Understanding the Costs of the Food Corporation of India

Author(s): Madhura Swaminathan

Source: Economic and Political Weekly, Vol. 34, No. 52 (Dec. 25-31, 1999), pp. A121-A132
Published by: Economic and Political Weekly
Stable URL: http://www.jstor.org/stable/4408768 .
Accessed: 12/09/2013 04:39

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .
http://www.jstor.org/page/info/about/policies/terms.jsp

.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact support@jstor.org.

Economic and Political Weekly is collaborating with JSTOR to digitize, preserve and extend access to
Economic and Political Weekly.

http://www.jstor.org

This content downloaded from 125.20.82.164 on Thu, 12 Sep 2013 04:39:44 AM

All use subject to JSTOR Terms and Conditions
 the
Understanding Costs of the Food
Corporation of India
Madhura Swaminathan

The debate on food subsidies in India affects millions of consumers and producers, and is of immense
importance in a country in which chronic hunger and malnutrition affects a majority of the population.
In this paper, I first examine the volume of the food subsidy over the last three decades. Secondly, using
data for the 1990s from the accounts of the Food Corporation of India (FCI), I examine the components
of the foodgrain subsidy, the relative contribution of different components to total costs, and the growth
in different components during the current decade. Thirdly, I attempt to compare the efficiency of the FCI
with that of private trade.
The data from FCI performance budgets show clearly that the increase in procurement price was a critical
factor in the increase in economic costs of rice and wheat. The procurement price, however, is a variable
over which the FCI has no control as the central government sets the procurement price based on the
recommendations of the Commission on Agricultural Costs and Prices. Despite the absolute increase in
many components of costs, there was an improvement in the operational efficiency of the FCI during the
1990s.
Lastly, no unambiguous finding emerged from a comparison of wholesale market prices and economic
costs of the FCI. The FCI compared favourably with private traders in the distribution of rice in a large
number of states. In the states of northern and central India, however, private traders had lower prices
for grain. The presence of systematic rural-urban price differences indicated the imperfections of grain
markets.

I the improved supply situation, it would an evaluation of the system of public

Introduction seem appropriateto phase out government distribution.4 As mentioned earlier, the
controls as well as current procurement term 'the food subsidy' refersto budgetary
IN 1991, with the introductionof new operations" [Radhakrishnaand Subbarao, payments to the FCI to meet is operational
programmes of structuraladjustmentand 1997:76]. In other words, it is recom- deficit.
fiscal tightening in India, government mended that the FCI "go out of the opera-
officialsandtheireconomicadvisersbegan tions of procurement and supply to the
II
to call for a reductionin subsidies, in- PDS" (ibid p 77).2 The underlyingassump-
Volume of Food Subsidy, 1966-67 to
1997-98
cluding food subsidies. They spoke in tion here is, of course, that private trade
termsof a 'rationalisation','reduction'or is more efficient than the FCI.3 Graph 1 plots the food subsidy at current
'withdrawal'of the food subsidy.1One The debate on food subsidies in India prices and constant prices (deflated by the
aspectof the discussionwas on ways of affects millions of consumers and pro- GDP deflator) from 1966-67 to 1997-98.
targetingthe public distributionsystem ducers, and is of immense importance in Graph 2 plots the nominal food subsidy
(PDS),a systemof rationingthatprovides a country in which chronic hunger and as a proportionof GDP and as a proportion
a set of basic commoditiesat subsidised malnutrition affects a majority of the of government expenditure over the last
pricesthroughfair-priceshops. Another population. I shall begin, first, by examin- 30 years. Innominalterms,thefood subsidy
aspectof thediscussionrelatesto thecosts ing the volume of the food subsidy and has been rising rapidly, particularly after
andefficiencyof theFoodCorporation of how it has varied over time. Secondly, the mid-1980s, and it took a big jump in
India (FCI), the government-owned using data for the 1990s from the accounts 1993-94. At constant prices, the increase
organisationin charge of procurement, of the FCI, I examine the components of is more subdued. Graph 1 shows that
storageand distributionof foodgrain. the foodgrain subsidy, the relative con- expenditureon the food subsidy, atconstant
Sincethecategory'foodsubsidy'in the tribution of different components to prices, rose in the mid-1980s and then
annualbudgetof the governmentof India total costs, and the growth in different remained unchanged until about 1989-90.
is nothingbut the operationaldeficit of components during the current decade. It dipped between 1990-91 and 1992-93
the FCI,it has been arguedthatthe food By doing so, we can identify the specific and rose in 1993-94. The food subsidy
subsidycan be reducedby privatisingor components that are high or rising and the fell slightly in the following years but
reducingthe role of the FCI. It is argued measuresrequiredto containthem.Thirdly, was higher than in the 1980s. However,
thatthe governmentshouldget out of the I attempt to compare the efficiency of the the food subsidy as a share of GDP ap-
task of distributingbasic commodities, FCI with that of private trade on the basis pears to have remained more or less
includingfoodgrain,andallowthemarket of price comparisons. unchanged over the last 20 years, peaking
to distributefood.A recentstudyprepared The main focus of this paper is on the at about 0.63 per cent in 1985-86 and
fortheWorldBankproposedthefollowing: costs of the FCI, the changes therein, and 0.64 per cent in 1993-94. Interestingly,
"Giventheweaknessesandconstraintsof how these compare with the costs incurred the food subsidy as a share of government
FCI,thepervasiveroleof well-developed by private tradersin grain. The paper does expenditure shows large year-to-year
foodgrainmarkets,and bearingin mind not deal directly with consumers or with fluctuations, and the share was higher in

