‘Leakages’ from the Public Distribution System – that is, grain released by the Food Corporation of India that fails to reach consumers – stood at around 42% in 2011-12 at an all-India level. Passed in 2013, the National Food Security Act guaranteed food security and mandated PDS reforms. As anticipated, based on data from the Household Consumption Expenditure Survey, Reetika Khera shows that leakages were down to 22% by 2022-23.
During the discussion on the National Food Security Act (NFSA), 2013, there was nervousness about guaranteeing food security through the Public Distribution System (PDS). The nervousness stemmed from the poor record of the PDS – according to the National Sample Survey (NSS) data, in 2011-12, leakages were at 41.7% at the all-India level.
The main argument in favour of continuing with the PDS was that states that had undertaken PDS reforms had witnessed major improvements. Between 2004-05 and 2011-12, several early reforming states saw a dramatic reduction in leakages, such as Bihar (from 91% to 24%), Chhattisgarh (from 52% to 9%) and Odisha (from 76% to 25%). With the same package of PDS reforms being mandated by the NFSA, 2013, there was hope that more states could improve.
The Household Consumption Expenditure Survey (HCES) data of the NSS for 2022-23 bears out this hypothesis. The HCES is the first large-scale nationally representative dataset after the implementation of the NFSA, 2013. It suggests that PDS leakages were down to 22% in 2022-23.
Understanding the data and methodology
PDS ‘leakages’ refer to the proportion of PDS rice and wheat released by the Food Corporation of India (FCI) that fails to reach consumers. Leakages are estimated by matching NSS data on household PDS purchases with ‘offtake’ data reported in the Monthly Food Grain Bulletin of the Food Ministry.
During the reference period (August 2022 to July 2023), PDS ration card holders were getting NFSA grain (5 kilogram (kg) per capita per month for “Priority” households and 35 kg per month for “Antyodaya” households1. In addition, they got Pradhan Mantri Garib Kalyan Anna Yojana (PMGKAY)2 grain till December 2022, when PMGKAY Covid-19 relief was discontinued.
To arrive at the estimates presented here, I match offtake under NFSA including tide-over rations, non-NFSA and PMGKAY on the one hand, with HCES 2022-23 PDS purchase of (paid for and free) wheat and rice by households on the other.
It is worth bearing in mind that these are estimates, the best possible with available data. Any mismatch between offtake and purchase is attributed to leakage, though there could be other reasons (transport losses, lags in supply, and so on). For instance, Table 1 below shows how lagging offtake by a month (July-June) yields a lower leakage estimate (17.6%), than if we match the offtake with the NSS reference period (August-July) – 18.2%. Both, however, are underestimates.
Adjusting for state contributions
Why so? Some states run an ‘expanded PDS’, providing PDS grain to non-NFSA beneficiaries using central contributions (for example, tide-over rations and non-NFSA allocations) as well as state contributions (for example, local procurement). For instance, Chhattisgarh’s State Food Security Act passed in 2012 made the PDS quasi-universal using local procurement. The all-India leakage estimates of 17.6-18.2% are underestimates because they only take into account central contributions, not state contributions.3 In such states, FCI offtake needs to be adjusted for state contributions.
There are 140 million non-NFSA beneficiaries, of whom at most 60 million are supported entirely through state contributions. If we adjust total offtake, for the state’s contribution to provide for non-NFSA beneficiaries, the all-India leakage estimate rises to 22%.
In Table 1, each row reports how leakage estimates vary depending on whether we use lagged (first row) offtake data or match it exactly with the HCES reference period (second row).4 The two columns present unadjusted (first column) and adjusted for state contributions (second column). The lowest leakage rate (17.6%) is when we used lagged offtake, without any adjustments for state contributions. The highest (22.1%) is if we use offtake for August-July adjusted for state contributions.5
Table 1. All-India estimates of PDS leakages (%), 2022-23
|
No adjustment for state contribution |
Adjusted for state contributions |
Lagged offtake, July 2022 to June 2023 |
17.6 |
21.6 |
Matched offtake, August 2022 to July 2023 |
18.2 |
22.1 |
A word on PMGKAY
Beginning in April 2020, PMGKAY continued till December 2022. Over this period, extensions were made in an ad-hoc manner. According to the Monthly Food Grain Bulletin, there were seven instalments of PMGKAY allocations during this period. Ad-hoc extensions are likely to have had the effect of increasing leakages: when people know what they are entitled to, they learn to claim it; but when there is lack of clarity about their entitlements, it enables cheating by PDS dealers. Drèze and Khera (2015) describe a similar situation with respect to large, ad-hoc APL (above poverty line) allocations, where card holders were unaware of their entitlements. This earlier experience suggests that PMGKAY allocations may have resulted in higher leakages during the reference period.
