1758 - Revival of the Public Distribution System: Evidence and Explanations by Reetika Khera - EPW

VOL 46 No. 44 and 45 November 05 - November 

 SPECIAL ARTICLE
november 5, 2011 vol xlvi no 44 & 45  EPW Economic & Political Weekly36
Revival of the Public Distribution System: 
Evidence and Explanations
Reetika Khera 

 

I thank Diane Coffey, Jean Drèze, Sudha Narayanan and Dean Spears for comments and suggestions that helped improve this paper and Karuna Muthiah and Dipa Sinha for helping to coordinate the survey in Tamil Nadu and Andhra Pradesh, respectively. Most importantly perhaps, I would like to thank the investigators who participated in the survey and provided us new insights into and a deep understanding of the working of the PDS in the survey states. I acknowledge financial support from the Indian Institute of Technology, New Delhi. A pilot study was supported by the Planning and Development Unit, Department of
Economics, University of Allahabad.
 

Reetika Khera (reetika.khera@gmail.com) is at the Department of
Humanities and Social Sciences, Indian Institute of Technology, Delhi
and a visiting scholar at the Centre for Development Economics, Delhi
School of Economics.

 

 

Contrary to a common belief that India’s public
distribution system is irreparably dysfunctional, a
nine-state survey of the pds finds that the respondents
received 84-88% of their full entitlement. The implicit
subsidy for households below the poverty line from pds
foodgrains alone is roughly equivalent, in many states, to
a week’s nrega wages every month. The revival of the pds
can be traced, in large part, to a renewed political
interest which manifests itself in state initiatives such as
expanded coverage, reduced prices, computerisation of
stock management, etc. 
A large majority of the respondents preferred to receive
in-kind food transfers rather than cash transfers, except
in Bihar where the pds is still in very poor shape. Their
testimonies, and the survey findings, point to many
good reasons to be wary of a hasty transition to cash
transfers. Further improving the pds seems like a more
sensible way forward.
1 Introduction
T
he public distribution system (PDS) is among the most
important food security programmes of the Government
of India. Since the introduction of the Targeted PDS
(TPDS) in 1997, most field-based studies of the PDS have focused
on  inclusion and exclusion errors (see, e g, Swaminathan and
Misra 2001;  Hirway 2003; Khera 2008; Mahamallik and Sahu 2011).
 Using National Sample Survey (NSS) data, others have focused on
 leakages from the PDS (see, e g, Jha and Ramaswami 2010;
Himanshu and Sen 2011; Khera 2011c, among others). Several
other equally important questions, e g, the nutritional impact of
the PDS, policy changes in the past decade (especially at the state
level), the role of the PDS in providing food security, have
 remained neglected.
To address some of these gaps, and understand the ground
 realities, a survey of the PDS in nine Indian states was undertaken
in May-June 2011 (hereafter “PDS Survey”). This exercise is
 important for two reasons: one, there have been few independent
evaluations of the PDS in recent years, and, two, the PDS is
 expected to play an important role in the proposed National Food
Security Act.1 Given that cash is also being discussed as an alter-
native to the PDS, this survey included an effort to understand
people’s views on cash transfers.
Contrary to a common belief in policy debates (and academic
circles) that the PDS is an irreparably dysfunctional scheme, this
survey finds that many state governments have undertaken a
 series of measures that have made the PDS functional. By provid-
ing assured supply of foodgrains at nominal prices, the PDS has
become an important source of food security for rural house-
holds, or more precisely for eligible rural households. With the
notable exception of Bihar, below the poverty line (BPL) house-
holds in the nine sample states reported receiving almost their
full entitlement of grain (rice and wheat) on a regular basis. In
the three months preceding the survey, households reported re-
ceiving 84-88% of their full entitlement of PDS grain – up to 40 kg
of grain per household per month. Further, the implicit transfers
associated with these food entitlements are now quite substantial
due to the combined effect of an increase in market prices and
reduced PDS prices.
The main weakness of the current system is that it is restricted
to certain categories of households and that there are large
 exclusion errors. Until recently, the PDS was targeted mainly at
BPL households in most states. In many states,  increasing dissatis-
faction with this approach has led to a move towards a more
 inclusive, if not a universal, system. As discussed below, the
SPECIAL ARTICLE
Economic & Political Weekly EPW november 5, 2011 vol xlvi nos 44 & 45 37
 revival of the PDS in many states appears to be linked with this
expanded coverage.
Along with foodgrain, the PDS provides sugar and kerosene in
most states. In addition, several states have started providing
pulses and edible oil at subsidised prices. However, PDS supplies
of non-foodgrain commodities (pulses, oil, kerosene, sugar, salt,
etc) were not the main focus of this survey.
Drawing on the survey, this report examines different aspects
of the PDS: its role in ensuring food security, corruption and other
complaints, state-level diversity in the way it works and reform
measures undertaken by state governments. We also discuss peo-
ple’s views on cash transfers as an alternative to subsidised food
from the PDS. The reasons that were given for preferring cash or
food are enlightening and they help us understand rural realities
from the perspective of those who are directly affected. While
people’s preferences are not the last word on this issue, they do
need greater attention than they have received so far.
1.1 The Survey
The nine sample states are: Andhra Pradesh, Bihar, Chhattisgarh,
Himachal Pradesh, Jharkhand, Orissa, Rajasthan, Tamil Nadu
and Uttar Pradesh.2 In an earlier paper based on secondary data
for 2004-05 (Khera 2011c), the major Indian states were grouped
into three categories based on the performance of the PDS: “func-
tioning”, “reviving” or “languishing”.3 The selection of states for
the PDS survey relied on that grouping of states. We chose three
functioning states (Andhra Pradesh, Himachal Pradesh and
Tamil Nadu), three reviving (Chhattisgarh, Orissa and Uttar
Pradesh) and three languishing (Bihar, Jharkhand and Rajasthan).
This survey finds further support for this categorisation, except
that two more states from the languishing category (Jharkhand
and Rajasthan) seem to be “reviving”.
Two blocks in each of two sample districts in each state were
visited by the survey teams.4 Sample districts and blocks were
selected to maintain a balance between more and less developed
areas of the state, and also between different agro-climatic
 regions. For instance, in Bihar, Nalanda (in “south Bihar” and
two hours from the state capital) and Katihar districts (in “north
Bihar”, on the border with Nepal) were selected. In Tamil Nadu,
both sample districts (Dindigul and Dharmapuri) are known as
“backward” districts. Given Tamil Nadu’s exceptional PDS per-
formance, we were interested in seeing how the scheme fared in
the worst areas of the state.
In each of the 20 sample districts, six villages (three each in
two different blocks) were selected for investigation. Selected
villages were a random sample from the list of census villages
with a population of 500-1,500.5 Investigators were supposed to
interview 12 randomly selected households in each village.
In six out of nine sample states, surveyed households were a
random sample selected from the BPL list. In the remaining three
(Andhra Pradesh, Himachal Pradesh and Tamil Nadu), where the
PDS is universal or quasi-universal, the voter list was used for
sampling.6 As discussed below, the BPL lists are far from reliable,
but they were good enough for our purposes, with the partial ex-
ception of Jharkhand. If a household on the BPL list (or voter list)
happened to have an above poverty line (APL) card or no card at
all, investigators were instructed to skip that household; house-
holds with Antyodaya cards, meant for the poorest of the poor,
were included (these households are generally, but not 
always, sele cted from the BPL list). The survey (hereafter “PDS
Survey”) covered 1,227 house-
holds in the nine states.
Along with the household ques-
tionnaire, the survey collected in-
formation on basic  socio-economic
features of the sample village.
This included information on con-
nectivity, basic amenities, access
to local markets and so on. The
village questionnaire also included
an interview with the fair price
shop (FPS) dealer who was asked
about the running of the shop.7
The survey was conducted by
students from the University of
Delhi and other universities, along
with local volunteers,  after careful
training. Many of them had also
participated in a series of earlier
surveys of the PDS and other
social security programmes such
as the  National Rural  Employment
Guarantee Act (NREGA), Mid-Day
Meal Scheme and Integrated Child
Development Services (see, e g,
Drèze 2001, 2002a; Bhatia and
Table 1: Socio-economic Characteristics of Sample Households
     All    States     AP     BI     CH     HP     JH     OR     RJ     TN     UP
Female respondents (%) 48 63 37 44 45 42 50 43 63 44
Illiterate respondents (%) 66 58 86 79 33 67 79 70 54 62
Living in kachha homes (%) 49 24 50 76 53 73 62 38 13 50
Proportion (%) of 
 Dalits 28 22 46 17 49 10 10 48 15 43
 Adivasis 24 17 6 52 1 68 59 7 2 1
 OBCs 35 46 27 29 3 20 30 36 80 40
 Other 12 15 20 2 47 2 1 9 3 16
Proportion (%) of households with
 No land 36 29 73 43 7 22 25 59 36 26
 Less than one acre of land 61 62 99 69 70 70 56 76 56 81
Main occupation
 Self-employed in agriculture 37 37 13 38 56 34 55 12 41 53
 Casual labour 49 41 76 50 31 53 36 72 40 33
 Regular employment 3 1 1 0 5 1 1 7 11 2
 Other 11 21 10 11 8 13 9 9 8 12
Standard of Living Indexa

