Dangers of narrow targeting

Published : Oct 18, 1997 00:00 IST

The shift from a public distribution system with universal application to a targeted one can leave the poor worse off than before.

MADHURA SWAMINATHAN

FOOD subsidies have come under attack in recent years from proponents of structural adjustment; food subsidies are seen as being excessive and reducing food subsidies is viewed as a way of controlling the fiscal deficit. The problem of food security requires high priority in India and given the scale of chronic hunger and malnutrition, food subsidies in India are not very large. It should be self-evident that the food policy should identify the most effective ways of providing food to those who need it and not merely aim to cut back government expenditure.

The two key components of the system of food management are the procurement of foodgrain and its distribution to consumers through the public distribution system (PDS). The PDS has many problems and limitations - the biggest one is its inadequacy in relation to the scale of hunger and economic vulnerability in India - and requires major overhaul. Instead of reforming the system, the Central Government of the day seems bent on dismantling it. The specific solutions offered in the post-1991 period of liberalisation have been disastrous. The Revamped PDS introduced by the Congress Government in 1992 was a failure and the Targeted PDS proposed by the United Front Government in January 1997 is likely to worsen food availability among the poorest households.

FOOD security can be defined as access for all people at all times to food that is sufficient to lead an active and healthy life. Food security requires that certain minimum quantities of food be available to all persons at affordable and stable prices; the establishment of food security requires intervention in food markets. For sustained food security, other forms of economic security are also essential, that is, people must have the means to acquire the required amounts of food.

Hunger and malnutrition are prevalent on a mass scale in India, and there are still areas of the country where deaths by starvation occur every year. Average cereal consumption per capita is low and has declined since the 1950s, from 17 kg in 1952 to 13 kg in 1993-94 ( Table1). Per capita cereal availability has fallen in the years following the "new economic policies": it declined from 510 grams per capita per day in 1991 to 468 grams in 1992, and the provisional figure for 1996 is 498 grams.

The National Family Health Survey 1992-93 collected data on the extent of malnutrition in 24 States and found that 53 per cent of children in the age group 0-4 were undernourished in terms of a weight-for-age criterion. Furthermore, 21 per cent of the children surveyed were severely undernourished (that is, fell below three standard deviations of the norm). There are, of course, large regional variations. The extent of severe malnutrition was, for example, 6 per cent in Kerala, 21 per cent in Maharashtra, 24 per cent in Uttar Pradesh and 31 per cent in Bihar.

CONTRARY to popular perceptions, the consumer food subsidy in India fell, in real terms, in the early 1990s and has not changed much as a share of gross domestic product (GDP) in recent years (Table 2). The Central Government's food subsidy declined between 1989-90 and 1992-93. There was a significant increase in 1993-94, but the subsidy declined once again thereafter. Food subsidies, both absolute and as a share of GDP, reached a trough in 1992-93.

The limited data available for international comparisons of food subsidies indicate that per capita food subsidy in India is lower than in many other countries (such as Pakistan, Brazil, Sri Lanka) and that the expenditure on food subsidies as a proportion of GDP is also lower than in other countries.

It is also important to note that the expansion of subsidies over time, and particularly after 1993, is not primarily because of an increase in the gap between procurement prices and issue prices. It is mainly because increases in intermediate costs and costs of procurement, storage, buffer-stock operations and transportion, and the burden of inefficiencies in the system of storage and distribution (for example, due to wastage) are being passed on to consumers through higher prices. Between 1975 and 1989, the distribution costs of the Food Corporation of India (FCI) increased by 274 per cent whereas the procurement costs increased only by 70 per cent. And in 1992-93, for example, the per quintal cost of procurement and distribution of the FCI was almost as high as the per quintal support price.

For consumers, the period of structural adjustment has been one in which there was a steep increase in the prices of foodgrains supplied through the public distribution system. At the national level, between 1991 and 1994, the Central Issue Price of the common variety of rice rose 85.8 per cent and the issue price of wheat rose 71.8 per cent. During the same period, the Index of Wholesale Prices rose 44.4 per cent. Between 1990-91 and 1994-95, the Consumer Price Index for Agricultural Labourers rose 53.1 per cent. In other words, the cumulative increase in the prices of major foodgrains sold through the PDS was higher than the corresponding increase in other general price indices. Consequently, the price differentials between the PDS and private markets narrowed or even disappeared. Price data from a market in Delhi, for example, showed that the difference between market and PDS prices for wheat fell from Rs. 1.11 a kg in January 1991 to 33 paisa a kg in February 1994.

ON January 26, 1997, the the United Front Government announced the introduction of a Targeted PDS scheme from June 1. The declared objective of targeting was, predictably, to focus the PDS on the poor. The scheme differs from the earlier system in three major respects. First, a distinction between the below-poverty-line and above-poverty- line populations is introduced. Secondly, the monthly allocation for the poor has been limited to 10 kg of grain per household, now raised to 15kg under pressure from State governments and the Left parties. Thirdly, for non-poor households, a transitory allocation is envisaged based on past levels of utilisation.

In a situation where resources are said to be fixed, targeting is often presented as a means of giving more to selected sections of the poplation, for instance, the poor. In reality, where targeting is inappropriate, the shift from a scheme with universal application to a targeted one can leave the poor in a population worse off than before.

Targeting carries many costs, and these can be high. First, administrative costs for a targeted scheme are usually high, around 2 to 5 per cent of total expenditures, and much higher than for a universal scheme; the costs depend on the type of identification required.