Impact of NFSA on PDS coverage
One of the major PDS reforms included in the NFSA, 2013 was an expansion of PDS coverage, that was aimed at reducing exclusion errors but had a ‘spillover effect’ on reducing leakages. In 2011-12, before NFSA, less than 50% of households had ration cards, and the proportion of households getting anything from the PDS was around 40% (see Table 2).6 That was an improvement over 2004-05, when less than a quarter (24%) of households were accessing the PDS.
The improvement between 2004-05 and 2011-12 was driven by early reforming states like Chhattisgarh and Odisha where the PDS coverage increased substantially in this period. For instance, in Chhattisgarh, there was a three-fold increase (from 21% to 63%) in the proportion of households using the PDS. According to HCES data, in 2022-23 the proportion of households buying from the PDS has increased further to 70%.7 In large part, this is the result of the rollout of the NFSA.
In spite of the improvement in coverage, the Centre is falling short of the coverage mandated by the NFSA (75% of rural population and 50% of urban population), that is, around 66%. Earlier work based on administrative data, suggested that only 59% had access to the PDS as NFSA beneficiaries (Khera and Somanchi 2020). Similarly, according to HCES, out of the 70% who access the PDS, only 57-61% have NFSA ration cards. The rest (5-9% according to the HCES) are non-NFSA beneficiaries.8
PDS reforms and leakages
PDS reforms undertaken by early reforming states such as Chhattisgarh and Odisha included a reduction in PDS prices, doorstep delivery of foodgrains, digitisation of records, deprivatising management of PDS outlets by handing over to panchayats (village councils), self-help groups, etc. These reforms were incorporated in the NFSA, 2013 (Chapter V).
As Table 2 shows, by 2011-12, PDS leakage estimates in the early reforming states had already crashed. Bihar more or less ‘copy-pasted’ the experiences of Chhattisgarh and Odisha, and the leakage rate went from 91% in 2004-05 to 24% by 2011-12.9 The 2022-23 HCES data suggest that leakage rates in these early reforming states were at 26.5% (Bihar), 14.8% (Chhattisgarh) and 13.7% (Odisha).
Madhya Pradesh (MP)_experienced a remarkable turnaround post-NFSA. Its turnaround was captured in small household surveys conducted in 2013, 2015 and 2016. In these surveys, we calculate the ‘purchase-entitlement ratios’ (PER), that is., the proportion of entitlements that PDS households actually buy. The PER is the converse of leakages. In the 2013 survey, the PER for the MP sample was 37%; when we revisited those households in 2015, the PER had risen to 96%. In another survey in 2016, PER in MP was at 83% (Drèze et al. 2019). The HCES corroborates the improvement in MP (26% leakage rate).
The 2016 survey included Bihar, Chhattisgarh, Jharkhand, Odisha and West Bengal, where the rollout of the NFSA was more or less complete.10 The improvement recorded in that survey (PERs were more than 80%) is confirmed by the HCES 2022-23. According to HCES 2022-23, more states have succeeded in reducing leakages: in Rajasthan it was 9%, 21% in Jharkhand and 23% in Uttar Pradesh (all one-time basket cases).