Low (SLI<=10) 70 30.2 96.0 86.7 41.5 95.7 85.2 81.5 18.0 92.5
 High (SLI>20) 4 10.3 0.0 1.4 7.6 0.0 0.7 0.0 12.2 0.8
Per capita (Rs/month) expenditure, non-food 450 524 375 289 649 289 356 663 631 342
NREGA employing in the preceding 12 months
 % getting no employment (0 days) 37 20 62 31 22 30 46 52 22 50
 % getting 100 days 7 8 0 5 9 2 4 23 13 0
 Average days per household 29 39 12 25 41 28 19 33 55 14
a
 The “Standard of Living Index” has been created by weighting the ownership of assets in this manner: four for a pucca house, two for a semi-pucca
house and zero for kachha; four for a flush toilet and zero for none; two for electricity and zero for none; two for piped water supply and one for
public sources of water supply; four for five or more acres of land, three for two to five acres of land, two for up to two acres of land and zero for no
land; four for ownership of a four-wheel vehicle and three for ownership of two-wheel vehicle; three each for ownership of a television set or fan;
one for ownership of a pressure cooker.
SPECIAL ARTICLE
november 5, 2011 vol xlvi nos 44 & 45 EPW Economic & Political Weekly38
Drèze 2002; Drèze and Goyal 2003; Khera 2002, 2006, 2008,
2011b, 2011c; Sinha 2008; Citizen’s Initiative for the Children
u nder Six 2006, among others). Each of the nine states covered by
the PDS survey was also covered in one or more of these earlier
 surveys. This, along with secondary data, helped to understand
how things have changed over time in these areas.
1.2 Socio-economic Background of Respondents
This section briefly describes the socio-economic characteristics
of the respondents and their households. Nearly half (48%) of all
 respondents were women (Table 1, p 37): the highest proportion
(63%) of female respondents were interviewed in Andhra
Pradesh and Tamil Nadu; the Bihar sample had the lowest (37%).8
Two-thirds of all respondents were illiterate, though there are
large interstate variations. For instance, Himachal Pradesh had
the lowest proportion of illiterate respondents (33%) and Bihar
had the highest (86%). Just over half (52%) of all respondents
were dalits  (scheduled castes) and adivasis (scheduled tribes),
one-third  belonged to the “Other Backward Classes” and 12%
belonged to other communities.
Table 1 provides further details of the socio-economic back-
ground of sample households. A large majority of households
were poor, with a small asset base (e g, more than one-third were
landless and 61% owned less than one acre of land); a quarter of
the households lived in kachha homes without electricity.
A “Standard of Living Index” (SLI) along the lines of the index
used in the National Family Health Survey (NFHS) has been
created using asset data collected in the PDS Survey.9 House-
holds score between 1 and 30 on the SLI constructed here. In
six survey states, more than 70% of the households had a “low”
SLI score (i e, SLI less than 10, see Table 1 for details). The
two southern states (AP and TN) and Himachal Pradesh in the
north are economically better off with 18-42% of households
with a low SLI.
2 The Public Distribution System
Under the TPDS introduced in 1997, households are provided
APL, BPL, or Antyodaya (introduced in 2001) cards.10 Antyodaya
cards, which enjoy a larger subsidy than BPL households, are
meant for the “poorest of the poor”. APL households were effec-
tively excluded from the PDS in 2001 when the central govern-
ment raised the APL issue price above market prices. However,
they gradually re-entered the system in recent years, in many
states, as market prices shot up while the issue price remained
unchanged (some states have also used the APL quota to expand
the coverage of the PDS – see below).
The proportion of BPL families that the central government is
willing to subsidise in each state is fixed in accordance with the
Planning Commission’s poverty estimates. The 1993-94 poverty
estimates are still used for this purpose, on account of a Supreme
Court order in the “right to food” case (PUCL vs Union of India and
Others, Writ Petition Civil, No 196 of 2001). To take into account
the increase in population, in 2000 the centre revised the BPL
figures for each state by applying the 1993-94 poverty estimates
to the projected population in 2000.11 Note that, Andhra Pradesh
and Tamil Nadu never really adopted the central government’s
BPL caps, and do not have a BPL category for the PDS.12 In Andhra
Pradesh, those who are entitled to PDS commodities are given
“white” cards and in Tamil Nadu these cards are referred to as
“rice cards”.13
Since 2003, many state governments have felt that the caps on
BPL cards imposed by the central government are too stringent.
In response to this, many have initiated state schemes (e g, the
Mukhya Mantri Khadya Sahayog Yojana in Chhattisgarh,
 Rajasthan and Madhya Pradesh) to issue more BPL cards (these
are referred to as “state BPL cards” in this paper). Household enti-
tlements of state BPL cardholders tend to be the same as those of
regular BPL households. In our sample, 62% of sample households
had BPL cards, and nearly one-tenth (8%) had state BPL cards.
Just over a quarter (27%) of the sample households possessed
 Antyodaya ration cards.14
2.1 PDS Entitlements
There have been two important developments with respect to the
PDS at the state-level in the past few years. One, there has been a
renewed political interest in the PDS, especially since 2007. There
are two manifestations of this: (a) a rejection of the APL-BPL
distinction imposed on state governments by the centre for the
purpose of the PDS; and (b) a reduction in PDS prices in many
states, including free grain in Tamil Nadu since June 2011. A 
second, more recent and less widespread development is the in-
clusion of pulses and edible oils among the commodities distrib-
uted at  subsidised prices.
There are wide state-wise variations in the list of commodities
supplied by the PDS. Table 2 (p 39) presents state-wise entitle-
ments of BPL cardholders including non-grain commodities that
are supplied through the PDS in each of the survey states. While
wheat and rice are generally provided only to BPL or Antyodaya
households, kerosene is meant for all households. Sugar is  another
commodity provided through the PDS. As mentioned earlier, this
survey focused primarily on the supply of wheat and rice.
Another point worth noting is that given the massive build up
of foodgrain stocks with the Food Corporation of India (FCI) in
2010, the Supreme Court ordered that excess stocks be reduced
by increasing PDS entitlements. Some states have taken advan-
tage of the additional allocation of grain by providing an
 additional 5 kg/month per BPL household.
2.1.1 Towards Universalisation
An encouraging trend observed in the states was one towards a
much more inclusive (even universal, in some cases) PDS. Tamil
Nadu has had a universal PDS for some time. Himachal Pradesh
also has a universal PDS, albeit with a difference – APL house-
holds pay a higher price than BPL households, though the entitle-
ments (in terms of commodities and quantities) of all households
are the same. In Andhra Pradesh and Chhattisgarh, the system is
“quasi-universal” with nearly 80% of the population entitled to
PDS commodities.
This trend towards universalisation of the PDS could be seen in
other states too. In Orissa, the PDS has been universalised in the
hunger prone “KBK region” (originally the Kalahandi-Bolangir-
Koraput districts, now divided into smaller districts).15 Rajasthan,
SPECIAL ARTICLE
Economic & Political Weekly EPW november 5, 2011 vol xlvi nos 44 & 45 39
Bihar and Jharkhand have all expanded their BPL lists to include
more rural households (see Table 3 in Khera 2011c). Uttar Pradesh
is the only exception among the nine sample states.
Expanded coverage has been made possible by using one (or
both) of two measures: state governments pay for expanded 
coverage (e g in Andhra Pradesh, Chhattisgarh and Tamil Nadu)
or central allocations to states are “spread thinner”. The central
government currently allocates 35 kg for each BPL household (so
long as state governments adhere to the Planning Commission’s
poverty estimates). While Chhattisgarh, Himachal Pradesh,
Jharkhand and Uttar Pradesh still give 35 kg per household per
month, in other sample states household entitlements are lower
than 35 kg per month (Table 2). Two states (Andhra Pradesh and
Tamil Nadu) have moved to per capita entitlements whereas in
Bihar, Orissa and Rajasthan BPL households are entitled to 25 kg
per month.
2.1.2 Reduction of PDS Prices
Along with increasing the number of households entitled to PDS
grain, six out of nine state governments have reduced issue prices
below the centrally fixed issue prices for BPL households – Rs 4.65/
kg for wheat and Rs 6.15/kg for rice. In Tamil Nadu, the govern-
ment provides 20 kg of free grain; in Chhattisgarh, Rajasthan,
Orissa and Andhra Pradesh, grain is provided at Rs 2/kg; in
Jharkhand the price is Re 1/kg (and free for Antyodaya house-
holds belonging to the “primitive tribal group” category).
 Meanwhile market prices have increased, so that there has been
a sharp increase in the implicit subsidy from the PDS.
To illustrate, consider the case of Chhattisgarh, where BPL
households are entitled to 35 kg of rice at Rs 2/kg (much the same
as in Jharkhand, Orissa, Andhra Pradesh and Tamil Nadu). The
value of this monthly ration at local market prices, net of what
people pay for it, is around Rs 600. This, in turn, is the equivalent
of about five days of NREGA wages every month, or 60 days per
year. To put this in perspective, the sample households in Chhat-
tisgarh had worked for 25 days on NREGA in the preceding
12 months, on average. The PDS is doing more than twice as
much as NREGA for them, and the two together now give them a
very important protection from poverty and hunger.
2.1.3 Beyond Rice and Wheat
Another positive trend is the tendency to expand the list of com-
modities made available through the PDS to include more nutri-
tious items. In Himachal Pradesh, all ration cardholders (irre-
spective of whether they are APL or BPL) can buy at least 1 kg of dal
and 1 kg of oil each month. Larger households can buy up to 3 kg
of dal and 2 kg of edible oil. Andhra Pradesh and Tamil Nadu also
provide dal and edible oil. In Uttar Pradesh, households  reported
 intermittent supply of “matar ki dal” (split peas); in Rajasthan too,
households reported getting dal and oil briefly during 2009-10,
when market prices suddenly
jumped. There was an expec-
tation among res pondents that
these would be reintroduced.
The Chhattisgarh state govern-
ment has introduced chana dal
on a pilot basis in Bastar district.
2.2 BPL Lists
Two BPL censuses (in 1997 and
2002) have been conducted so
far by state governments. BPL
lists prepared on the basis of the
2002 Census were held up on
account of a stay order from
the Supreme Court in the “right
to food” case. The Supreme
Table 2: PDS Norms for BPL Cardholdersa
     Grain    Entitlements     Other    PDS    Commodities
Andhra Rice: 4 kg/capita (Rs 2/kg) Kerosene, sugar, salt 
 Pradeshb Wheat: Fortified atta Pulses (1-2 kg, Rs 34-50/kg)
  Oil (1 litre, Rs 35/litre)
Bihar Rice: 15 kg (Rs 6.7/kg) Kerosene
 Wheat: 10 kg (Rs 5.22/kg)
Chhattisgarhc Rice: 25 kg (Rs 2/kg) Kerosene, sugar, salt
 Wheat: 10 kg (Rs 2/kg) Chana dal on experimental basis
Himachal  Rice: 15 kg (Rs 10/kg for APL  Kerosene, sugar, salt
 Pradesh and Rs 6.85/kg for BPL) Pulses (1-3 kg, Rs 20-35/kg)
 Wheat: 20 kg (Rs 8.5 for APL  Edible oil (1-2 litres, Rs 40)
 and Rs 5.25 for BPL)
Jharkhand Rice: 35 kg (Re 1/kg) Kerosene
 Wheat: Discontinued
Orissa Rice: 25 kg (Rs 2/kg in KBK  Kerosene, sugar
 region; Rs 9.3/kg for
 APL elsewhere)
 Wheat: Unclear
Rajasthan Rice: 0 Kerosene, sugar
 Wheat: 25 kg (Rs 2/kg)
Tamil Nadu Rice: 20 kg (free)d Kerosene, sugar, salt
 Wheat: Fortified atta Pulses (2 kg, Rs 30/kg) 
 (1 kg, Rs 11/kg) Oil (1 litre, Rs 25/litre)
Uttar Pradesh Rice: 20 (Rs 6.15/kg) Kerosene, sugar
 Wheat: 15 (Rs 4.65/kg)
a
 Antyodaya cardholders get 35 kg/month/household in all states, and they pay Rs 3/kg for rice
and Rs 2/kg for wheat. However, in Orissa rice is given to them at Rs 2/kg; Chhattisgarh and
Jharkhand have reduced the prices further to Re 1/kg. In Tamil Nadu and for so-called "primitive
tribal groups" with Antyodaya cards in Jharkhand rice is free.
b
In Andhra Pradesh entitlements are specified in per capita terms, with each person entitled to
four kg/month, with a maximum of 20 kg per family.
c
 In Chhattisgarh, there are three additional state-specific of ration cards. Grey cards (for STs and
SCs) and Saffron (OBCs) enjoy the same entitlements as BPL cardholders. Widows and senior
citizens are also given saffron cards, but with smaller entitlements (10 kg rice at Rs 2/kg). Persons
with disabilities are given green cards with the same entitlements as widows and senior citizens.
d
 Single-person households are entitled to 12 kg of rice and 16 kg for two-person households.
Table 3: Purchase of PDS Grain
     Average    Purchase    (Kg/Month)     Average     Purchase-Entitlement
     BPL     Antyodayaa     Entitlement    for     Ratio    (PER)    in    the    Past
     March     April     May     Average     March     April     May     Average     BPL    Cardholders     Three    Months
                                                 (Kg/Month)b     BPL     Antyodaya
Andhra Pradesh 14.8 15.1 14.9 14.9 35.3 35.3 30.9 33.8 15.1  100 97
Bihar 11.5 10.8 11.4 11.2 19.3 14.9 14.6 16.3 25  45 47
Chhattisgarh 33.1 34.0 32.8 33.3 34.1 33.1 35 34.1 35  95 97
Himachal Pradesh 41.6 36.9 32.9 37.1 43.1 38.1 31.8 37.6 35-40d  92-100d 94-100
Jarkhand 29.5 22.5 22.8 24.9 30.4 22.1 19.4 23.9 35  71 68
Orissa 20.9 26.0 40.6c 29.2 31.6 32.7 40.6 35.0 25-30d  97-100d 100
Rajasthan 25.6 25.3 27.1 26.0 35 35.3 35 35.1 25-30d 86-100d 100
Tamil Nadu 17.3 18.6 17.8 17.9 33.5 28.5 30.3 30.8 19.5  92 88
Uttar Praesh 31.2 31.0 29.8 30.7 33.7 34.4 33.9 33.9 35-40d  77-88d 85
All states 23.7 23.5 24.7 24.0 32.6 29.6 29.1 30.4 27.3-28.7 84-88 87
a
For Antyodaya the entitlements are 35 kg/month everywhere except in Uttar Pradesh where it is 40 kg (temporarily) on account a Supreme Court
Order in the right to food case (see text).
b
Entitlements applicable to sample households, based on the norms given in Table 2.
c
Due to "clubbing" (e g, several households getting nothing in March and April, and then getting 100 kg in May).
d
Entitlements are reported as a range to account for the five kg extra due to a Supreme Court order in the right to food case (see text). The range gives 
 the lower and upper limit for entitlements as well as PER.
SPECIAL ARTICLE
november 5, 2011 vol xlvi nos 44 & 45 EPW Economic & Political Weekly40
Court lifted the stay in 2005, with the proviso that names of
households on the 1997 BPL list could not be struck off the 2002
BPL list and that  effective appeal procedures should be put in
place for all households. Even after the stay was lifted, not all
states adopted the new BPL list. Meanwhile, as noted earlier, state
governments began to spend resources on expanding the PDS
with the introduction of state BPL cards.16 As a result of this chain
of events, in each of the states where the BPL list was used for
sampling (with the notable exception of Chhattisgarh), there was
