REETIKA KHERA
June 12, 2011
Obituaries for the PDS system are a bit premature, based as they are on outdated data and presumptions. Despite flaws like pilfering and leakages, the system shows signs of improvement in certain states.
Cash transfers (CTs) are increasingly advocated as an alternative to the Public Distribution System (PDS). The proponents of cash transfers typically begin from the assumption that the PDS is a proven and irreversible failure. This assumption, however, is based on outdated or selective estimates of PDS leakages derived from National Sample Survey (NSS) data for 2004-5 or earlier, or worse, no data at all. For instance, Pranab Bardhan declares that the PDS is “dysfunctional” on the grounds that less than a quarter of PDS grain reaches the poor, but provides no source for this “rough estimate”. Citing a paper that estimates leakages in 2001-2 for rice and wheat separately, Kaushik Basu focuses on diversion of wheat alone, conveniently ignoring rice (with a much lower diversion rate) even though rice accounts for the bulk of all PDS grain.
Recent field reports suggest that these obituaries for the PDS might be premature. Some of the signs of revival are discussed below, including declining leakages in some states. Though much of the evidence of revival pertains to the period since 2007, NSS data for 2007-8 (the latest available round) do capture some of these improvements.
First, access to the PDS has been rising steadily. “Access” depends on the proportion of households that the government is willing to subsidise (i.e., proportion of BPL, or “below poverty line”, households) and on whether the PDS delivers to BPL households. Between 2004-5 and 2007-8, the proportion of households getting any grain from the PDS has progressively risen, from about a quarter of rural households (27 per cent) to just over one-third (35 per cent). In several states, this proportion increased by more than 10 percentage points (e.g., from six per cent to 18 per cent in Uttar Pradesh (UP), from 22 per cent to 36 per cent in Orissa, from 25 per cent to 47 per cent in Chhattisgarh and from 40 per cent to 59 per cent in Kerala). In addition, in states such as Andhra Pradesh (AP), Himachal Pradesh (HP) and Tamil Nadu (TN) access was high to start with — 81, 69 and 88 per cent respectively.
INCREASING OFF-TAKE
Second, per capita purchase of grain from the PDS has also been increasing. Apart from the four southern states, where per capita grain purchase has been relatively high (between 2.2 kg per month in Karnataka and 4.98 kg per month in TN in 2007-8), there are some surprises among the Hindi-speaking states too. In Chhattisgarh, per capita purchase witnessed a five-fold increase between 2004-5 and 2007-8, from a measly 600 grams to 3.2 kg per month. Uttarakhand's performance is similar — up from 640 grams to 2.15 kg. Another fact that has gone unnoticed is the performance of HP and Jammu and Kashmir, where per capita purchases are higher than in some of the southern states. These two indicators suggest that in 2007-8, on average, more people were getting more grain per capita from the PDS than before.
Of course, not all is well with the PDS — far from it. Focussing on purchase alone presents an incomplete picture. To get the full story, one needs to look at the difference between offtake by the states and per capita purchases. Leakages, thus estimated, are the sum of transport and storage losses on the one hand, and illegal diversion on the other. At the all India level (rural and urban areas combined), in 2007-8, 44 per cent of the PDS grain did not reach entitled households. State patterns in diversion rates are plausible. Not surprisingly, the lowest diversion rate is in TN (four per cent) and the highest in Bihar (90 per cent).
There are interesting surprises in the state trends, which are consistent with field reports. Three states — Chhattisgarh, Orissa and UP — usually considered “basket cases”, are beginning to show signs of revival. In Chhattisgarh, leakages of around 50 per cent in 2004-5 were much lower by 2007-8, even virtually eliminated according to one estimate. In Orissa and UP, states that are notorious for corruption, leakages came down from a whopping 76 per cent to 50 per cent and from 58 per cent to 27 per cent respectively. While leakages from the PDS remain unacceptably high, these states show that the situation is not irreversible.
CONTINUING IMPROVEMENTS
NSS data beyond 2007-8 are not yet available. This is unfortunate, because recent field evidence suggests that these improvements have been consolidated since then. The most impressive example is Chhattisgarh: Ration shops open regularly in the first week of each month and most people get their full quota of 35kg at the right price. In Orissa, two field studies conducted in 2010 point to a drastic improvement: In both, 75 per cent of the respondents reported getting their full quota. Even in one of the most dysfunctional blocks of Jharkhand (Manika Block, Latehar District), informal investigations in eight villages earlier this year revealed that BPL cardholders get between 30-32 kg/month. Though this is less than their entitlement (35 kg/month), it is a huge improvement. Earlier, supplies were so unpredictable that people find it hard to recall how much they were getting! A similar picture emerged from enquires in Khunti District too.
Much of this revival can be attributed to a renewed political interest in the PDS which has manifested itself in an expansion of PDS coverage and a reduction in PDS prices. Many states have been emulating a policy associated with Chhattisgarh, but originally traceable to Tamil Nadu and Andhra Pradesh. This is the “populist” promise of one rupee rice for all. The popularity and electoral payoffs of PDS reforms in Chhattisgarh provided the necessary nudge in several other states, including Jharkhand, Kerala, Madhya Pradesh, Orissa and Rajasthan.
