PDS leakages: the plot thickens

It is well understood that a substantial proportion of the grain, mainly wheat and rice, that is meant to be distributed to eligible families under the Public Distribution System (PDS) ends up being sold in the open market by corrupt intermediaries, including some dealers who manage PDS outlets. The extent of this “diversion” of PDS grain has been a matter of speculation for some time. Two recent surveys shed further light on the matter.

The diversion ratio (proportion of PDS grain “diverted” to the open market) has been estimated by several researchers in the past by matching National Sample Survey (NSS) data on household purchases with Food Corporation of India (FCI) data on “offtake.” The former tell us how much grain people are buying from the PDS. The latter tell us how much grain has been lifted by State governments from FCI godowns under the PDS quota. The difference is a rough estimate of the extent of diversion.

Based on this method, the estimated diversion ratio was around 54 per cent in 2004-05, the last year for which detailed data are available from a “thick round” of the NSS. Needless to say, this is an alarming figure. Tamil Nadu had the lowest diversion rate (around 7 per cent); the rate was well below the national average in the other southern States also (around 25 per cent in each case). By contrast, the estimated diversion rates ranged between 85 and 95 per cent in Bihar, Jharkhand, Assam, and Rajasthan. These estimates, if proved correct, suggest a comprehensive breakdown of the PDS in these States at that time.

Having said this, the reliability of NSS figures with respect to PDS purchases is not clear. There are two reasons to assume that they are not wildly off the mark. First, the State-wise averages for 2004-05 are broadly consistent with corresponding figures from the India Human Development Survey (IHDS) for the same year. Second, the inter-State patterns are more or less as one would expect, with, for instance, very little diversion in Tamil Nadu and a huge amount of it in Bihar. Nevertheless, this approach requires independent corroboration, not just because of the uncertain accuracy of NSS data, but also because of other difficulties in this method. Incidentally, among these difficulties is the utter lack of transparency in data on “offtake”: both the FCI and the Food Ministry seem to be doing their best to divulge as little as possible of it — they would do well to read Section 4 of the Right to Information Act.

Further evidence on these matters is available from a recent survey, conducted in June 2011 by student volunteers under our guidance (hereafter “PDS Survey”). The survey covered about 1,200 randomly-selected BPL households in nine sample States (Andhra Pradesh, Bihar, Chhattisgarh, Himachal Pradesh, Jharkhand, Orissa, Rajasthan, Tamil Nadu and Uttar Pradesh). The investigators were carefully trained to record the respondents' PDS purchases, in three different ways. The purchases were then compared with “entitlements” — what BPL households are supposed to get from the PDS in different States. For instance, BPL households are entitled to 25 kg of grain a month in Orissa and Rajasthan, and 35 kg a month in Chhattisgarh and Jharkhand. It turned out that in most States (with the notable exception of Bihar), BPL households were getting the bulk of their entitlements. The ratio of purchases to entitlements was 84 per cent in the sample as a whole. Here again, there were significant inter-State variations: this ratio was above 90 per cent in Andhra Pradesh, Chhattisgarh, Himachal Pradesh, Orissa and Tamil Nadu, but as low as 45 per cent in Bihar. The sample average of 84 per cent, however, suggests much lower rates of diversion (even in Bihar) than emerged from the earlier method — at least under the BPL quota.

The findings of this survey confirm other recent evidence of substantial improvements in the PDS around the country. In most of the sample States, there have been major initiatives in the recent past to improve the PDS, and it seems these efforts are showing results.

Also of interest are provisional figures on PDS purchases for 2009-10 (the latest “thick round” of the NSS) computed by the National Sample Survey Organisation. Starting with the good news, these figures suggest that on average PDS purchases of wheat and rice have more or less doubled between 2004-05 and 2009-10. This, again, is consistent with independent evidence of a revival of the PDS in recent years.

NSS-based estimates of diversion rates, however, remain high. Applying the method described earlier to these provisional figures, the diversion rate for 2009-10 seems to be around 41 per cent. This is 13 percentage points lower than in 2004-05, but still very high. The diversion rates improved (that is, declined) in almost every State, with big improvements in some States: down from 23 per cent to 8 per cent in Andhra Pradesh, from 85 to 47 per cent in Jharkhand, from 76 to 30 per cent in Orissa, and from 52 to 11 per cent in Chhattisgarh. Interestingly, these are four States where the PDS Survey also found evidence of major improvements. In 2009-10, none of India's major States had an estimated diversion rate higher than 75 per cent (the top rate, found in Bihar), in contrast with 2004-05 when as many as eight major States had that distinction.

This broad-based improvement is good news, but needless to say diversion rates remain unacceptably high. The question remains how these high diversion rates (41 per cent at the national level) square with the fact that BPL households in the PDS Survey were able to secure 84 per cent of their PDS entitlements. Even if the comparison is restricted to the nine sample States, a similar contrast applies.

There are at least two possible explanations. First, the PDS Survey is more recent: it took place two years after the NSS survey. And as mentioned earlier, there is consistent evidence of steady improvement in the PDS in recent years in many States. However, it is difficult to believe that progress has been so rapid as to explain, on its own, the full contrast between the two surveys. Second, the PDS Survey is restricted to BPL households in rural areas.

Diversion rates may be higher (possibly much higher) under the APL quota, and perhaps also in urban areas. Indeed, the APL component of the PDS, which has expanded steadily since 2004-05 (with a big upward jump in 2009-10), is devoid of any transparency. There are no specific entitlements for APL households, and no clear allocation norms. This segment of the PDS remains highly vulnerable to corruption, as it is possible for large quantities of grain to disappear without anyone feeling the pinch.

If this tentative line of explanation is correct, two conclusions can be drawn. First, both surveys (the PDS Survey, and the 66th Round of the NSS) add to growing evidence of steady improvements in the PDS in recent years. There is still a long way to go in achieving anything like acceptable levels of functionality, especially under the APL quota, but recent progress shows that the PDS is not a “lost cause” — far from it. Second, one thing that really helps to prevent corruption is to give people a strong stake in the system (large quantities, low prices), and make sure that they are clear about their entitlements. That has already happened, to a large extent, with the BPL quota: it has become much harder to cheat the recipients, because they know their due and clamour for it if need be. As Bhukhan Singh, a resident of Kope gram panchayat in Jharkhand, put it, when the price of PDS rice for BPL households in Jharkhand was slashed to Re. 1 a kg, awareness of the new entitlements spread quickly and people made up their mind that they “would not let this go.”

The recent turnaround of the PDS in Chhattisgarh (or, for that matter, Orissa) also built largely on this simple insight, as well as on the related fact that broad coverage strengthens public pressure for a functional PDS. There is an important lesson here for the proposed National Food Security Act.