confusion  regarding to which BPL survey the BPL list corre-
sponded (1997, 2002 or something more recent).
In Jharkhand the situation was particularly messy. The distri-
bution of BPL cards was largely based on the 1997 survey. How-
ever, the distribution of PDS rations was on the basis of a separate
list (let us call it the “PDS list”) which the FPS dealer had, and did
not quite coincide with the BPL list. During the survey, we came
across households whose names were on the BPL list and who had
a BPL card, but who did not get rations. On the other hand, a few
households did not have a ration card, but were getting rations
because their names appeared on the PDS list! This situation is
the result of expansion of the PDS beyond the BPL list in recent
years and of poor governance in the state.17 The lack of clarity in
terms of who is entitled to PDS rations opens the door to corrupt
practices on the part of dealers and other officials.
The misclassification of households (inclusion and exclusion
errors) on BPL lists is well established (e g, Drèze and Khera 2010a
and the literature cited earlier). Documenting such errors was
not explicitly part of the survey, yet the investigators came across
serious inclusion and especially exclusion errors.
On the brighter side, the BPL lists are becoming more transpar-
ent (see section 7.3 below). Most BPL lists are online. In Chhattis-
garh and Orissa, the teams found that an exercise to update and
verify BPL lists had been conducted recently. However, even this
exercise has had some negative fallouts – the ration cards of those
who were not available during the verification drive (e g, mi-
grants) have been cancelled. The drive has also given the FPS
managers an additional excuse to deny some people their ration.
Further, since BPL lists were used for sampling purposes it gave
us a chance to investigate the presence of “ghost” cards on the
BPL list. The teams found hardly any bogus names on the BPL
lists. Across all states, there were less than 10 names (out of 1,227)
on the BPL lists that the survey teams were unable to locate.
Nearly all households (93%) possessed only one ration card. A
small proportion (6%) of households had two; of these, one-third
were joint families. This is significant in the light of claims that
bogus and/or duplicate cards are a major source of corruption in
the PDS. A possibility remains that bogus cards are floating “out-
side” the official BPL lists; this could be verified by checking
whether actual PDS allocations (say, at the state or district level)
exceed what one would expect based on the official BPL lists.
2.2.1 Poor Coverage
In Bihar, Jharkhand and Uttar Pradesh, the BPL lists were very
patchy and there were enormous “exclusion errors”. For instance,
in Jharkhand, the team found entire hamlets that did not have
BPL cards. In Uttar Pradesh, the survey team sometimes found it
difficult to locate 12 BPL households in one village. In Orissa’s
Nuapada district (in the KBK region), many households com-
plained of having no ration card at all. Often, younger house-
holds – i e, newly married couples – did not have any ration card.
Aside from well-known general problems associated with the
entire BPL identification exercise, one important source of exclu-
sion errors is the absence of arrangements to update the BPL list
over time. The current system of issuing ration cards is inflexible
– centrally imposed caps on the permissible number of BPL cards
combined with population increase (in many states, the last BPL
census was held about 10 years ago) lie behind the rigidity in the
system. The block supply officer in Nuapada’s Sinapalli block told
us that the order for issuing new cards has to be given from Bhu-
waneshwar, the state capital. The district or block administration
does not have the authority to do so. In many states, a satisfactory
system for issuing new cards needs to be put in place (e g, whereby
names for new ration cards could be sent from the gram pancha-
yat to the district or state capital for approval).
2.2.2 Supreme Court Order on Antyodaya Cards
The Supreme Court has issued an interim order in the right to
food case directing certain groups to be included compulsorily on
the Antyodaya list. The team found several violations of this – e g,
many eligible widows did not have Antyodaya cards. There were
34 landless widows in the sample with BPL cards, although by
virtue of being landless alone they should have been on the
 Antyodaya list.18 Another group that is entitled to Antyodaya
card is that of so-called “primitive tribal groups” (or, PTGs). In the
case of PTGs, there has been some attempt to cover all house-
holds, but even here the team found exclusion errors, e g, in
Gopikander block of Jharkhand several Parhaiya households
(classified as PTGs) did not have any ration card.
3 Corruption and the PDS
As mentioned earlier, corruption in the PDS has been one the big-
gest concerns in recent years (see, e g, Jha and Ramaswami 2010;
Himanshu and Sen 2011; Kotwal, Murugkar and Ramaswami
2011; Khera 2011c), including in this survey.
The survey questionnaire included a very careful accounting
of purchases of PDS grain by sample households. Respondents
were asked three complementary questions to arrive at a reliable
estimate of the extent to which they were able to secure their full
entitlements: (1) how much they “normally” get from the FPS;
(2) how much they got the last time they bought PDS grain; and
(3) how much they bought, month by month, from March to 
June 2011.
3.1 Improvement in ‘Purchase-Entitlement Ratio’
The purchase-entitlement ratio (PER) refers to the proportion of
full entitlement that is purchased by BPL households. A low PER
could be due to corruption in the system or lack of demand
(possibly related to low quality of PDS grain). One of the major
findings of the survey is that as far as BPL and Antyodaya
 cardholders are concerned, there has been a marked improve-
ment in the PER, except in Bihar (Table 3, p 39).19 As Table 3 shows,
average  purchase in the past three months ranged between
SPECIAL ARTICLE
Economic & Political Weekly EPW november 5, 2011 vol xlvi nos 44 & 45 41
24.0 and 30.4 kg/month (for BPL and Antyodaya cards respec-
tively). The average entitlement of BPL cardholders (averaged
over all states) is  between 27.3 and 28.7 kg/month.20 The aver-
age PDS purchase in the past three months (24 kg/household
per month) is at least 84% of the monthly entitlement (28.7 kg/
household per month).21
A more demanding benchmark is to look at the proportion of
households for whom PER equals 100, i e, the proportion of house-
holds who get their full entitlement normally. Three-quarters of
respondents reported getting their full quota in the nine survey
states.22 This rises to 80-81% if we exclude Bihar from the sam-
ple. The two important exceptions to this general pattern of high
PERs are Bihar and Jharkhand. In Bihar, the situation remains
grim (even though it seems to have improved even there, at least
by Bihar’s standards).23 Only 18% of BPL households reported get-
ting their full quota in Bihar, followed by Jharkhand where 25%
get their full entitlement (Table 6, p 42).
NSS data on PDS purchases from 2004-05 onwards show a simi-
lar revival in the PDS (Himanshu and Sen 2011; Khera 2011c).24
Some of the measures that have contributed to this turnaround
are discussed below. Many household respondents and others
(the ration dealer, village head, etc) attributed the improvement
in the PDS to the increase in the gap between market and PDS
prices mentioned earlier. The increase in implicit subsidy has had
the effect of enhancing voice, and no one seems willing to forego
their ration any more.
Leakages in wheat supply among “rice states” has perhaps not
been properly captured in the survey.25 For instance, in Tamil
Nadu, respondents complained that the supply of wheat/atta is
irregular. Respondents in other rice states (Andhra Pradesh,
Chhattisgarh and Orissa) also voiced such complaints. Much of
this is also linked to uncertainty among both dealers and card-
holders regarding their wheat entitlements. However, wheat dis-
tribution in these rice-eating states is relatively small.
3.2 Corruption in Other Commodities
The PDS survey did not include a similar careful accounting for
other PDS commodities (e g, sugar, dal, edible oil and kerosene).
Sample households often complained about corruption in the
 distribution of kerosene and sugar. In Tamil Nadu, some card-
holders complained of irregular supply of dals and pulses. This
survey did not match kerosene and sugar entitlements of house-
holds with  actual purchase. There are two reasons why this is
difficult to do: one, entitlements of these commodities are not
clear, not even to FPS managers in some cases. Two, interviews
with the FPS managers and block supply officers suggested that
the supply of these commodities to the district/block is irregular.
In such cases, it is not clear whether cardholders do not get these
commodities because of supply issues or because commodities
have been siphoned off.
3.3 Awareness of Entitlements
As far as awareness of grain entitlements is concerned, generally
people knew what they were entitled to. For instance, in the case
of Antyodaya cardholders, less than 3% were unable to answer the
question on grain entitlement correctly. Among BPL cardholders,
the correct answers ranged between 84% and 100%.26 However,
in the case of other commodities such as sugar, kerosene, dals
and oils there was a lack of clarity and irregularity in the supply
of these commodities.
3.4 Overcharging
There were hardly any instances of “overcharging” (charging
more than the official issue price) for PDS grain. In Jharkhand
(where BPL households now pay Re 1/kg for rice) we heard some
complaints of people paying Rs 35 each month, even though they
were given only 32 kg of rice. One factor that has helped to pre-
vent overcharging is the reduction in prices (e g, from Rs 4.65/kg
of wheat to Rs 2/kg in Rajasthan) and the use of round figures. It
is much easier for people to calculate the total amount due when,
say, wheat is priced at Rs 2/kg for 25 kg, than to multiply Rs 4.65
by 35.
Here again, as far as other commodities are concerned, there
was a lot of confusion and also more reports of overcharging.
Among the worst cases were FPS managers in Dumka (Jharkhand)
who openly told us that though the official price of kerosene is
Rs 13.50/litre, they charge Rs 16/litre as this was the “norm”.
4 Assessment of the PDS
Respondents were asked about the importance of the PDS in their
lives. A large majority (80%) considered it “very important”. If
one includes those who said it was “quite important” the propor-
tion increases to 98%. This sentiment is also conveyed in the
qualitative data from the survey, especially the responses to
cash transfer as alternative to the PDS. Many respondents
pleaded that the FPS not be closed (“Society nahi band honi cha-
hiye”).  Respondents in Warangal district (Andhra Pradesh) said
“even if you give me Rs 1 lakh, I will opt for rice”, or “you want to
deny us our food?” A female respondent in Tamil Nadu followed
the survey team to persuade them that closing down the ration
shop was a bad idea. Similar sentiments were echoed in most
states (see e g, Nag 2011 and Puri forthcoming).
People’s perception of the overall change in the PDS over the
past five years were also recorded. In Himachal Pradesh,
Jharkhand, Orissa and Tamil Nadu, half or more respondents felt
that things had improved. One-tenth of all respondents said that
things had become worse over the past five years.27
Table 4: Main Complaints with Fair Price Shops
     Proportion    (%)    of    Respondents
Quality related concerns
 Poor quality of grain at the time of last purchase 15
Lack of predictability
 Days of opening are not fixed 38
 Inadequate information regarding the opening of FPS 20
Access to FPS
 Distance to FPS more than two km (3 km) 19 (9)
 Last visit (travel + queuing) took more than 
 two hours (three hours) 45 (30)
Days of opening are not adequate 48
Reasons for not buying full ration in the past three months
 No supply to the ration shop 10
 Supplies ran out by the time respondent went to the FPS 10
 FPS dealer refused to give full quota 18
Dealer's attitude is unhelpful 18
SPECIAL ARTICLE
november 5, 2011 vol xlvi nos 44 & 45 EPW Economic & Political Weekly42
4.1 Quality
Overall, 38% of respondents reported getting good quality
grain at the time of their last purchase; nearly half (47%) said it
was of “fair” quality and 15% said that the grain they received
was of poor quality (Table 4, p 41). In Andhra Pradesh and
Himachal Pradesh there were hardly any complaints about the
quality of PDS grain. In Bihar, however, close to one-third (33%)
of the respondents reported getting poor-quality grain at the
time of their last purchase.
The qualitative data, however, suggest that perhaps the 
quantitative data do not capture grain quality issues adequately.
Many interviews were interspersed with complaints regarding
quality. This was especially true in Pali district (Rajasthan)
where some households reported finding 2-3 kg of stones in the
wheat out of the 25 kg they bought each month. In Chhattisgarh,
respondents had clear preferences with respect to “usna” 
(parboiled) and “arwa” (fine) rice. The complaints in these 
areas often related to getting the less preferred type of rice. 
In Tamil Nadu, some households said that they prefer to pay
more and get better quality rice than get free or cheap rice 
of poor quality. They said that the rice they get has to be 
mixed with other rice to be edible and is suitable only for 
making idlis and dosas. Further evidence of the  quality of grain
being an issue comes from the fact that when people were 
presented with the option of buying PDS grain or getting cash,
one-fourth of all sample households cited grain quality among
the relevant considerations.
4.2 Regularity and Predictability
There has been much improvement in the predictability and
regularity of PDS distribution and opening of FPSs. Most state
governments have now instituted a system of pre-announced
and fixed dates for distribution of PDS rations. For instance,
in Rajasthan, FPSs are open from the 15th to the 22nd of
each month.
Regularity and predictability in opening days and hours is a
major convenience for rural households, and also important to
prevent diversion of PDS quotas. In many of the sample states, it
used to be common for private dealers to open the FPS whenever
they liked. This meant that households may or may not hear
about the FPS being open. By the time they did get to hear of it,
the dealer would tell them that stocks are over. Further, when
opening days are predictable, households can plan their budgets
better and arrange to have cash in hand on those days. Reports
from households of being told that their quota had “lapsed” have
certainly gone down. More than 70% of sample households said
that their quota had not been diverted to the open market even
once in the past six months (others either said it had or were not
sure). This simple and effective reform measure needs to be
 enforced strictly in all states.
This step was welcomed by the sample households and yet it
was clear that this rule needs better enforcement. Though FPSs