In 2007, PDS coverage in Chhattisgarh was extended to 80 per cent of rural households, well beyond the Planning Commission's stingy poverty estimates (45 per cent). The expansion makes a lot of sense, since the Central government's attempt to target the poor by identifying “below poverty line” (BPL) households failed miserably. In 2004-5, half of the poorest 20 per cent of rural households did not have a BPL card. Further, in a country where the proportion of under-nourished far exceeds the proportion of poor, linking PDS coverage to official poverty estimates does not make sense in the first place.
Further, many states have also reduced PDS prices below the centre's prices. In AP, Chhattisgarh, Orissa and Kerala, rice is given at Rs. 2/kg; in Rajasthan wheat is sold at Rs. 2/kg since May 2010; in TN and Jharkhand, rice is currently priced at Re 1/kg. Lower PDS prices combined with rising market prices have dramatically raised people's stake in the PDS.
LESSONS FOR THE FUTURE
What lessons can we draw from all this for the Food Security Act? First, PDS coverage must be delinked from poverty estimates. Linking the two leads to large exclusion errors and hardship for the rural poor. The NAC proposal, where the coverage of “priority” households is effectively pegged to poverty estimates, is a step backward — several states have already expanded coverage. This is all the more worrying as recent pilots to test the proposed methodology of the next BPL Census (based on the Saxena Committee's recommendations) confirm the danger of large exclusion errors.
Two, since 2007, states have demonstrated the political will to invest in the PDS, by putting in state resources to make it work. Measures that enhance voice (e.g., expanded coverage and reduced PDS prices) combined with other measures such as computerising records, improving grievance redressal procedures and so on, have begun to show results.
Three, whatever the possible merits of cash transfers, the system is not ready for that: Opening of bank accounts is incomplete and access to banks remains very poor for large sections of the population. The UID infrastructure, which is necessary for a cash-transfers regime, is also far from ready — in the first six months of enrolment, only three million residents had been enrolled.
The bottom line is that the PDS can deliver. In TN, where rice is sold at Re. 1/kg, some households are able to meet their basic monthly food needs (including dals and edible oil) for Rs. 100! Other states are moving in that direction: In HP and AP, rural residents can now buy subsidised grain, pulses and oil at the local PDS outlet. Importantly, the signs of revival are not restricted to the usual suspects, but include “basket cases” like Orissa and UP. The PDS is a system that people have learnt to use. Foisting a completely new regime of cash transfers on them without first putting systems in place would be very dangerous.
Cash transfers (CTs) are increasingly advocated as an alternative to the Public Distribution System (PDS). The proponents of cash transfers typically begin from the assumption that the PDS is a proven and irreversible failure. This assumption, however, is based on outdated or selective estimates of PDS leakages derived from National Sample Survey (NSS) data for 2004-5 or earlier, or worse, no data at all. For instance, Pranab Bardhan declares that the PDS is “dysfunctional” on the grounds that less than a quarter of PDS grain reaches the poor, but provides no source for this “rough estimate”. Citing a paper that estimates leakages in 2001-2 for rice and wheat separately, Kaushik Basu focuses on diversion of wheat alone, conveniently ignoring rice (with a much lower diversion rate) even though rice accounts for the bulk of all PDS grain.
Recent field reports suggest that these obituaries for the PDS might be premature. Some of the signs of revival are discussed below, including declining leakages in some states. Though much of the evidence of revival pertains to the period since 2007, NSS data for 2007-8 (the latest available round) do capture some of these improvements.
First, access to the PDS has been rising steadily. “Access” depends on the proportion of households that the government is willing to subsidise (i.e., proportion of BPL, or “below poverty line”, households) and on whether the PDS delivers to BPL households. Between 2004-5 and 2007-8, the proportion of households getting any grain from the PDS has progressively risen, from about a quarter of rural households (27 per cent) to just over one-third (35 per cent). In several states, this proportion increased by more than 10 percentage points (e.g., from six per cent to 18 per cent in Uttar Pradesh (UP), from 22 per cent to 36 per cent in Orissa, from 25 per cent to 47 per cent in Chhattisgarh and from 40 per cent to 59 per cent in Kerala). In addition, in states such as Andhra Pradesh (AP), Himachal Pradesh (HP) and Tamil Nadu (TN) access was high to start with — 81, 69 and 88 per cent respectively.
INCREASING OFF-TAKE
Second, per capita purchase of grain from the PDS has also been increasing. Apart from the four southern states, where per capita grain purchase has been relatively high (between 2.2 kg per month in Karnataka and 4.98 kg per month in TN in 2007-8), there are some surprises among the Hindi-speaking states too. In Chhattisgarh, per capita purchase witnessed a five-fold increase between 2004-5 and 2007-8, from a measly 600 grams to 3.2 kg per month. Uttarakhand's performance is similar — up from 640 grams to 2.15 kg. Another fact that has gone unnoticed is the performance of HP and Jammu and Kashmir, where per capita purchases are higher than in some of the southern states. These two indicators suggest that in 2007-8, on average, more people were getting more grain per capita from the PDS than before.
