Some reflections on the National Food Security Act

The Food Security Bill became an Act with little parliamentary opposition. Yet the public debate has lingered. Would subsidised food grains reduce malnutrition? Won’t it be better to invest in health and education instead? Can we afford the cost of subsidising food for such a large chunk of the population? Should we continue to waste money on the flawed PDS system? How will the grain markets be affected? This column offers a perspective on these important questions.

Impact on nutrition and health

India has the dubious distinction of being at the top of the chart for malnourished children and at the bottom of the chart for health indicators. In the minds of many, this is a disgrace, and the reason why India needs a scheme to provide food security. What impact would the National Food Security Act (NFSA) have on malnourishment and health?

Some components of the NFSA such as mid-day meals and maternity benefits (nutritional supplements and cash) are directly targeted at nutrition. What about the supply of cheap rice and wheat?

Some people hope that subsidised food grains will reduce malnourishment by inducing beneficiaries to consume more food grains and thereby more calories. However, malnutrition stems from many things including micronutrient deficiency (Malhotra 2012), being ill informed about nutrition (Malhotra 2013), and poor sanitation (Hammer and Spears 2013). The fact that other things matter has led critics to dismiss the nutritional impacts of food subsidies. This is incorrect too.

Poor households typically consume about 10 kg of grain while the subsidy will be offered on a fraction of this amount (5 to 7 kgs) (Kaul 2013). Thus, the subsidy releases resources that could be spent on the purchase of more food grains or other foods or even non-foods. That is why the scheme to subsidise food grains is essentially an income transfer programme denominated in terms of the price of food grains. Kaul (2013) gives estimates of the elasticities1 of per capita calorie intake from different food groups with respect to rice subsidy per capita: 0.123 for cereals, 0.151 for lentils, 0.237 for fruits and vegetables and 0.169 for meat. Therefore, the income effect through the subsidy will allow beneficiaries to buy more food grains and also more nutritious food. Since the additional income can be spent on non-foods as well, the impacts will not just be restricted to nutrition.

Rationale for income transfers

People could use the extra income for medical or educational expenses; farmers could use it to supplement their expenses for farm inputs. Two third of Indian households are rather poor and living on the brink, and an income transfer of even Rs. 3000 ($48 approx.) a year can be handy in avoiding it2. It can allow them a chance to live their life with dignity.

In addition to the welfarist argument above, there are some instrumental arguments as well. First, the Indian economy has grown relatively fast over the last few decades but the gains have gone disproportionately to the top layer. The envy and resentment caused by such a pattern of growth can be politically unhealthy. It tears the social fabric leading to deterioration in institutional performance and the functioning of society.

Second, it is indisputable that the direct nutritional interventions such as meals for lactating and pregnant mothers as well as mid-day meals for school children would have a positive impact on the human capital of the next generation and should therefore be considered an investment. Third, an income transfer programme of this sort can also lead to greater output through its impact on human capital and on the risk-taking ability of the poor. Consumption can thus translate into investment.

This is not just wishful thinking. The successes of conditional transfers such as Bolsa Famila in Brazil and Progressa in Mexico are well known. Recent research from Africa have also shown positive impacts of unconditional transfers in terms of reducing hunger, increasing livestock holdings and investment in new trades (Blattman, Fiala and Martinez 2013, Haushofer and Shapiro 2013).

Of course, these schemes are different in scale and intent from the income transfer programme under the NFSA in India. But the relevant point is that a well-implemented income transfer programme can enable the poor to overcome the constraints of poverty and become more productive. The cliché that the poor will waste the money on bad habits is not borne out by the evidence.

But can we afford such an expensive programme?

The Act has expanded the coverage from 44% of the population under the Targeted Public Distribution System (TPDS) to 67% while reducing the subsidised amount from 7 kgs to 5 kgs per person for all except the very poor (beneficiaries of Antyodaya3). It is easy to see at a rough level that the amount of grain commitment by the centre cannot have changed much as 7 x .44 = 3.08 and 5 x .67 = 3.33. The additional commitments are principally because of other welfare schemes that include maternity benefits and school feeding programmes but the magnitudes are not large. For this reason, Mishra (2013) calculates the incremental cost over the existing TPDS to be 0.2% of the Gross Domestic Product (GDP).

Some experts have speculated that the incremental costs are underestimated. For example, Gulati et al. (2012) believe that the Act will necessitate large investments in food production. We don’t see why such a small increase in PDS supply will have a big impact on required food production. Of course, irrespective of NFSA, agricultural investments are necessary. But those costs cannot be attributed to the NFSA.

Mishra (2013) also mentions that there could be further additional costs due to setting up of food commissions and grandfathering of existing beneficiaries. The magnitude of these costs cannot be firmly established without assumptions of how the Act will be implemented (Sinha 2013). Since these costs are difficult to gauge in advance, let us take the figure mentioned by Dr Raghuram Rajan as the incremental cost due to the Act - 0.5% of GDP. Is this affordable?