are supposed to remain open throughout the month (as per
 Supreme Court orders) or at least for a week (as per the newly
instituted schedule in many states), respondents complained that
shops would be open only for two to three days. This also  resulted,
in some areas, in long waiting hours (sometimes the whole day to
buy the month’s quota).
It was not uncommon to find FPSs that remain open through-
out the month in Himachal Pradesh and Tamil Nadu. Even in
states such as Chhattisgarh and Orissa, some outlets remain open
through the month. In many places, respondents said that deal-
ers were willing to sell them their ration even if the opening days
of the FPS were over. Some households also said that they could
buy two months’ ration together, if they had been unable to pur-
chase it in the previous month.
Table 5 brings out the association between the PER and
 regularity of opening of the FPS. The pattern is striking (more
regularity, less diversion) and is not coincidental. Bihar is the
state with highest levels of reported  under-purchase (82%) and
lowest  levels of predictability in the opening of the FPS – only 7%
of respondents felt that opening hours were predictable. In
Jharkhand, similarly, reported  under-purchase is relatively high
along with lack of predictability in the opening hours of the
 ration shop. Rajasthan lies at the other end of the spectrum
where 91% of respondents said that the opening days of the FPS
were known to them in advance; in  Rajasthan, most sample
households  reported getting their full  ration “normally”.
4.3 Accessibility
Generally, the accessibility of the PDS is quite remarkable: the
 local FPS was 1.4 km away from the respondent’s house on aver-
age, and within 3 km for 91% of respondents (Table 6). Respon-
dents reported that, on average, the last round-trip to the FPS
 (including commuting and waiting time) took two hours. Distance
to the FPS remains an issue in some of the remoter regions of the
sample areas.
The accessibility of the PDS is a major convenience for rural
households. This is not an insignificant achievement given the
earlier literature on urban bias in the PDS (Dev and Suryanaray-
ana 1991; Howes and Jha 1992; Suryanarayana 1995).
 Table 5: Predictability of Fair Price Shops
     Average    "Purchase-     Proportion    (%)    of     Proportion    (%)    of    Respondents   
     Entitlement    Ratio"         Respondent    Who     Reporting    That    Days    of   
     (PER)    in    the    Past         Report    That    They    “Normally”     Opening    of    FPS    Are    Fixed   
     Three    Monthsa     Get    Their    Full    Entitlement
Rajasthan 86-100 100 91
Tamil Nadu 92 85 95
Chhattisgarh 95 97 72
Orissa 97-100 98 73
Himachal Pradesh 92-100 86 71
Andhra Pradesh 100 79 70
Uttar Pradesh 77-88 77 56
Jharkhand 71 25 9
Bihar 45 18 7
All states 84-88 75 60
a
 "Purchase-entitlement ratio" refers to the proportion of full entitlement purchased by
households (see Table 3 for further details).
Table 6: Access to Bank or Post Office vs PDS
     Average     Average    Time         Proportion    (%)    of    Respondents    Who    Reported   
     Distance    (Km)     Taken    on    Last     Availability    within     Taking    More    Than   
          Visita    (Hours)     2        km    (3    km)     One    Hour    at    Last    Visit
Bank/post office  5.2b 3.25 37 (53) 76
Fair price shop  1.4 2.10 81 (91) 53
a
 "Time taken on last visit" includes travel and queuing time. b If one includes two outliers, the
average distance to the bank increases to 5.4 km.
SPECIAL ARTICLE
Economic & Political Weekly EPW november 5, 2011 vol xlvi nos 44 & 45 43
5 Hunger and Food Intake
5.1 Hunger in the Survey Areas
An important function of the PDS is to ensure food security in a
limited sense, viz, protection from hunger. Survey respondents
were asked if any member of the household had to skip a meal or
sleep hungry in the three months preceding the survey.28 More
than one-fifth (22%) said yes (Table 7).
The interstate contrasts are striking. The proportion reporting
hunger is very low in Himachal Pradesh (6%), Orissa (9%), Tamil
Nadu (6%) and Uttar Pradesh (7%). In Andhra Pradesh and
Chhattisgarh, it is around 15%. In the other states, the figures are
not negligible: over a quarter of the households in Jharkhand
 reported skipping meals. In Rajasthan, the proportion rises to
36% and a whopping 70% in Bihar! It is perhaps not an accident
that the highest levels of incidence of hunger (by a long margin)
were in Bihar, where the PDS is in bad shape. Bihar is of course
very poor, too, but so are, say, Chhattisgarh and Orissa. At the
other end, the two states with an exemplary PDS (HP and TN)
 report very little hunger. While the state of the PDS may not be
the only factor behind these contrasts, they are certainly consist-
ent with the idea that the PDS has become an important tool of
food security in rural India.
5.2 Diversification of Diets
Diets in many survey areas remain heavily cereal-based
 (Table 7). Per capita monthly cereal consumption, as reported in
the  survey, is nearly 15 kg on average.29 Not only is cereal con-
sumption high, there does not seem to be much dietary diversity
in people’s meals. Overall, 7% of sample households ate only rice
or roti for their evening meal on the day before the survey. The
figures for Bihar and Rajasthan are alarmingly high – 13% and
29% respectively.30
Half of all sample households reported eating vegetables
every day in the week preceding the survey the corresponding
 proportions for fruit, eggs and meat are negligible: 5.4%, 0.7%
and 0.4%, respectively.
The last three columns of Table 7 report the proportion of
households which did not consume dal, fruit or eggs and meat
on even once in the week preceding the survey. The state 
patterns are interesting: not surprisingly, in states where the
PDS supplies subsidised dal, this proportion is negligible (1-3%
for Andhra Pradesh, Himachal Pradesh and Tamil Nadu). The
lowest consumption of dal is reported in Bihar, Jharkhand, 
Rajasthan and Uttar Pradesh. Consumption of animal proteins
(eggs and meat) is abysmally low in Himachal Pradesh, Rajasthan
and Uttar Pradesh.31
5.3 On Millets
It is a common perception among policymakers that people are
not interested in buying cereals other than wheat or rice from the
PDS.32 Contrary to this perception, a majority of respondents
(79%) said that they would buy millets or maize should they be
provided at subsidised prices under the PDS. In Tamil Nadu, some
respondents even said that they would be willing to take ragi 
in lieu of rice.
There is a strong case for the introduction of millets and maize
in the PDS. First, these grains are nutritionally superior to “fine”
grains such as wheat and rice. Second, these crops are suited for
dryland farming conditions that prevail in large parts of the
country. Third, there is a huge potential for procurement of
 millets – in 2008-09, only 4% of total millet production was
 procured by the government. Fourth, procurement of millets/
maize would also be cheaper than buying wheat and rice. For
 instance, the economic cost of bajra in Rajasthan was half of that
of paddy in 2010-11. Finally, the introduction of these grains
would also impart a “self-selecting” character to the PDS as these
grains are known to be more popular among the poor than richer
rural households.
6 Cash vs Food
In each of the sample households, the investigators had detailed
discussions with the respondents regarding their views on cash
and food. The question posed to the respondents was, how they
would feel if the FPS was closed and instead of PDS grain, the
government opened an account for them and deposited cash each
month in that account. The amount deposited would be equal to
the market value of the commodities they currently purchase
from the FPS (i e, whatever it would cost them to buy these
 commodities on the market).33 It was clarified that this amount
would be adjusted with price increases, so that it always enables
people to buy from the market whatever they are getting today
from the PDS.
Based on the experience of pilot testing the questionnaire, we
were prepared for some respondents being unable to engage eas-
ily with this hypothetical question. Given that our interest was as
much in learning about people’s preferences as with understand-
ing how they think about these issues, investigators were told to
play the “devil’s advocate” if need be, presenting the advantages
of the other option to the respondent. We were, however, pleas-
antly surprised to find that a large majority of households were
able to articulate their preferences. Only 7.5% of all respondents
were unclear or gave inconsistent replies. The remaining 92.5%
were able to state clearly whether they preferred food or cash, or
expressed a “conditional preference” for one of those two options
Table 7:  Hunger and Dietary Diversity
     Monthly                                      Proportion    (%)    of    Households    That
     Cereal         Reported    Any         Reported             Did    Not    Consume    the    Following           
     Consumption     Member    Having     Eating    Only         Items    Even    Once    in    the    Past    Week
     (Kg,    Per    Capita)     Skipped    Meals    in     Rice    or     Dal     Fruit     Eggs    or    Meat
          the    Preceding     Roti    at    the
          Three    Months     Last    Meal        
          Preceding   
Bihar 18.5 70 13 15 88 66
Rajasthan 17.7 36 29a 15 90 91
Jharkhand 13.9 26 3 12 47 55
Chhattisgarh 17.2 17 5 9 82 56
Andhra Pradesh 11.7 16 4 1 54 9
Himachal Pradesh 13.2 6 1 3 66 87
Orissa 15.8 9 2 4 72 41
Tamil Nadu 12.1 6 5 1 34 36
Uttar Pradesh 13.4 7 2 14 79 84
All states 14.9 22 7 8 68 58
a
In Rajasthan, the questionnaire did not capture certain frequently consumed food items, such
as buttermilk (see text).
SPECIAL ARTICLE
november 5, 2011 vol xlvi nos 44 & 45 EPW Economic & Political Weekly44
(e g, “I prefer food if good quality of grain is assured” or “I prefer
cash if we are assured of the money being deposited on time
every month”). Most of the respondents were also able to explain
why they preferred one or the other, and in many cases the expla-
nations were very insightful. Their thoughtfulness is also reflected
in the fact that about one-fifth of the respondents were able to list
at least one problem with both cash and food.
People’s preferences depended on a combination of pragma-
tism, shrewdness and deep understanding of the local circum-
stances. For example, a widow in a remote Maoist-affected block
of Nuapada district (Orissa) with no local transport but with a
functional PDS was as likely to opt for food as a widow from a
Bharatpur village (Rajasthan) with easy access to banks and
markets where she suspected the PDS dealer was cheating her
was to opt for cash.
When some households were asked what they would opt for
if the cash amount was double the market value of their current
PDS entitlements, those who switched from food to cash made
mental calculations about whether the additional hassles
 associated with taking cash would be compensated for by the
larger amount being offered. A mid-survey informal review
(conducted when teams had concluded the survey in the first of
the two districts in each state) suggested that most households
were in  favour of food over cash transfers. To get a sense of
whether and to what extent this decision was linked to the
amount of cash  offered, investigators were asked to offer
 double the market value of their current PDS entitlement, as a
follow-up question.
6.1 People’s Preference between Cash and Food
There is a very clear pattern in the preference between cash and
food – in states where the PDS is functional, most people did not
want to hear of cash transfers (Table 8 and Figure 1). However, in
areas with a dysfunctional PDS, notably Bihar (where the system
functioned so poorly that some households had received their
 entitlements only three times in the past 12 months), many
 respondents were quite open to the idea of cash transfers. The
general sentiment in these villages was, “give us something, so
long as it reaches us”.
Overall, more than two-thirds of the respondents expressed a
clear preference for food over cash; less than one-fifth (18%)
were in favour of cash over food (Table 8).34 Clubbing conditional
and unconditional preferences for food, the proportion wh0
p refer food rises to 73%. State-wise preferences are reported 
in Table 8.
Another interesting finding is that even in states where 
the proportion of respondents preferring food over cash is com-
paratively “low” (e g, 60% in Rajasthan), the converse prefer-
ence (i e, for cash over food) remains uncommon. In Rajasthan,
only 15% of respondents were in favour of cash over food. A 
related point is that many of those who were in favour of food
felt very strongly about it. As mentioned earlier, some respond-
ents were quite  agitated when the cash proposal was discussed
with them.
Interestingly, the responses of men and women were quite
similar. Among female respondents, 71% opted for food  compared
to 63% among male respondents. While men were more favour-
ably disposed to the cash option (21% said they preferred cash),
the corresponding figure for women is only slightly lower (15%).
Among the social categories, adivasis in the sample were most
 favourably disposed to food (83%) with only 8% in favour 
of cash.
6.2 Readiness of the System
One very basic indicator of readiness of the system for a transi-
tion to cash transfers is whether people have bank or post office
accounts. Nearly three-quarters of sample households already
had a bank or post office account – most of them were opened for
NREGA wage payments. The lowest coverage was in Bihar (46%)
and the highest in Himachal Pradesh (87%). Of those households
that had a bank or post office account, nearly two-thirds (68%)
had an account either in a nationalised bank (46%) or in “other
banks” (22%). Post office accounts were also common – 30%
 reported having a post office account. Only 5% of households had
more than one account.
In the light of the relatively high rates of “financial inclusion”
already achieved, the overwhelming preference for food in most
sample states is all the more interesting. It suggests that the
 preference for food is not just a reflection of playing safe (i e, the
rejection of cash is not just on account of being completely unfa-
miliar with the new system). Further, this also suggests that
“readiness” to transition to cash transfers goes beyond opening
bank or post office accounts, as suggested by some commentators.
Table 8: Cash vs Food
     Proportion    (%)    of    Respondents    Who:   
     Prefer     Prefer     Have    a    “Conditional     Have    a    “Conditional         Were    Undecided,   
     Food     Cash     Preference"         Preference"     Unclear    or       
               for    Food     for    Cash     Inconsistent
Andhra Pradesh 91.3 5.6 0.8 0 2.4
Bihar 20.8 54.2 18.1 1.4 5.6
Chhattisgarh 90.3 2.1 2.1 1.4 4.2
Himachal Pradesh 81.4 9.3 1.7 0 7.6
Jharkhand 66.0 22.2 3.5 1.4 7.0
Orissa 88.3 5.8 0.7 0 5.0
Rajasthan 59.6 14.7 7.4 8.1 10.2
Tamil Nadu 70.6 10.5 8.4 2.1 8.4
Uttar Pradesh 41.5 34.1 6.7 0 17.7
All states 67.2 17.9 5.6 1.6 7.5
See text for details.

