Mending the Food Security Act
Jean Drèze IN THE HINDU
|The National Advisory Council has proposed a framework for the National Food Security Act. But its potential could be wasted by a flawed approach to the PDS.|
Two years have passed since the Central government announced that a draft National Food Security Act (NFSA) would be posted on the Food Ministry’s website “very soon.” After prolonged deliberations, a detailed framework for this Act has recently been proposed by the National Advisory Council (NAC), and a draft is on the anvil. This is a “compromise draft” of sorts, heavily influenced by the government’s own concerns and priorities.
The NAC framework includes important provisions relating, for instance, to child nutrition, reform of the public distribution system (PDS), and redress of grievances. It has the potential to put all food-related schemes on a new footing, in a rights framework. However, this potential is in danger of being wasted by a flawed approach to the PDS.
In this approach, the PDS rests on a three-way division of the population, among “priority,” “general” and “excluded” households. (This article focusses on rural areas.) Priority households, covering at least 46 per cent of the rural population at the all-India level, are to get 35 kg of grain a month at “Antyodaya prices” (Rs. 3 a kg for rice, Rs. 2 for wheat and Re. 1 for millets). General households will get 20 kg at no more than half of the Minimum Support Price. And excluded households, which account for 10 per cent of the rural population, will get nothing.
This framework is problematic. First, it hinges on a lasting division of the population into three groups, without any clarity as to how the groups are to be identified. In the absence of any obvious alternative, the NAC is effectively falling back on the Below Poverty Line census to identify priority groups. This is a major setback — the NAC’s entire work began with a virtually unanimous rejection of BPL-based targeting for the PDS. Exclusion errors in earlier BPL censuses were very large, and the next BPL census is unlikely to fare much better, judging from the pilot survey.
Second, since identification criteria are left to the Central government, with some discretion for State governments, nobody has guaranteed PDS entitlements under the Act, except for a few ultra-marginalised groups (such as the so-called Primitive Tribal Groups) which have a right of “automatic inclusion” in the priority list. Other households have no legal entitlement to be included in the priority list or, for that matter, in the general list. Therefore, they have no guaranteed PDS entitlements at all. This undermines the basic purpose of the Act.
Third, the transition from the current Above Poverty Line-Below Poverty Line framework to the NAC framework is likely to be disruptive. There are at least three major sources of disruption: the creation of an “excluded” category; the transition to a new BPL list; and the switch from household to per capita entitlements. Each of these changes entails a loss of entitlements for significant numbers of households. Meanwhile, the entitlements of other households will be enhanced. Can we expect this transition to happen without major tensions, or even to be completed at all?
Fourth, the NAC framework fails to “de-link” PDS entitlements from official poverty estimates, and to prevent a rapid shrinkage of PDS coverage over time. It is well understood by now that official poverty lines in India are abysmally low, and that undernutrition is not confined to households below the “poverty line.” In the NAC framework, 46 per cent coverage of priority groups in rural areas corresponds to the proportion of the population below the “Tendulkar poverty line,” plus a margin of 10 per cent for targeting errors. This is significantly higher than the current BPL coverage of about 33 per cent. But except for ruling out any reduction of PDS entitlements before the end of the 12th Five Year Plan (which is only a few years from now), nothing in the draft NFSA prevents the government from reducing PDS coverage in tandem with official poverty estimates over the years.
Fifth, the idea of a universal PDS in the poorest 200 districts was dropped from the NAC framework (after being agreed and placed on record). This was an important idea, because any targeting process here is likely to lead to massive delays, fraud, and exclusion errors. In many of these districts, the local administration has little credibility. Large numbers of poor households are outside the BPL list, and are likely to remain excluded from the proposed “priority” list. Further, targeting is pointless in areas where an overwhelming majority of the population is vulnerable to food insecurity. Launching a universal PDS in these districts would have addressed a large part of the food insecurity problem in rural India in one go, at a small extra cost.
Sixth, the NAC abandoned another important idea as it went along: the automatic inclusion of all Scheduled Caste and Scheduled Tribe (SC/ST) households in the priority list — unless they come within the standard exclusion criteria. This will be a major protection against exclusion errors, and a well-justified form of positive discrimination in favour of SC/ST families. But the idea was dropped, on the grounds that it is difficult to reconcile with pre-specified “caps” on the coverage of priority groups at the State level based on poverty estimates. Punjab, for instance, has a low poverty ratio but a high proportion of SC/STs in the population — there is no obvious way to handle this.
In short, the NAC framework not only perpetuates the flaws of BPL targeting but also institutionalises artificial social divisions under the law. It is not difficult to imagine the Act being used as a foothold to extend these divisions to other domains.
The obvious alternative, a universal PDS, is a ‘no-no’ for the Central government. Is there another way to repair, or at least contain, the damage? I believe there is. Before coming to that, let me mention an interesting finding of recent BPL identification studies (by Reetika Khera, Sabina Alkire, and Himanshu, and others). These analyses, mainly based on the 2004-05 data from the National Sample Survey or the 2005-06 data from the National Family Health Survey, suggest that about 25 to 30 per cent of households in rural India meet simple, transparent and verifiable “exclusion criteria,” such as having a government job, owning a motorised vehicle, or living in a multi-storied pucca house.
This suggests a simple but far-reaching modification of the NAC framework: expand the excluded category, but extend “priority” entitlements (35 kg of grain at Antyodaya prices) to all other households. With an exclusion ratio of, say, 30 per cent, the foodgrain requirements will be the same as in the current NAC framework. The financial cost will be a little higher (because all entitled households will pay Antyodaya prices), but the extra cost will be a small fraction of the total food subsidy.
In this “quasi-universal” framework, every rural household will be entitled, by law, to 35 kg of grain a month at Antyodaya prices, unless it comes within the well-defined “exclusion criteria.” Everyone will be clear about their legal entitlements. The burden of proof, so to speak, will fall on the government to exclude a household, and poor households will be well protected from exclusion errors. State governments will be free to move even closer to universalisation, if they wish, by waiving some exclusion criteria and contributing additional resources to the PDS (as many States are already doing). Automatic inclusion of SC/STs (unless they come within the exclusion criteria) will be built in. PDS entitlements will be de-linked from the APL-BPL rigmarole, and from poverty estimates. And while some social division will remain, it will be “at the top,” without undermining solidarity among disadvantaged groups.
Two further modifications of the NAC framework will round up this proposal quite nicely. First, the idea of a universal PDS in the poorest 200 districts could easily be reinstated, by waiving exclusion criteria in these districts for an initial period of, say, 20 years. Second, the Act could be gradually extended to the whole country, over a period of, say, three years, starting with the poorest 200 districts. This will make it easier to meet the additional foodgrain requirements in a phased manner.
This approach is not perfect, but it seems much preferable to the confused, impractical and divisive framework that has emerged from the NAC (or rather, from protracted discussions between the NAC and the government). It will be easy to adapt the current NFSA draft to this approach, while retaining the valuable work that has been done by the NAC on other aspects of the draft. This small modification could make a big difference.
(The author is a Visiting Professor at the University of Allahabad. The views expressed here are his own.)
- Urban BPL Survey is Expected to Enable the Effective Design and Delivery of Inclusive Programmes for the Urban Poor (equalityindia.wordpress.com)
- Poverty, caste and religion to be simultaneously mapped for census (hindu.com)
- India ‘redefines’ poverty for new survey (urbanhealthupdates.wordpress.com)