The answer depends on how you evaluate the benefits and opportunity costs of the extra resources used. One difficulty is that the benefits of the subsidy that actually reaches the targeted beneficiaries are intangible and what part of the allotments actually reaches them is dependent on the extent of corruption and leakage. What is pertinent is the opportunity cost of these resources. For example, what if this 0.5 % of GDP were used as investment in education or healthcare? Would that not do a lot more good than wasting resources on PDS that is known to be a leaky bucket? Indeed, this seems like the right question to ask but implicitly the leaky bucket is being compared to well-functioning systems of delivering education and healthcare. The trouble is that the public education and healthcare in India are both poorly delivered (ASER 2012, Muralidharan et al. 2013). Many parents, even poor parents, are putting their faith in private schools, and public healthcare facilities are poor. A wasteful and inefficient system is being compared with another equally flawed system. In short, it is hard to evaluate benefits in either case and give a clear answer in terms of the opportunity costs of the food subsidy. What is crucial, therefore, is to assess opportunities to improve the working of the system.

TPDS – a flawed system

We believe that affordability is not the real problem and has served to distract our attention from the main problem – the seriously flawed system of delivering the food subsidy, namely TPDS.

Some of the perennial problems plaguing the TPDS are:

It is very difficult to identify the poor. Exclusion errors are huge. According to Jha and Ramaswami (2012), only about 30% of the poor derive some benefit from the PDS.
Fair price shop owners have a built-in incentive to divert grain to the open market. In 20042005, about 54% of subsidised grain was estimated to be lost in this manner (Jha and Ramaswami 2012)⁴. In 2009-2010, the estimate is 40% (Ramaswami and Murugkar 2012).
Farmers from many states where coarse grains are important lose by having subsidised grains dumped in their area. In several states such as Karnataka, Maharashtra and Rajasthan, one or more coarse grains are important staples. These coarse grains are typically cultivated in arid areas where the soil is of a low quality and the farmers are poor.

Note that the targeting problem can only be solved by making the coverage almost universal. But substituting the in-kind transfer system with a well-designed cash transfer system could solve the other two problems.

The influence of PDS on market prices of food grains⁵

What happens to the market price of grain under cash and in-kind transfers, respectively? The question is important because, in practice, it is difficult to devise a perfect safety net. Some of the poor could be left out even if the coverage was meant to be universal. Moreover, if a policy intervention causes a rise in the market price of grain, the non-poor who are not entitled to a food subsidy would be adversely affected, and this would make the scheme politically difficult to implement. Finally, even the poor beneficiaries would source part of their food grains from the market.

Grain markets have a well-defined seasonal pattern. Price levels are at their lowest at harvest time and then rise through the year to cover the costs of carrying stocks. Grain prices can be higher either because of a higher harvest price or because of greater margins of storage and distribution.

When governments procure, the initial harvest price is determined not by the forces of supply and demand but by the support price set by the government. For politicians, the demand for a higher support price affords an opportunity to mobilise a constituency. In India, most of the grain (especially wheat) is procured from a few states. As a result, the support price, and hence the harvest price, is typically determined through bilateral bargaining between the central government and the state governments. Given the nature of parliamentary democracy in India, the ruling party cannot ignore the votes in these surplus states, and consequently the outcome of the bargaining game is a price that is higher than it would otherwise be (cash transfers). The power of the farm lobby to dictate prices does vary with circumstances. Shortage in the world market reduces the threat of imports and increases their power, but their power diminishes if government stocks are far in excess of need.

An additional complication is that the Central government is extremely sensitive to the possibility of under-supply to the PDS. Politicians and bureaucrats perceive the costs of insufficient supplies but nobody is held accountable for excessive stocks and high prices. Predictably, the errors are in one direction. Since the early 1990s, procurement has consistently exceeded PDS sales. This is why there have been recurrent crises of excess stocks and consequently, of storage capacity. High procurement prices and large government stocks displace private trade and therefore, bumper procurement and stocks continue until the momentum is broken by an exceptional event such as a drought or by ad-hoc dumping of grain in the domestic (open market sales) or international market (exports).

The implication is that in our structure of procurement, there is always a tendency to accumulate excess stocks which in turn high prices. If the expanded obligations under the Food Security Act are met by a mix of transfers in-kind and in cash, it will restrain the pressures on procurement and public stocks. Of course, the problem vanishes in a world where cash transfers completely replace the transfers in-kind.

A case for cash transfers⁶

We have already mentioned how the substitution of the in-kind transfers by a cash transfer scheme would solve the most common problems that plague TPDS.

An immediate objection to cash transfers is infeasibility. How can cash be transferred? Does a poor country have the systems to implement it? A cash transfer system is constructed on two pillars: a payments system to distribute the cash; and an authentication system to verify that the transaction is with the intended beneficiary. Conventional payment systems are brick-and-mortar banks and post offices. By definition, such infrastructure is not well developed in the poor remote areas of low-income countries. This has been a barrier to the use of cash transfers.