0
.2
.4
.6
.8
1
0 20 40 60 80 100
Figure 1: Performance of the PDS and Views on Food vs Cash
Purchase-entitlement ratio (proportion of full entitlement bought in the past three months),
village average.
 0 20 40 60 80 100
1
.8
.6
.4
.2
0
P
ro
p
o
r
t
io
n
 o
f
 r
e
sp
o
n
d
en
ts
 w
h
o
 p
re
fe
r
 fo
o
d
, v
i
l
la
g
e
 a
ve
ra
g
e
SPECIAL ARTICLE
Economic & Political Weekly EPW november 5, 2011 vol xlvi nos 44 & 45 45
For instance, readiness is also a question of accessibility, conven-
ience and efficiency of the banking or post office systems.
6.3 Reasons for Preferring Food
Respondents had a range of reasons for preferring the PDS to cash
transfers: food security, poor access to banks and post offices,
unimpressive record of other cash transfer programmes, under-
developed rural markets, apprehensions regarding possible
 misuse of cash, and familiarity with the existing system, among
 others. Some of these are elaborated below.
Food security was a major concern among sample households.
The sense of security that poor households derive from getting an
assured quota of grain every month through the PDS was palpa-
ble among many of those who expressed a preference for food.35
This was especially true for single women (often widows), the
elderly and also poor households.
Respondents expressed a range of reasons for being apprehen-
sive about dealing with banks and post offices. The first among
these is the distant location of banks and post offices, exacer-
bated by the lack of public transport facilities in most of the sur-
vey areas (Himachal Pradesh and Tamil Nadu are exceptions to
this). Second, households were also concerned with overcrowd-
ing at banks and post offices. This is because of the poor density
of branches in rural areas as well as poorly staffed offices. Third,
in some parts many respondents were scarred by their bitter ex-
perience (delays, overcrowding, repeated trips, etc) with banks
and post offices in the context of NREGA wage payments. Since
2008, NREGA wages have been paid through bank and post office
accounts and that transition has resulted in delays of up to one
year (Khera 2011d). Apart from delays, in some areas respond-
ents had encountered corrupt practices in the payment of NREGA
wages even through banks and especially post offices. Fourth,
respondents also said that the PDS provided greater flexibility in
the sense that any household member, relative or neighbour
could be sent to purchase the household’s monthly quota. They
were worried that this would not be possible if a transition to
bank payments were to be made. Fifth, high rates of illiteracy
among respondents makes them uncertain of their ability to deal
with banking procedures.
Among respondents who had some experience with other
forms of cash transfers (e g, social security pensions, Janani
Suraksha Yojana or Indira Awas Yojana), many voiced another
range of concerns, such as irregular payments and corruption
(especially in the case of the Indira Awas Yojana) in the form of
“cuts” being demanded by local officials.
 Another set of concerns related to under-developed rural mar-
kets. Primary among these was the distance to the market.
Households felt that combined with poor access to banks and
post offices, this would increase the transaction costs substan-
tially especially in comparison with the current PDS system. (As
Table 6 shows, on average, the bank/post office is 3.8 km further
away than the FPS.) Second, not only are markets far, but in some
parts (especially Chhattisgarh and Orissa) respondents were con-
cerned about where they would buy their rice, as it seems rice is
not easily available in the markets of those regions and not
throughout the year. Some people were also anxious that making
frequent retail purchases would raise their food budget. A third
factor brought up often – directly or indirectly – was the lack of
trust in local markets and traders. Respondents said that if the
PDS shop closes down, there is every likelihood of local private
traders taking advantage of the lack of that fallback option for
poor households and raising prices. Even when respondents did
not articulate this distrust of local traders directly, they
 expressed apprehensions regarding being entirely reliant on
private  traders. The current situation, where they are only
partly dependent on private traders (with the rest of their needs
being met out of home-produce and the PDS) seemed a source of
comfort for them.
Note that these concerns were expressed in spite of the pres-
ence of more than one kirana shop in the village (on average each
village had 3.6 kirana stores). The reason why the kirana shop
does not count as a real alternative to the PDS or the local market
is that they do not necessarily stock grains and pulses and house-
holds seem to resort to purchases from them in emergency situa-
tions (e g, when there are unexpected guests). As a result, the
correct market alternative available to rural households seem to
be bulk grocery stores generally located in the nearest market.
Those who preferred food were 2.8 km from the bulk grocery
store and 5.6 km from the nearest market. Those who preferred
cash were closer to the bulk grocery store (1.9 km) and the near-
est market (3.7 km). The food-walas were 2.5 km from the post
office, 5.7 km from the bank, and nearly 1 km from an all-weather
road. The cash-walas were marginally closer to these facilities.
Apart from this, some respondents appeared to be “risk averse”
in the sense that they were happier with the current situation –
with its various faults discussed earlier – than with transitioning
into a system which they only partially understood. While the
payment of NREGA wages through banks and post offices has
given them an opportunity to understand the banking system,
that experience has not always been positive.
6.4 Reasons for Preferring Cash
Corruption in the PDS and dissatisfaction with the quality of grain
were important reasons for preferring cash. While overall only
13% complained of corruption in the PDS, among those who
 preferred cash this proportion was much higher (47%). However,
these proportions were very high in Bihar (54%) and UP (28%).
Similarly, where respondents were unhappy with the quality of
PDS grain, they were more likely to prefer cash: e g, compared
with an overall average of 24% complaining about the quality of
PDS grain, this proportion rises to 48% in Bihar, 39% in Tamil Nadu
and 35% in Rajasthan. The only state where respondents preferred
cash on account of irregularity of the FPSs was Bihar (50% of
 respondents reported that as a reason for preferring cash).
Some respondents also valued the fact that cash could be spent
however they liked, including on diversification of diets and pur-
chasing better-quality grain.
Based on a pilot survey, we expected those in favour of cash to
generally be the better off households. This applies to some
 extent, but in many states the pattern was not particularly pro-
nounced, perhaps because the survey was confined to BPL (or
Antyodaya) households. However, in Tamil Nadu, where the PDS
SPECIAL ARTICLE
november 5, 2011 vol xlvi nos 44 & 45 EPW Economic & Political Weekly46
is universal, this pattern was very clear. For instance, average per
capita expenditure was almost twice as high among households
preferring cash than among households preferring food. But
 interestingly, even the better-off households in Tamil Nadu often
mentioned that while they preferred cash for themselves, the
 situation of poorer people, who might prefer food, should also be
taken into account.
7 PDS Reforms
There is evidence of improvement in the PDS in most sample
states. Some of the measures undertaken by state governments
are discussed here (see also Drèze and Khera 2010b; Singh Sawh-
ney 2011 for more on PDS reforms at the state level).
7.1 Management Practices in the PDS
This section discusses management practices in the PDS, covering
the viability of outlets, changes in systems and the complaints 
of dealers.
7.1.1 Viability of Outlets
In the current system, the financial viability of FPSs depends on the
volume of grain they handle and official commissions (Rs/quintal).
In 1997, the number of ration cards handled by each FPS dealer
shrank with the introduction of the targeted PDS. Low commis-
sions combined with fewer cards made most FPSs financially
unviable, a strong incentive (if not compulsion) to cheat.
In the past three years, barring Bihar, Jharkhand and Uttar
Pradesh, official commissions that FPS managers earn from the
sale of PDS commodities have been revised (increased). This has
contributed to improving the viability of the FPS. Simultaneously,
the number of BPL cards per FPS has also increased in Chhattis-
garh, Orissa and Rajasthan, as the coverage of the PDS was
 expanded. These two measures (higher commissions and more
 ration cards), combined with earnings from the sale of empty
gunny bags (boras) after the grain has been sold, seem to  reduce
the pressure on FPS dealers to indulge in corrupt practices. They
also make it much easier to replace private dealers with collective
management of FPS (e g, by gram panchayats or self-help groups),
as many states have already done.
7.1.2 Management Practices
There have been several important changes in the policies for
management of the FPSs. This includes, in some states, the
handing over of FPSs to cooperative societies, gram panchayats,
self-help groups and other community institutions. This was 
especially visible in Chhattisgarh, Himachal Pradesh, Orissa
and Tamil Nadu. In Chhattisgarh, Himachal Pradesh and Tamil
Nadu cooperative societies dominated; Orissa has taken the lead
in handing over management of FPSs to gram panchayats,
followed by Chhattisgarh and Andhra Pradesh. In Bihar and 
Uttar Pradesh all shops were run by private dealers. Private
dealers were running most FPSs in Andhra Pradesh, Jharkhand
and Rajasthan (between 67% and 75% of FPSs were managed by
them). Andhra Pradesh is the only state in which the PDS works
reasonably well in spite of the involvement of private dealers 
on a large scale.
In cases where FPSs are run by cooperative societies, salesmen
are appointed to run the FPS. As employees, the salesman receives
a reasonable monthly salary (more than Rs 4,000 in both Chhat-
tisgarh and Tamil Nadu). This is an important step towards
 reducing corrupt practices at the FPS.
Another major change in policy has been the implementation
of “door-step delivery” of grain to FPSs. Much of the diversion of
PDS grain is known to happen between the lifting of grain from
FCI godowns and the FPSs. When PDS dealers are responsible
for this step, the chances of diversion are high. In many sample
states, the food department has started making transport
 arrangements to deliver PDS commodities to the FPS. Half of the
FPSs reported door-step delivery (either through state transport
or contractors). Wherever this has been implemented, the PERs
are high. A measure that has reduced the woes of FPS dealers is
that commissions and transport reimbursements are adjusted
at the time of depositing money for the grain. This reduces the
risk of delays and harassment when private dealers try to
 recover their official commissions and transport reimburse-
ments ex post.
Improvement in the physical infrastructure for FPS has also
helped smoother functioning of these shops. Nearly 40% of FPS in
the sample had their own dedicated space. Another 30% were
 either housed in gram panchayat buildings or rented spaces. The
proportion of shops located in private houses (e g, of the private
dealer) remains high – 25%. Most shops (71%) used weighing
scales with standard weights (rather than using stones, etc) and
one-third even had electronic weighing machines. Calculators
and receipt books were found at two-thirds and one-third of all
FPSs in the sample respectively. These basic facilities, especially
separate buildings and standard weights, contribute to stream-
lining the running of the ration shop.
7.1.3 PDS Dealers’ Complaints
PDS dealers have various complaints. In some states, dealers did
complain about low commissions for certain commodities. Quite
surprisingly, in many areas dealers did not complain of having to
pay bribes at the FCI godown or while submitting their records at
the block office.36
Dealers also complained of under-staffing. For instance, in
Tamil Nadu, each salesman is expected to make entries in several
registers, make a bill, handle the cash, then bring out the five to
seven commodities being purchased by each cardholder and
weigh them. In other places, e g, Chhattisgarh and Orissa where
many shops are run by gram panchayats, it becomes difficult for
one person (panchayat secretary in Orissa) to handle all BPL
cardholders over a short spell of three to four days. To deal with
the work, in most survey areas, the practice is to hire one person
(e g, a labourer) when the shop is open and pay him out of their
own salary. However, these expenses are not reimbursed.
7.2 Computerisation of Records
One of the most significant initiatives to increase transparency in
the PDS has been the computerisation of records. This has several
advantages. One, it helps to streamline the entire chain of distri-
bution (from lifting to distribution at FPSs) and adopt more
SPECIAL ARTICLE
Economic & Political Weekly EPW november 5, 2011 vol xlvi nos 44 & 45 47
 effective management practices (e g, Chhattisgarh has been able
