall, in consequence, in the per capita offtake of foodgrain under the PDS in a large number of States, in circ umstances in which there has also been a marked increase in the real price of food. All this has happened alongside a steep rise - given that the level of procurement of foodgrain for the public system has not fallen - in the stocks which the Food Corpor ation of India (FCI) is maintaining. Stocks now greatly exceed the level that is believed necessary for providing for food security; and maintaining these stocks has added significantly to the costs of the FCI - though the operational efficiency of the Corporation (as given by the ratio of its operating costs to procurement prices) has, if anything, improved.

Perhaps the reduced offtake through the PDS should not be a matter for concern - if the foodgrain issued through the system is now being 'properly targeted'? There is not a great deal of evidence on the impact of the measures which have been introduced, but what there is confirms the general argument that efforts at targeting lead to serious errors of exclusion. Swaminathan's own study of the 'Revamped PDS' in Maharashtra - which was supposed to increase the coverage of the population in specified geog raphical target areas, and to increase the access of the poor within them - shows that the results have been the reverse of those intended. The contribution of the PDS to cereal consumption among poor people in the target areas was found to be very low. The later 'Targeted PDS', introduced in 1997, has set low prices for the 'Below Poverty Line' (BPL) population, to be sure, but it has also, in effect, reduced the entitlements of these families to such subsidised food, to 10 kg per month - which makes f or a very meagre contribution to minimum food requirements. This, together with the sharp increase in the issue prices of cereals for those defined as 'Above Poverty Line', will have the effect of increasing necessary expenditure on basic foodgrain "even for families correctly identified as poor". The consequences - in terms of increased deprivation - are plain.

The Union Budget for 2000-2001 has of course increased the allocation for those categorised as 'BPL' from 10 kg to 20 kg a month, but it has also increased prices. Substituting the revised allocation and the new price into Swaminathan's model example (pa ge 98) of a poor tribal family in Maharashtra, it seems that the total cost of the ration for this family will increase by about 25 per cent.

Given the levels of food insecurity which prevail in India, and the scale of hunger and malnutrition, a system of public provision is essential. The likes of Bajpai and Sachs do not disagree. The question at issue is how best to provide for basic food se curity. There is no doubt that the PDS, as it has been run hitherto, has failed to serve the vast majority of India's poor people; and there is a good deal of case material which provides evidence of corruption, misuse and mismanagement. But Swaminathan argues that such malpractice is not the principal reason for the failures of the PDS. These have to do in the first place with the very wide disparities among States in the scale and operation of the system. In 1995, for instance, Andhra Pradesh, Tamil N adu, Kerala and Karnataka accounted for almost half of the total offtake of PDS foodgrain in the country, and the BIMARU (Bihar, Madhya Pradesh, Rajasthan and Uttar Pradesh) States for only 10 per cent. Evidence given elsewhere (for example, in D.S. Tyag i's book Managing India's Food Economy; Sage; 1990) shows that this is a long-standing pattern. At the same time the indications from village surveys (such as those conducted recently by the United Nations Development Programme) are that as the co verage of the PDS becomes more extensive the poor buy progressively higher amounts of grain from the system. The PDS is more likely to reach the poor where the system operates on a significant scale. The divergences between States which are so crucial he re - and which are such that, for example, in 1986-87, whereas 87 per cent of rural households in Kerala purchased foodgrain from the PDS, the same was true for only 2 per cent of households in rural Bihar and Uttar Pradesh - cannot be explained by diffe rences in State income or in the incidence of poverty. They can be explained only in terms of differing political commitment to food security (on the conditions for which see my own article in Economic and Political Weekly, XXXIV, 48, 1999). But e ven among those States in which the PDS operates extensively there are significant variations in efficiency. On the whole it appears that the system operates most effectively in Kerala - the only State to have near-universal coverage - where there is a h igher level of public action than elsewhere and which ensures accountability.

These arguments supply the starting point for Swaminathan's statement of 'What Needs To Be Done'. Indeed this manifesto is very largely an adumbration of what has gone before, and it makes a strong case in terms of logic and of morality for a universal r ather than a targeted system, for one which supplies a greater share of basic requirements of food energy (and through rations rather than through food stamps), and for using the existing food stocks to increase food security and stabilise food prices (r ather than running up the costs of the FCI to no purpose). Swaminathan says much less, however, than might have been expected, or than the book seemed to promise, about the mechanics of operating the food security system. She refers briefly to the recomm endations of the Bureau of Industrial Costs and Prices, made in 1991, for improving the performance of the FCI, and flatly rejects the arguments of the World Bank (and of others. See, for example, R. Chelliah and R. Sudarshan, eds, Income Poverty and Beyond: Human Development in India, Social Science Press; 1999) in favour of its privatisation. While she is surely right that these arguments are ideologically driven, they call for a more substantial rebuttal than she offers. She gives a number of specific proposals, on the other hand, concerning the micro-management of the delivery system, emphasising the significance of information for consumers. She also makes the important point that while decentralisation can serve the administrative efficien cy of the PDS, it is no panacea. Decentralisation only serves the needs and interests of poor people where local power is contained, as it has been to a significant extent in Kerala and West Bengal, as a result of the more serious efforts in these States than elsewhere to implement agrarian reform. In the last part of her manifesto Swaminathan simply states, against the recommendations of Kirit Parikh, whom she cites, and of others (like Bajpai and Sachs, in fact, when they suggest targeting food subsid ies by linking them to school meals programmes), "that employment and other social security programmes should be complementary to PDS and not seen as a way of restricting food distribution only to households participating in employment or other welfare p rogrammes". There is no discussion at all of the implications for State-level finances of the proposals for extending the PDS. Yet these, given both the very considerable historical variations between States which are attested in the book, and the fiscal crunch which State governments are now facing, demand attention.

At the last, therefore, Madhura Swaminathan delivers only partially on the promise of her book. But the limitations of her manifesto do not detract from the fact that the book's criticisms of the oh-so-reasonable conventional wisdom that food security th rough subsidised public distribution must be 'properly targeted', are both penetrating, and trenchantly and effectively expressed. May they prompt not just more public debate than there has been on this matter, which is so crucial for the livelihoods and well-being of so many in the country, but also political action.

John Harriss is Reader in Development Studies, London School of Economics.