Computerisation of financial systems and the use of internet and mobile devices have broken through this impasse. Africa leads the world in the use of mobile phones to transfer cash (Mas and Rotman 2008, Mas and Kumar 2008). It has allowed urban migrants to remit money to their families still living in urban areas. Effectively, any retailer is potentially a point for banking transactions. In India, post offices have typically delivered cash payments in welfare programmes (such as those arising from pension and public works), but this process is vulnerable to capture by the intermediaries, which results in both delay and loss. Policy now emphasises the direct transfer to savings accounts of beneficiaries in banks and post offices. This is possible only because of computerisation of financial systems. This still does not address the issue of “last-mile” connectivity. An emerging model here is the use of intermediaries between the banks (situated in towns and larger habitations) and the beneficiaries (resident in villages). These intermediaries, called banking correspondents, provide services of withdrawal and deposit with the help of internet-enabled portable devices that record these transactions in real time. Internet connectivity is provided through the usual mobile phone networks.

Authentication systems require verification of the identity of the beneficiary. In a digital system, this can be done through a user-supplied numeric code or password. More secure systems rely on biometric identification such as Aadhaar⁷.

Another justification of in-kind transfers is that it leads to self-selection of only the truly needy. The effectiveness of self-selection unfortunately depends on the relative inconvenience of buying in a ration shop or even having a lower quality of food available in ration shops. The inconvenience of standing in a queue for buying something from a ration shop could be perhaps enough to deter the rich from taking advantage of it except for the fact that they can send their domestic help for such chores. A cash transfer with biometric identification would make the self-selection work more effectively, thus making even universal coverage affordable.

An infusion of cash in a local area could give rise to a sudden increase in prices. In an environment where markets are not well developed, the rise in prices may not trigger imports from other areas to bring down the prices in a short time. In-kind transfers of food may induce an increase in demand for non-food items but will not cause food price inflation. Clearly, this is a real concern about cash transfers, and it suggests that cash transfers are more appropriate for the areas where the markets are well developed.

The most serious objection to any sort of cash transfer is that food prices fluctuate and that a commitment to the poor in terms of a certain quantity of food per person cannot be maintained very easily. Consider the logistics of the problem. Suppose it is decided to give each household 25 kg of grain each month at a subsidised price and the subsidy amount required for a recipient to purchase that much grain is deposited into her account at the beginning of the month. If the market price has risen by 10% by the time the recipient goes to buy the grain, the subsidy amount would fall short of what is required. The subsidy amount should therefore be adjusted as the market price changes. It is, of course, expensive to adjust the subsidy amount too frequently, and the cost of not adjusting it frequently enough will be borne by the poor8. This can be an objection against any cash transfer scheme.

Concluding comments

The suggestions for reforms in the existing TPDS are the main benefits of the NFSA. The fact that the implementation of the Act would still require targeting the bottom 67% is still a big problem. It would have been a lot better if the coverage had been near universal requiring the exclusion of only the top income groups (for example, income tax payers) who could be easily identified. It is possible that the task of organising logistics of procurement and distribution of grain for such a large population is what kept the government from not entertaining that option. Cash transfers would obviate those concerns while reducing the waste and leakage at the same time. In a diverse country like India, however, it is always a mistake to impose solutions from the top. Different states are at different stages of development and the best way is to let them choose the means of delivering their income transfers.

A version of this column has appeared on Yojana, a Development Monthly of the Planning Commission of India. The authors would like to thank Arka Roy Choudhuri for research assistance.

Notes:

The elasticity of per capita calorie intake with respect to per capita rice subsidy is a measure of the percentage change in per capita calorie intake when there is a one percentage change in the per capita rice subsidy.
This is based on a household of five people, each getting subsidided grain of 5 kgs per month at a price that is Rs. 10 per kg less than the market price yielding Rs. 250 x 12 = Rs. 3000.
Antyodaya Anna Yojana (AAY), a central government scheme, was launched in December 2000 for one crore ‘poorest of the poor’ families. Initially, it was planned to provide 25 kg food grain per family per month, at Rs. 2/kg for wheat and Rs. 3/kg for rice.
Jha and Ramaswami (2012) examine the working of the PDS in India based on data in the Consumption Expenditure schedule of the 61st round (2004-2005) of the National Sample Survey (NSS).
This section is drawn from Kotwal, Murugkar and Ramaswami (2012).
For a more elaborate case for cash transfers please refer to Kotwal and Ramaswami (2013), the paper from which we have reproduced much of the material in this section.
Aadhaar is a 12-digit individual identification number issued by the Unique Identification Authority of India (UIDAI) on behalf of the Government of India, which serves as a proof of identity and residence through India.
Of course, the shortfall this month can be added to the amount sent to the consumer next month. In addition, the market prices can go down as frequently as they go up, so over a long time it can be a wash.