to regularly update the list of cardholders in the state allowing
them to weed out duplicates from the system). Two, it helps to
maintain better records (e g, accurate, consistent, and tamper-
proof), even in real time in some cases. Three, the discipline of
strict record-keeping makes corrupt practices more difficult.
However, the full potential of these computerised databases
for reducing corruption and fostering transparency is yet to be
realised. For instance, in Chhattisgarh and Tamil Nadu, details of
purchases by each ration cardholder are available in a computer-
ised database. However this database has not been made public,
as the NREGA Management Information System has been.37 If
these were public access databases, it would help to bring greater
transparency to the system.
7.3 Transparency Measures
7.3.1 Role of Ration Cards
Ration cards can play an important role in checking whether PDS
commodities have reached intended beneficiaries. Ration cards
also allow cardholders to keep track of their purchases.
In most states, ration cards were really old. The worst case per-
haps was Jharkhand where almost all BPL respondents still had
ration cards issued by the Bihar government in 1997! In Orissa,
there was no space to record entries for current purchases
 because ration cards were outdated.
The teams found few instances of ration card hoarding either
by dealers or by sarpanches. Overall, 93% of sample households
were in possession of their ration cards. The two states in which
ration card hoarding was an issue are Bihar and Uttar Pradesh.
The maintenance of ration cards varied quite a lot across states.
In many survey areas, ration cards were legible and regularly up-
dated. In others (e g, Orissa and Uttar Pradesh), while they were
updated regularly, the entries were barely legible. In Bihar (42%),
Jharkhand (51%) and Uttar Pradesh (42%), teams found that
 ration cards had some incorrect and inflated entries, especially in
the case of non-grain commodities.
7.3.2 Coupons
Bihar recently introduced a coupon system in the PDS as a “last
mile” tracking device. At the beginning of each year, each BPL
household is supposed to get 12 coupons, one for each month of
the year. Each time a household purchases grain they are expected
to deposit the coupons for that month with the dealer. Each
month, the FPS dealer can only get as much grain as the number
of coupons he deposits.38
However, far from acting as a transparency measure, the
 coupon system in Bihar was often creating more confusion
 (Dhorajiwala and Gupta forthcoming). One reason for this is the
haphazard distribution and redeeming of coupons. What seems
to have happened is that many BPL cardholders do no have the
coupons, while some non-BPL households were given coupons.
BPL households without coupons do not get rations. Further, the
team found piles of undistributed coupons with mukhiyas (vil-
lage heads). They also found that dealers were hoarding coupons.
In one  village, the team found that the FPS opens once every two
months; when BPL cardholders go to buy their ration, they are
sold only one month’s ration, but are expected to deposit coupons
for two months.
In this and other ways, dealers are able to undermine the use-
fulness of coupons as an accountability mechanism. That, at any
rate, seems to be the situation in Bihar at the moment. In the
past, Rajasthan has experimented with the use of a similar cou-
pon system – both for the PDS and in the context of food-for-work
programmes – and the experience there has, apparently, been
more encouraging.
7.3.3 Transparency of BPL Lists
In many states, we noticed that the names of BPL households are
painted outside the panchayat bhawan or FPS. This serves the
dual purpose of helping people to know whether or not they are
on the BPL list, and also of identifying “ghost” and “duplicate”
cardholders.
Chhattisgarh has gone a step ahead and painted a sign on the
door frame of each rural house indicating the colour, type of
 ration card the household has, associated entitlements and the
BPL census it was based on. This exercise has helped to clean up
the BPL list, create awareness regarding entitlements and also
shame richer households that were in possession of BPL or
 Antyodaya cards.
7.3.4 SMS Alerts in Tamil Nadu
In Tamil Nadu, the food department has put in place a system
whereby any ration cardholder can send an SMS with the FPS
number to receive instant information regarding the stock of
each PDS commodity available in that outlet. The survey teams
tested this at several FPSs and found that the stock register
matched the information received through the SMS alert. This is
an impressive arrangement, which illustrates the scope for effec-
tive IT-based transparency measures using straightforward tech-
nology. In a similar vein, the Chhattisgarh government provides
a service whereby mobile phone users can register themselves to
receive an SMS alert whenever a truck with PDS grain leaves for
their village. The SMS gives details regarding the truck, quantity
of grain being transported, etc.
7.4 Grievance Redressal
In Tamil Nadu and Chhattisgarh, there is (at least partially) a
functional system of grievance redressal. This includes providing
phone numbers – or helplines – for ration cardholders to call 
in case of complaints. In Tamil Nadu, the phone numbers of
 concerned officials are painted outside each FPS. Local
 organisations in Chhattisgarh claimed that the helplines were
 effective and that complaints lodged there usually led to some
effective action.
8 Performance of States
Before concluding, this section attempts an informal ranking of
states on eight parameters: degree of inclusiveness, integrity of
the system (assessed in terms of the PER, quality of PDS grain,
physical access (distance and time),  predictability (whether days
of opening of the FPS are fixed), reliability (whether days of
SPECIAL ARTICLE
november 5, 2011 vol xlvi nos 44 & 45 EPW Economic & Political Weekly48
 opening are adequate), record-keeping (maintenance of ration
cards and of the sales and stocks registers) and institutional
 arrangements (door-step delivery of grain and whether the FPS has
its own physical space). Evaluated in this way, we place Himachal
Pradesh at the top of the chart and Bihar at the bottom (Table 9).
The Himachal Pradesh model is of particular interest because
it is based on the principle of a “universal, not uniform” PDS. The
state has a BPL list and BPL households are entitled to PDS grain at
a lower price than APL households. However, unlike other states,
HP seems to have ensured APL quotas are lifted and supplied to
APL households. Importantly, non-grain PDS commodities (pulses
and edible oils) are provided to all households at the same price.
Here the entitlements are linked to the number of members in
the household, rather than ration card type. Finally, Himachal
Pradesh has achieved all this without resorting to the practice of
“spreading thinner”.
Many would argue that Tamil Nadu, in fact, should be ranked
first (even based on the indicators chosen in Table 9).39 Indeed, it
is ranked number one on half of the indicators used. There are
two important reasons for placing Himachal Pradesh above Tamil
Nadu: one, in terms of providing a nutritious basket of goods, the
HP model is better than the TN model – there is a larger provision
for pulses and edible oil and this is linked to household size. Two,
the achievements of Himachal Pradesh have to be evaluated in
the light of its difficult terrain.
A curious contrast between Himachal Pradesh and Tamil Nadu
is that Himachal’s performance has come about without strict
checks and balances of the sort that one sees in Tamil Nadu. On
the other hand, primary among Tamil Nadu’s achievements, are
the monitoring systems that have evolved there – be it the use of
people-friendly and low cost technology, simple and cost-
effective measures for ensuring transparency and account-
ability, no other state has invested so much careful thought into
putting systems in place (Alamu 2011; Meenakshi 2011; Mary
Joseph  forthcoming).
The PDS in Andhra Pradesh is in large part inspired by the
Tamil Nadu model – former Chief Minister N T Rama Rao rea-
lised early the advantages of reducing prices for winning votes as
well as reducing corruption. Andhra Pradesh also never used BPL
lists prepared according to central government guidelines for the
purpose of the PDS. They applied simple exclusion criterion and
the state contributed to covering the additional households (over
and above the central quota).
In more recent times, Chhattisgarh has applied some impor-
tant lessons from Tamil Nadu: this includes an expansion of PDS
coverage, reduction in PDS prices, computerisation, doorstep
 delivery of grain, “deprivatisation” of ration shops (handing over
ration shops to community institutions such as the gram pancha-
yats and self-help groups) and setting up proper channels for
grievance redressal. These measures along with other PDS  reforms
put Chhattisgarh fourth among the nine PDS states included in
this survey.
Since 2008, Orissa has been emulating the Chhattisgarh model
(including universalisation of the PDS in the “KBK” region); it is
ranked sixth. Its experiment with entrusting the management of
the FPS to gram panchayat secretaries has had encouraging
 results (Aggarwal 2011).
More recently, since May 2010, Rajasthan too has implemented
some of the reform measures. It is worth noting that Rajasthan
was selected as it belonged to the category of states with a “lan-
guishing” PDS in 2007-08. However, the results of this survey
suggest that the reforms introduced in 2010 have already had an
impact on the functioning of the PDS there.
At the bottom of the rankings are Uttar Pradesh (ranked
eighth), Jharkhand and Bihar. While there are some signs of
 improvement in Uttar Pradesh too (e g, respondents reported
 receiving 77-88% of their full entitlement of grain from the PDS),
it is not yet clear whether, and to what extent, this is a broad-
based improvement.
Jharkhand, though ranked last but one, is an interesting case:
this is because since 2009, the state slashed PDS prices of rice to
Re 1/kg. Along with this, other PDS reforms have also been initi-
ated (somewhat half-heartedly), e g, instituting a timetable for
the distribution of PDS grain to improve predictability in the sys-
tem and door-step delivery. However, its record on actual imple-
mentation of these measures is at best mixed. Further, Jharkhand
had possibly the most incomplete and outdated BPL lists among
all sample states. The reason why Jharkhand does not do as badly
Table 9: Performance Indicators of PDS and State Ranking
State     Type    of    PDS     Purchase-     Proportion    of    (%)         Access    to    FPS     Predictability    and    Reliability     Institutional    Arrangementsb     Record-Keeping
     Entitlement     Respondents     Distance    (Km)     Time    Taken     Proportion        (%)    of     Proportion    (%)    of    FPS     Agree    with     Registers
          Ratio    (PER)     Who    Got    “Poor”     to    Ration     (mins)    for    the     Respondents    Who         Reporting     Running     Rationcard     Available    for
               Quality    Grain    at     Shop     Last    Round    Tripa     Said    that    the         “Door-step     from    a     Entries    (%    of     Inspection
               the    Time    of    the               Opening    Days     Delivery”     Private         Respondents)d     (%    of    FPS)
               Last    Purchase               of    the    FPS    Are         of    Grain    c     Residence
                  Fixed Adequate
Himachal Pradesh Universal, not uniform 92-100 4 1.4 114 (50) 71 76 45 0 70 80
Tamil Nadu Universal 92 17 0.6 109 (56) 95 67 100 8 85 100
Andhra Pradesh Quasi-universal 100 5 0.7 50 (83) 70 61 92 10 NA 70
Chhattisgarh Quasi-universal 95 14 2.3 233 (13) 72 39 100 0 94 75
Orissa Expanded 97-100 14 2.4 194 (24)  73 44 8 9 96 53
Rajasthan Expanded 86-100 19 1.7 122 (46) 91 39 35 28 87 100
Uttar Pradesh Targeted 77-88 11 1.6 102 (55) 56 44 0 17 53 75
Jharkhand Targeted 71 19 0.7 98 (47) 9 26 83 50 43 91
Bihar Targeted 45 31 1.4 94 (54) 7 29 0 100 25 55
a
In brackets, proportion (%) of respondents who reported taking up to one hour for their last round trip to the FPS. bInformation on institutional arrangements and record-keeping is taken from
the village questionnaire and is based on the visit to the FPS and interview with the FPS manager. cIncluding direct (by government) or contracted door-step delivery. dExcluding households where
responses were "unclear" or missing.
SPECIAL ARTICLE
Economic & Political Weekly EPW november 5, 2011 vol xlvi nos 44 & 45 49
as Bihar in spite of the lack of a proper system can be attributed to
the reduction of the price of rice to Re 1/kg. The PDS survey and
secondary data suggest that at least on one parameter (integrity
of the system) there has been a substantial improvement in
Jharkhand (Drèze and Khera 2011).
This leaves Bihar at the bottom of the pile, way down. It seems
that in spite of the rhetoric of development and good governance,
things have not changed all that much there at least insofar as
the PDS and NREGA are concerned.40 Bihar is at the bottom of the
chart on seven out of eight indicators used in Table 9. Even here, though, it appears that things have improved in recent times.
 