For some states, Table 2 presents an upper and lower bound for leakages. These include Andhra Pradesh, Karnataka, Odisha, and West Bengal. These are among the states that make state contributions to run an expanded PDS. The lower bound for Karnataka and Odisha is the unadjusted estimate. For Andhra Pradesh, the lower bound corresponds to non-NFSA beneficiaries getting grain for seven months (as stated in the Economic Survey), and the upper bound corresponds to assuming that they got it for 12 months. For West Bengal, two sources provide different number of non-NFSA beneficiaries (19 versus 28.8 million), so we have an upper and lower bound.
In addition, state adjustments are made for Chhattisgarh (where we directly use the offtake provided in the State Economic Survey) and Telangana (where without adjusting for state contributions, the leakage rate is negative, so only the upper bound is reported).
Many believe that the integration of Aadhaar, especially Aadhaar-based biometric authentication (ABBA), led to the improvements in the PDS. Data from primary surveys, however, do not support this. Two surveys in 2017 in Jharkhand studied this issue. A study by Muralidharan, Niehaus and Sukhtankar (2020) found that leakages before the introduction of ABBA were already less than 20%. Drèze, Khera and Somanchi (2020) reported that the PER in offline villages and ABBA villages were virtually same – 94% and 93%, respectively. Both surveys found little evidence of ‘ghost’ cards.
Somewhat puzzlingly, leakages in states where the PDS traditionally worked better have not experienced further improvements; in fact, in some estimated leakages have increased (for example, in Tamil Nadu from 12% in 2011-12 to 25% in 2022-23).
Table 2. Estimates of PDS leakages and coverage
|
Estimated leakage (%) |
Proportion of households buying PDS grain (%) |
||||
2004-05 |
2011-12 |
2022-23 |
2004-05 |
2011-12 |
2022-23 |
|
Andhra Pradesh |
23.2 |
22.0 |
0.8-12.9* |
57 |
35 |
85 |
Assam |
88.7 |
50.7 |
26.5 |
9 |
30 |
75 |
Bihar |
91.0 |
24.4 |
26.5 |
2 |
14 |
72 |
Chhattisgarh |
51.7 |
9.3 |
14.8* |
23 |
61 |
90 |
Gujarat |
51.7 |
67.6 |
40.3 |
25 |
28 |
47 |
Haryana |
82.7 |
49.0 |
18.2 |
4 |
17 |
38 |
Himachal Pradesh |
27.0 |
27.1 |
30.3 |
48 |
86 |
74 |
Jharkhand |
85.2 |
44.4 |
20.7 |
6 |
23 |
77 |
Karnataka |
28.7 |
34.7 |
14.4-28.9* |
49 |
77 |
70 |
Kerala |
25.6 |
37.1 |
36.8 |
28 |
93 |
80 |
Madhya Pradesh |
50.1 |
51.5 |
25.6 |
21 |
42 |
64 |
Maharashtra |
49.3 |
48.2 |
35.6 |
22 |
36 |
54 |
Odisha |
76.3 |
25.0 |
5.8-13.7 |
20 |
55 |
80 |
Punjab |
93.2 |
58.8 |
[-2.5]# |
0.4 |
19 |
40 |
Rajasthan |
93.9 |
60.9 |
9.2 |
10 |
18 |
57 |
Tamil Nadu |
7.3 |
11.9 |
25.1 |
71 |
62 |
86 |
Telangana |
- |
- |
21.1* |
- |
82 |
69 |
Uttarakhand |
59.4 |
34.9 |
38.8 |
21 |
35 |
56 |
Uttar Pradesh |
58.0 |
57.6 |
23.0 |
6 |
23 |
74 |
West Bengal |
80.6 |
65.3 |
16.6-22.3* |
13 |
34 |
89 |
India |
54.0 |
41.7 |
22.1 |
24 |
39 |
70 |
Notes: * Adjusted for state contribution based on either state data or estimating state contribution using number of state supported non-NFSA beneficiaries from Khera and Somanchi (2020). For Chhattisgarh, state offtake is from the Chhattisgarh Economic Survey 2023-24.
# Unadjusted estimates based only on offtake recorded in the Monthly Food Grain Bulletins. Punjab distributes grain every six months (unlike other states which have a monthly cycle). That may have led to over-counting of PDS purchases in the HCES.