Secondly, a shift from a universal to a targeted scheme is associated with two types of errors (referred to as Type I and Type II errors in statistics): the error of inclusion of the non-poor and the error of exclusion of the poor. International evidence shows that a shift to targeted welfare schemes from universal schemes is accompanied by a big increase in Type II errors, that is error of excluding the poor. In Jamaica, for example, Type II mistakes increased by 50 per cent with a shift from a general food subsidy to restricted food stamps. In Sri Lanka, the per capita food subsidy declined sharply in real terms after the Government shifted from near-universal rationing to means-tested food stamps.

There are large welfare costs associated with the error of excluding those deserving from a scheme, and these are the social costs of having a population that is hungry and malnourished. The costs of wrong inclusion, however, are only fiscal costs.

Thirdly, targeting can lead to incentive distortions, whereby people alter their behaviour in response to targeting. Lastly, when resources are fixed, reducing the number of beneficiaries by means of targeting should increase the benefits available to each participant. However, it has been found in the experience of both developed and developing countries that targeted welfare schemes lack political support and, over time, the resources available for targeted programmes tend to decline.

THE population that can legitimately be excluded from a basic system of food security in India is small, and the administrative costs of excluding this small segment may outweigh the savings in expenditure. The need for a universal system in India is dictated by the scale of hunger and economic vulnerability in the country.

It is well-known that the food share or the share of household expenditure spent on food declines as income or expenditure rises. Based on this relationship, in the scholarly literature, the inverse food share (the ratio of total expenditure to food expenditure) is taken as a measure of poverty. In other words, poor households are likely to be characterised by large food shares. Food security in households that spend a very high proportion of their expenditure on food will be very sensitive to changes in food prices and incomes, and such households are vulnerable to small changes in the demands on their income.

A very striking feature of the data on consumer expenditures in India is the extent of the population that has a high food share (Table 3). In the United States, a criterion for eligibility for the food stamp programme, a scheme that gives coupons to the poor that can be exchanged for food, is that a family must spend on third or more of its total expenditure on food. By that criterion, most of India's population qualifies for state support in respect of food. The lowest or poorest two deciles of households spend around 70 per cent of their total expenditure on food. The share falls as expenditure increases but it is only the top decile in the urban areas and the top 5 percentile in the rural areas that have a food share of less than 50 per cent. By the criterion of food share, there is little doubt that the majority of the population requires a system of food security. Only the top 5 per cent of the urban population in India spends around one-third of total expenditure on food.

Given the quality and characteristics of commodities available in the PDS, the top 10 per cent or so of households are not likely to make purchases from the PDS. At the lower end of the income distribution, however, narrow targeting can worsen food security for many vulnerable households. This is both because it is difficult to identify who falls below a given poverty line and because of the possibility of excluding large numbers of persons who are in need of food security.

The official poverty line, a useful academic tool for understanding and monitoring poverty, cannot be easily used to demarcate one set of households from another on the ground. Can one say that a household with ten or fifty or one hundred rupees more than the poverty line level of income is not vulnerable as compared to another household with an income less than the poverty line? In a country where the majority of the population works in the informal sector, income is very difficult to measure. Further, most households that earn income from casual work or self-employment are likely to face fluctuations in income over time. For a household that spends 60 per cent of income on food, whether the household is defined as being "poor" or "not poor" in official statistics, even a small change in incomes or expenditure, say for urgent medical purposes, can be a major blow to a family's economic status and can, in turn, affect its access to food.

Under the proposed system, even if a family below the poverty line is able to buy its quota from a ration shop and thereby meet 25-30 per cent of its grain requirements, an increase in the open market price of 15-20 per cent can neutralise the effect of the lower price in the ration system.

Apart from the problems of accurate measurement of income, excessive reliance on a single income poverty line to identify households gives scope for discrimination, siphoning off of resources and exclusion of the neediest from the scheme.

The Targeted PDS has been introduced in a situation in which quantities distributed through the PDS have been declining (Table 4). The decline in distribution cannot be attributed to a fall in availability as the Government maintained large buffer stocks during this period. The decline in purchases is on account of the rise in issue prices, the poor delivery system, and the lack of purchasing power among consumers. The low take-up is often on account of poor infrastructure.

Many poor households are already excluded from the PDS, and the introduction of targeting can lead to the further exclusion of vulnerable households from the system. Data collected by the National Sample Survey in 1986-87 showed that around 40 per cent of rural households in the lowest four expenditure deciles had not purchased any commodity from the PDS.

THERE are large gaps in the existing system of delivery of food, and large numbers of the income-poor are excluded from the PDS. The quantities supplied are often inadequate and prices have not been low in recent years. The PDS is badly administered in many States with large-scale diversion of grain, wastage, low quality and unreliability of provisioning. The inefficiencies in the operation of the FCI have added greatly to the subsidy burden. No serious effort has been made by the Central Government and by a majority of State governments in the last few years to alter this situation by improving the delivery system and by specifically attempting to include the poor in the PDS. On the contrary, the new schemes introduced after 1991 have been designed with the aim of reducing expenditures and reducing the size of the target group for schemes for providing food to the people through public channels. They are likely to worsen food security for the poorest in society and lead to further deterioration in the system of the public delivery of food.

Madhura Swaminathan is Associate Professor, Indira Gandhi Institute of Development Research, Mumbai.

Sign in to Unlock member-only benefits!
  • Bookmark stories to read later.
  • Comment on stories to start conversations.
  • Subscribe to our newsletters.
  • Get notified about discounts and offers to our products.
Sign in

Comments

Comments have to be in English, and in full sentences. They cannot be abusive or personal. Please abide to our community guidelines for posting your comment