9 Discussion and Concluding Remarks
 

An important lesson from the PDS Survey 2011 is the need for a careful and nuanced evaluation of the PDS. Coverage, entitlements and implementation of the PDS at the state level vary so much that it now makes little sense to evaluate it only at the  national level. From a universal PDS that supplies grain, dals and edible oil in Tamil Nadu with hardly any leakages to Bihar’s tar- geted PDS where much of the grain does not even reach the rural
poor, in each survey state the PDS has unique features.
 

The most encouraging finding from the PDS survey is that there is a broad-based trend of improvement in the sample states, linked with recent initiatives to make the system work. Some of these measures (e g, computerisation of records, “de-privatisation” of FPS management, rationalising official commissions and so on) have been outlined here. An important part of the revival, without which perhaps the reform measures mentioned above would not have been undertaken, is a new political will to make the PDS work. In spite of these positive developments, there  remains much scope for improving the PDS, especially in terms of better quality of grain, monitoring and grievance redressal systems.
 

Meanwhile, the PDS is already considered an important lifeline by many rural households. As we saw, for BPL households in many states, the implicit subsidy from PDS foodgrains alone is roughly equivalent to a week’s NREGA wages (without having to work) every month. Combined with NREGA employment (another two days per month, on average, among sample households), the PDS has become one important foundation for India’s much needed
social security system.
 

An important caveat is that the PDS survey focused mainly on BPL (and Antyodaya) households. It is possible that serious cor- ruption persists in the APL quota; indeed there is some tentative evidence to that effect from the 66th round of the NSS (on this see Drèze and Khera 2011). One possible  reason why the PDS works better for BPL households than APL households is the lack of clear entitlements for the latter and arbitrariness of the APL quota.
 

The PDS performs a very useful role in ensuring food security for the rural poor in another important sense: it ensures a regular supply of foodgrains even in the remotest parts of the country. As things stand, rural markets are under-developed and private markets seem to fail at the last mile in many areas. It is not easy to predict what would happen in the event of a dismantling of the PDS. This “stabilisation” role of the PDS is not adequately  recognised today.
 

People’s aversion to cash transfers must be read in the light of this observation. A hasty dismantling of the PDS in favour of cash transfers would expose rural households to considerable uncer tainty and possibly exploitative transactions.
 

While this and related obstacles (such as the poor reach of the banking system) can be removed, this is a big challenge, and there is no “quick fix”. For instance, one may suggest the use of business correspondents (“the BC model”) to improve access to banks. In this context, Andhra Pradesh’s experiment with the BC model for NREGA wage payments in some areas has been sober- ing. Four years after the model was introduced with a team committed to making it work, there are still major issues related to accountability, corruption and delays. The BC model comes with its own set of problems.
 

Similarly, some commentators suggest that the problem of maintaining the purchasing power of cash in times of inflation has an easy fix – viz, to link cash transfers to an appropriate price
index. However, this deceptively simple solution raises difficult questions such as whether local prices can be tracked with ade- quate precision and speed. Apart from the “technical” issues, there is also a political aspect to indexation. The recent freeze on NREGA wages (in money terms), lasting two years at a time of double-digit food inflation, shows that political dimensions also need to be factored in.
 

 In the end, it is worth recalling that the PDS infrastructure ex- ists and in many ways performs a useful role. On the other hand, currently the infrastructure for cash transfers seems sorely lack- ing in rural areas. For those respondents who were dissatisfied with the functioning of the PDS, the correct question to ask seemed to be how to improve the PDS and reduce leakages fur- ther. In their view, the PDS, though a leaky ship, was nowhere close to sinking, and in fact, it helps them stay afloat. And there- fore, to them, it seemed premature to jump into untested waters.
 

Notes
 

 1 There are two nationally representative surveys which asked about the PDS (NSS and the India Human Development Survey conducted by the national Council of Applied Economic Research), but neither goes much beyond possession of  ration cards and PDS purchase.
 

 2 As a follow-up activity, Wayanad and Palakkad districts of Kerala were also surveyed in July 2011. Those findings are not discussed here.
 

3 This classification of states is based on per capita monthly purchase of PDS grain. “Functioning” states are those where per capita purchase re- mained above 1 kg/month in the post-1999-2000 period; “Languishing” where it remained below 1 kg/month and “Reviving” states are those where per capita purchase was below 1 kg/month at the beginning of the period, but crossed that thresh- old at the end of the period (2007-08). Note that
purchases were averaged over all households, not just BPL households.
 

 4 The sample Districts are Chittoor and Warangal (Andhra Pradesh), Katihar and Nalanda (Bihar), Mahasamund and Surguja (Chhattisgarh), Mandi and Sirmaur (Himachal Pradesh), Dumka and  undivided Ranchi (Jharkhand), Nuapada and Sundergarh (Orissa), Bharatpur and Pali (Rajas- than), Hardoi and Jaunpur (Uttar Pradesh) and Dharmapuri and Dindigul (Tamil Nadu).
 

 5 The proportion of the sample district’s rural pop- ulation living in villages in the 500-1,500 popula- tion range varies from 9-17% Andhra Pradesh, 19-21% Bihar, 52-63% Chhattisgarh, 38-44% Himachal Pradesh, 42-45% Jharkhand, 49-60% Orissa, 29-35% Rajasthan, 5-13% Tamil Nadu and 26-36% Uttar Pradesh. The survey findings should be read in the light of these figures.
 

 6 In Dumka (Jharkhand), we were unable to find the BPL list; he voter list was used.
 

 7 The terms “FPS Dealer” and “FPS Manager” are used interchangeably in this article for different management types (self-help groups, private dealerships, cooperative societies, etc). Both  survey instruments are available online at http://web.iitd.ac.in/~reetika/.
 