Concluding remarks
The PDS is now a functional instrument of social policy, guaranteeing a modicum of food security to many. During the Covid-19 lockdowns, it formed the backbone of India’s relief efforts, along with MNREGA (Mahatma Gandhi National Rural Employment Guarantee Act). However, the PDS remains an endangered instrument of social policy, constantly subjected to ‘innovations’. These include cash transfer experiments, door-step delivery to people’s homes (in Delhi), imposing inappropriate technologies (such as Aadhaar-based biometric authentication), etc.
Instead of expending energy on ill-conceived measures (such as the ongoing e-KYC drive for the PDS) that can easily derail the PDS, scarce government capacity should be channelled to expediting the delayed Census, that is leading to the exclusion of over 100 million people. Other demands such as inclusion of more nutritious items like pulses and edible oil remain relevant.
The author would like to thank Ishan Anand for generously sharing his HCES data files. She is also grateful to Neha Arya, Jean Drèze, Md. Asjad and Raghav Puri for assistance, comments, discussions, and feedback. The views expressed in this post are solely those of the author, and do not necessarily reflect those of the I4I Editorial Board.
An abridged version of this post first appeared in The Hindu.
Notes:
- Antyodaya households are the poorest households identified by state governments.
- Under the 2020 PMGKAY, an additional quantity of rations was made available to NFSA recipients at zero cost.
- In cases such as Chhattisgarh, Andhra Pradesh, Telangana, West Bengal, the state contributions are large enough that, without adjusting for them, we get negative leakage estimates.
- In the HCES questionnaire, the reference period for PDS purchases is the “last 30 days”, which is why for some households, the state offtake in July matters.
- Ideally, we need state-wise information on ‘top-ups’, but this information is hard to come by, if available at all. If we allow for double-counting of offtake and assume that all 140 million non-NFSA beneficiaries are supported entirely through state contributions, the leakage estimate rises to 27%.
- This implies that approximately one-tenth of ration card holding households in 2011-12 were either not entitled to PDS grain, or were entitled but did not get any due to corruption.
- The increase from 40% (in 2011-12) to 70% (in 2022-23) is a combination of more entitled households and an increase in the proportion entitled households who actually get their entitlements.
- Four ration card categories in the HCES that could be NFSA cards are Antyodaya (5.6%), BPL (below poverty line) (24.3%), Priority (27.3%) and other (4.3%), and they add up to 61.4%. Around 5% reported having state ration cards, 15.7% had APL cards and 18.2% had no cards at all. According to Khera and Somanchi (2020), 11% accessed the PDS as non-NFSA beneficiaries.
- An important exception was that Bihar did not de-privatise PDS dealerships like Chhattisgarh and Odisha did.
- After the NFSA 2013 was passed, implementation was staggered. In some states (like Rajasthan) rollout was immediate whereas others took time (for example, Uttar Pradesh). Interestingly, often the big push came in the run up to the State Assembly elections (this was true in Bihar, Rajasthan, Uttar Pradesh and West Bengal).
Further Reading
- Drèze, Jean and Reetika Khera (2015), “Understanding Leakages in the Public Distribution System”, Economic and Political Weekly, 50(7).
- Drèze, Jean, Himanshu, Reetika Khera and Abhijit Sen (2015), “Clarification on PDS Leakages”, Economic and Political Weekly, 50(39).
- Drèze, Jean, Prankur Gupta, Reetika Khera and Isabel Pimenta (2019), “Casting the Net”, Economic and Political Weekly, 54(6).
- Khera, R and A Somanchi (2020), ‘A Review of the Coverage of the PDS’, Ideas for India, 19 August.
- Drèze, J, R Khera and A Somanchi (2020), ‘Balancing Corruption and Exclusion: A rejoinder’, Ideas for India, 28 September.
- Khera, Reetika (2011), “Trends in the Diversion of PDS”, Economic and Political Weekly, 46(21).
- Muralidharan, K, P Niehaus and S Sukhtankar (2020), ‘Identity Verification Standards in Welfare Programs: Experimental Evidence from India’, NBER Working Paper No. 26744.
Comments will be held for moderation. Your contact information will not be made public.