Of course, not all is well with the PDS — far from it. Focussing on purchase alone presents an incomplete picture. To get the full story, one needs to look at the difference between offtake by the states and per capita purchases. Leakages, thus estimated, are the sum of transport and storage losses on the one hand, and illegal diversion on the other. At the all India level (rural and urban areas combined), in 2007-8, 44 per cent of the PDS grain did not reach entitled households. State patterns in diversion rates are plausible. Not surprisingly, the lowest diversion rate is in TN (four per cent) and the highest in Bihar (90 per cent).
There are interesting surprises in the state trends, which are consistent with field reports. Three states — Chhattisgarh, Orissa and UP — usually considered “basket cases”, are beginning to show signs of revival. In Chhattisgarh, leakages of around 50 per cent in 2004-5 were much lower by 2007-8, even virtually eliminated according to one estimate. In Orissa and UP, states that are notorious for corruption, leakages came down from a whopping 76 per cent to 50 per cent and from 58 per cent to 27 per cent respectively. While leakages from the PDS remain unacceptably high, these states show that the situation is not irreversible.
CONTINUING IMPROVEMENTS
NSS data beyond 2007-8 are not yet available. This is unfortunate, because recent field evidence suggests that these improvements have been consolidated since then. The most impressive example is Chhattisgarh: Ration shops open regularly in the first week of each month and most people get their full quota of 35kg at the right price. In Orissa, two field studies conducted in 2010 point to a drastic improvement: In both, 75 per cent of the respondents reported getting their full quota. Even in one of the most dysfunctional blocks of Jharkhand (Manika Block, Latehar District), informal investigations in eight villages earlier this year revealed that BPL cardholders get between 30-32 kg/month. Though this is less than their entitlement (35 kg/month), it is a huge improvement. Earlier, supplies were so unpredictable that people find it hard to recall how much they were getting! A similar picture emerged from enquires in Khunti District too.
Much of this revival can be attributed to a renewed political interest in the PDS which has manifested itself in an expansion of PDS coverage and a reduction in PDS prices. Many states have been emulating a policy associated with Chhattisgarh, but originally traceable to Tamil Nadu and Andhra Pradesh. This is the “populist” promise of one rupee rice for all. The popularity and electoral payoffs of PDS reforms in Chhattisgarh provided the necessary nudge in several other states, including Jharkhand, Kerala, Madhya Pradesh, Orissa and Rajasthan.
In 2007, PDS coverage in Chhattisgarh was extended to 80 per cent of rural households, well beyond the Planning Commission's stingy poverty estimates (45 per cent). The expansion makes a lot of sense, since the Central government's attempt to target the poor by identifying “below poverty line” (BPL) households failed miserably. In 2004-5, half of the poorest 20 per cent of rural households did not have a BPL card. Further, in a country where the proportion of under-nourished far exceeds the proportion of poor, linking PDS coverage to official poverty estimates does not make sense in the first place.
Further, many states have also reduced PDS prices below the centre's prices. In AP, Chhattisgarh, Orissa and Kerala, rice is given at Rs. 2/kg; in Rajasthan wheat is sold at Rs. 2/kg since May 2010; in TN and Jharkhand, rice is currently priced at Re 1/kg. Lower PDS prices combined with rising market prices have dramatically raised people's stake in the PDS.
LESSONS FOR THE FUTURE
What lessons can we draw from all this for the Food Security Act? First, PDS coverage must be delinked from poverty estimates. Linking the two leads to large exclusion errors and hardship for the rural poor. The NAC proposal, where the coverage of “priority” households is effectively pegged to poverty estimates, is a step backward — several states have already expanded coverage. This is all the more worrying as recent pilots to test the proposed methodology of the next BPL Census (based on the Saxena Committee's recommendations) confirm the danger of large exclusion errors.
Two, since 2007, states have demonstrated the political will to invest in the PDS, by putting in state resources to make it work. Measures that enhance voice (e.g., expanded coverage and reduced PDS prices) combined with other measures such as computerising records, improving grievance redressal procedures and so on, have begun to show results.
Three, whatever the possible merits of cash transfers, the system is not ready for that: Opening of bank accounts is incomplete and access to banks remains very poor for large sections of the population. The UID infrastructure, which is necessary for a cash-transfers regime, is also far from ready — in the first six months of enrolment, only three million residents had been enrolled.
The bottom line is that the PDS can deliver. In TN, where rice is sold at Re. 1/kg, some households are able to meet their basic monthly food needs (including dals and edible oil) for Rs. 100! Other states are moving in that direction: In HP and AP, rural residents can now buy subsidised grain, pulses and oil at the local PDS outlet. Importantly, the signs of revival are not restricted to the usual suspects, but include “basket cases” like Orissa and UP. The PDS is a system that people have learnt to use. Foisting a completely new regime of cash transfers on them without first putting systems in place would be very dangerous.