 8 As we sampled “Households” rather than individ- uals, investigators were requested to maintain a rough balance between male and female  respondents. SPECIAL ARTICLE
november 5, 2011 vol xlvi nos 44 & 45 EPW Economic & Political Weekly50
This is important in its own right, and especially in the context of their views on cash and food (see Section 6).
 

 9 In the NFHS, the Standard of Living Index is the sum of scores assigned to the ownership of land, television sets, fans, pressure cooker, four- and two- wheelers, type of house, toilet, energy and water source available to the household (see  International Institute for Population Sciences 2000). A modified version of SLI is used here (see Table 1 for details).
 

10  Annapoorna cards (entitled to 10 kg of free grain each month) are for the aged who do not get a pension. There were three Annapoorna card- holders in the sample. These have been excluded from the discussion.
 

11  It is not clear whether there have been further  revisions to take into account the increase in pop ulation since then.
 

12  This does not preclude the possibility that these states have a BPL list which is used to identify households for other government schemes such as the Indira Awas Yojana and Rashtriya Swasthya Bima Yojana.
 

13  In Tamil Nadu, one can choose either a “Rice Card” or a “Sugar Card”. The former entitles households to rice and other commodities; those with the “Sugar Card” forego their rice quota for an addi tional quantity of sugar. See http://www.tncsc.tn.gov.in/html/pds.htm for further details.
 

14  The remaining 3% were either APL or Annapoorna cardholders.
 

15  This was done by abolishing any differentiation of entitlements between APL and BPL households. An important caveat though is that many house- holds in the KBK region do not have any (APL or BPL) ration card. According to some estimates nearly one-fifth of all households do not have any ration card, so the universalisation of the PDS in the KBK region remains incomplete.
 

16  Interestingly, the survey team in Chhattisgarh found that households brought under the PDS as a result of the state’s decision to expand the PDS were much the same as previously identified “BPL” households (see Puri forthcoming).
 

17  The expansion of PDS coverage in Jharkhand has come in phases, with one expansion during the drought of 2008-09 and another in 2010. It was not easy to trace when these expansions were  ordered, what the entitlements of the new card- holders are or the criteria that were used for the selection of new cardholders.
 

18  An order dated 3 August 2004 was issued by the Ministry of Food and Consumer Affairs to implement the second expansion of the Antyodaya list.  According to an interim order dated 2 May 2003 in the right to food case, widows are supposed to receive Antyodaya cards.
 

19  For instance, in a 2002 survey of 400 households in Rajasthan, I found that BPL households were buying only 12.6 kg/month, 36% of their 35 kg quota. In the PDS survey 2011, the correspond- ing figure for the Rajasthan sample is 85-91%.
 

20 As some states provide an additional 5 kg/house- hold per month after a Supreme Court  order, the entitlements are expressed as a range (27.3 to 28.7 kg/household per month).
 

21  Another way of calculating a lower bound is to look at the lowest purchase from the PDS in the three months preceding the survey (i e, March- May), as a proportion of the full entitlement. Doing so for BPL households, over all states, house holds get 82% of their full quota.
 

22  Whether we use the “Normal” purchase criterion, or the “Last Purchase” criterion, the same conclusion applies.
 

23  In 2004-05, 90% of PDS grain in Bihar did not reach ration cardholders (Khera 2011c)  according to NSS data; in our PDS survey, the  average PER for Bihar is 45-47% (Table 3).
 

24  More recent data pertaining to 2009-10 from the 66th round of the NSS also confirm that revival among most sample states has been consolidated between 2004-05 and 2009-10 (Drèze and Khera 2011).
 

25  The “Rice States” are Andhra Pradesh, Chhattisgarh, Orissa and Tamil Nadu, where wheat is   either not distributed at all or forms a very small share of total PDS grain entitlements.
 

26 In the case of Rajasthan, data on awareness was recorded only for about half of the households. Among these, awareness levels are high. The rem- aining households include missing observations and those who were unaware of their entitlements.
 

27  Some households (nearly 14% of the sample) were unable to answer this question as they had been in- cluded in the PDS in the past one or two years only.
 

28 The figures on hunger reported here are not comparable to those reported by the NSS. The latter reports whether everyone in the household gets two square meals throughout the year whereas in this survey, we asked if any person in the household had ever had to sleep hungry/skip a meal in the past three months. During the debriefing workshop, investi-
gators felt that there were variations in people’s willingness to talk about hunger in the house-
hold, and the data presented here should be read in that light.
 

29 If one excludes outliers, it comes down to 14.5 kg/ capita/month.
 

30 In the case of Rajasthan, it is possible that the questionnaire did not capture some food items that might be ordinarily consumed by households (e g, buttermilk).
 

31  This pattern can be explained partly by vegetarian diets for significant sections of the population.
 

32  Statements to this effect were often made in the consultations on the National Food Security Act.
 

33  Note that this amount exceeds the cash equivalent of PDS entitlements. The latter is equal to the market value of PDS commodities, minus what people pay for them at the FPS. This is still quite substantial (e g, Rs 4.65 per kg for wheat) in states that charge the “Central Issue Price”.
 

34  Excluding the Bihar sample, the corresponding proportions are 73% (food) and 13% (cash) respectively.
 

35  See Jose (2011), Nag (2011) and Puri (forthcoming).
 

36 In the past, PDS dealers have given us detailed  accounts of the corrupt practices they indulge in to break even when official commissions are coupled with low volumes and pressures to pay bribes (see, e g, Khera 2011b).
 

37  The official NREGA website (www.nrega.nic.in) has a publicly accessible database which contains employment information for all job cardholders (including days of employment, wages earned, worksite employed on and so on). This makes it possible for people to verify their own records and that of others, apart from facilitating social audits.
 

38 This is not the same as the “Food Stamps” advocated by others which can be redeemed at any shop, in the private market or an FPS (e g, Basu 2011).
 

39 See Meenakshi (2011) and Alamu (2011).
 

40 In Table 1, one finds that Bihar 62% of sample households reported getting no work on NREGA in the past 12 months and the average days of NREGA employment in the same period was just 12 days. Not a single household had got 100 days of work.
 

Secondary data also suggest that Bihar is among the worst states in terms of the scale of NREGA employment (see www.nrega.nic.in and Khera 2011d).
 

References
 

Aggarwal, Ankita (2011): “The PDS in Rural Orissa: Against the Grain”? Economic & Political Weekly, Vol 46, No 36, 3 September.
 

Alamu, R (2011): “It Just Works in Tamil Nadu”, The Hindu, 25 September.
 

Bhatia, Bela and Jean Drèze (2002): “Still Starving in Jharkhand”, 16 August, Frontline.
 

Citizen’s Initiative for the Rights of Children Under Six (2006): “Focus on Children under Six (New Delhi: Right to Food Campaign)”, available online at http://www.righttofoodindia.org/icds/icds_in- dex.html. Last accessed on 15 September 2011.
 

Dev, S Mahendra and M H Suryanarayana (1991): “Is PDS Urban Biased and Pro-Rich? An Evaluation”, Economic & Political Weekly, Vol 26, No 41, 12 October.
 

Dhorajiwala, Sakina and Aashish Gupta (forthcoming): “Can Coupons Cure  Corruption?”, Frontline.
 

Drèze, Jean (2001): “No More Lifelines: Political  Economy of Hunger in Orissa”, Times of India, 17 September.
 

 – (2002a): “Ending Destitution”, The Hindu, 27 July.
 

 – (2002b): “From the Court to the Streets”, The Hindu.
 

Drèze, Jean and Aparajita Goyal (2003): “The Future of Mid-Day Meals”, Economic & Political Weekly, Vol 38, No 44, 1 November.
 

Drèze, Jean and Reetika Khera (2010a): “The BPL Census and A Possible Alternative”, Economic & Political Weekly, Vol 45, No 9, 27 February.
 

 – (2010b): “Chhattisgarh Shows the Way”, The Hindu, 13 November.
 

 – (2011): “PDS Leakages: The Plot Thickens”, 12 August, The Hindu.
 

Himanshu and Abhijit Sen (2011): “Why Not a Universal food Security Legislation?”, Economic & Political Weekly, Vol 46, No 12, 19 March.
 

Hirway, Indira (2003): “Identification of BPL House- holds for Poverty Alleviation Programmes”, Eco- nomic & Political Weekly, Vol 38, No 45, 8 November.
 

Howes, Stephen and Shikha Jha (1992): “Urban Bias in the Indian Public Distribution System”, Economic & Political Weekly, Vol 27, No 19, 9 May, pp 1022-30.
 

International Institute for Population Sciences (2000): “National Family Health Survey 1998-99 (NFHS-2): India” (Mumbai: International Institute of Population Sciences).
 

Jha, Shikha and Bharat Ramaswami (2010): “How Can Food Subsidies Work Better? Answers from India and the Phillippines”, ADB Economics Working Paper Series, No 221, September, Asian Development Bank, available online at http://www.adb.org/documents/working-papers/2010/ economics-wp221.pdf. Last accessed 20 September 2011.
Jose, Jijo (2011): “The PDS Learning Curve”, Down to Earth.
 

Khera, Reetika (2002): “Mid-Day Meals in Rajasthan”, The Hindu, 13 November. 

 – (2006): “Mid-Day Meals in Primary Schools: Achievements and Challenges”, Economic & Polit- ical Weekly, Vol 41, No 46, 18 November.
 

 – (2008): “Access to the Targeted Public Distribu- tion System: A Case Study in Rajasthan”, Econom- ic & Political Weekly, Vol 43, No 44, 1-7 November.
 

 – (2011a): “The UID Project and Social Welfare”, conomic & Political Weekly, Vol 46, No 9.
 

 – (2011b): “India’s Public Distribution System: Utili- sation and Impact”, Journal of Development Studies, Vol 47, No 3, pp 1-23.
 

 – (2011c): “Trends in Diversion of Grain from the Public Distribution System”, Economic & Political Weekly, Vol 46, No 26, 21 May.
 

– (2011d): ed. “Battle for Employment Guarantee (New Delhi: Oxford University Press).
 

Kotwal, Ashok, Milind Murugkar and Bharat Ramaswami (2011): “Don’t Water Down Food Security”, Indian Express, 19 January.
 

Mahamallik, Motilal and Gagan Bihari Sahu (2011): “Identification of the Poor: Errors of Exclusion and Inclusion”, Economic & Political Weekly,  Volume 46, No 9, 26 February.
 

Mary Joseph, Aleesha (forthcoming): “Lessons in Grass- roots Economics”, Frontline.
 

Meenakshi, Swathi (2011): “A Case for Uniform and Universal System”, The Tribune, 7 September.
 

Nag, Kuber (2011): “People’s Demand: ‘No Substitute for the PDS’”, 4 September, The Hindu.
 

Puri, Raghav (forthcoming): “Cash Transfers? No, thanks”, Frontline.
 

Sinha, Dipa (2008): “Social Audit of Mid-day Meal  Scheme in AP”, Economic & Political Weekly, Vol 43, No 44, 1 November.
 

Sawhney Singh, Ria (2011): “The PDS Is Not Ailing or Failing”, The Tribune, 7 September.
 

Suryanarayana, M H (1995): “PDS: Beyond Implicit Subsidy and Urban Bias – The Indian Experience”, Food Policy, Vol 20, No 4, pp 259-78.
 

Swaminathan, Madhura and Neeta Misra (2001):  “Errors of Targeting”, Economic & Political Weekly, Vol 36, No 26, 30 June.