Economic and Political Weekly December 25, 1999 A-121

This content downloaded from 125.20.82.164 on Thu, 12 Sep 2013 04:39:44 AM

All use subject to JSTOR Terms and Conditions
 1976-77(3.84 per cent) thanin 1993-94 in differentyears. In 1993-94, for example, agency for the procurement,storage, inter-
(3.56 per cent). about 86 per cent of the total food subsidy state movement or transport and distri-
There are two significantpoints here. was on account of the cereal subsidy.5 The bution of food commodities. In short, the
First,the food subsidyas a shareof GDP rest of this paper deals only with the FCIis responsiblefor implementingcentral
has not changedvery muchover the last foodgrain subsidy, that is, the subsidy for government policies on procurement,
20 years.In otherwords,the 'burden'of rice and wheat, which as is clear, is lower storage, transport and distribution. In
food subsidyhas not risen. Secondly, in than the total food subsidy.6 certainoperations such as the maintenance
proportionate terms,the food subsidybill The FCI is responsible for buffer stock of national buffer stocks, the FCI has sole
in Indiais not very high when compared operations and the total foodgrain subsidy responsibility whereas in certain other
with expendituresin other developing includes the costs associated with main- operations sach as procurement, the FCI
countries.In Sri Lanka,after the intro- taining buffer stocks (such as handling has to work with state government
s ductionof means-tested food stampsand costs, costs of storage, interest payments organisations (such as state marketing
a steep reductionin food subsidies,they and administration). The total costs of federations) and within the purview of
still accountedfor 1.3 per cent of GDP storing and distributing the foodgrain state government policies.8 Despite these
(in 1984),orroughlytwice theproportion procured by the FCI are apportioned, on constraints, the achievements of the FCI
in India [Jayawardenaet al 1988]. In the basis of certain principles, to distri- areconsiderable.Today, it is a majorplayer
Mexico, in 1984, when general food bution through the PDS and to the costs in the foodgrain trade:procuringfoodgrain
subsidieshad been eliminated,the food of maintaining buffer stocks.7 Data from from surplus regions, storing grain,
subsidy was 0.63 per cent of GDP FCIperformancebudgetsshow veryclearly transporting it over very large distances
[Pinstrup-Andersen et al 1991]. In Egypt thatthe subsidy incurredon bufferstocking across states and making it available for
in 1982, the expenditureon food rations operations rose rapidly in absolute and distribution in all parts of the country,
was about15 percent of GDP [Subbarao relative terms in the 1990s. In 1994-95, including very remoteareas.Throughthese
et al 1997]. In Tunisia, food subsidies for example, the shareof the subsidy given operations, it has contributed to several
accountedfor about4 per cent of GDP to the FCI on account of buffer stocking objectives of food policy. This is the
in 1984;theywerereduced,aftertargeting, operations was 44 per cent of the total context in which the operations and costs
toaround2 percent of GDPin 1993[Tuck foodgrain subsidy [World Bank 1996]. of the FCI need to be reviewed.
andLindert1996]. In India,over the 31- The management of buffer stocks does
OFFCI
COSTS
yearperiod1966-1997,the food subsidy benefit consumers via means of price
accounted,on average,for 0.31 per cent stabilisation. The rest of this paper, Data from the performance budgets of
of GDP and 2.35 per cent of central however, deals with the direct subsidy the FCI were made available to me and
governmentexpenditure.These numbers involved in the distribution of foodgrain they provide a break-up of the costs
areimportant(as NoraLustighas said of to consumers. incurred on rice and wheat separately for
Mexico) because they show that even
III TABLE 2: GROWTH RATE OF COMPONENTSOF
eliminatingfood subsidiestotallywould FCI: Objectives and Costs ECONOMIC
COST, 1990-91 TO 1998-99
not solve the fiscal problems of the
government. The broad objective of food policy in Wheat Rice
India has been to make available food to Procurement
priceof grain 11.51 9.80
SUBSIDY ON FOODGRAIN
the people at reasonable prices. The more Total procurementincidentals 8.86 10.92
The food subsidy as defined in the specific objectives include providing Statutorycharges 11.02 9.10
Labourand transportcharges 6.49 10.77
governmentbudget includes the entire remunerative prices to cultivators; sup- Total chargespaid to stategovts 5.67 17.86
deficitof thestate-ownedFCI. plying food at subsidised prices to the
operational Total distributioncosts 6.84 9.04
The four major items of food that are undernourished; controlling inflationary 8.52 9.57
Freight
handledby the FCIarerice (andpaddy), pressures;stabilising prices forconsumers Handlingexpenses 14.05 14.23
wheat, importededible oils and sugar. and producers; reducing fluctuations in Storagecharges 9.97 10.76
Now the total central food subsidy (as food availability; and achieving self- Interestcharges 3.03 7.60
shown in Graphs1 and 2) includes the sufficiency in foodgrain production. The Transitand storagelosses -9.45 3.85
Establishmentcharges 8.58 8.74
subsidyon sugar,andthisis likelyto vary FCI was set up in 1964 to act as a nodal
TABLE 1: ECONOMICCOSTS AND SUBSIDY ON RICE AND WHEAT, 1990-91 TO 1998-99
(Rs/Qtl)
1990-91 1991-92 1992-93 1993-94 1994-95 1995-96 1996-97 1997-98(RE) 1998-99(RE) 1990-98
Growth
(% pa)
Wheat
Economic cost 356.5 390.79 504.1 532.03 551.17 583.95 640.16 800.5 807.95 9.99
Sales Realisation 239.95 251.68 279.36 355.88 407.89 411.94 433.2 395.87 388.33 7.18
Subsidy 116.55 139.11 224.74 176.15 143.28 172.01 206.96 404.63 419.62 13.56
Total costs other than
procurementcosts 151.82 179.59 208.59 206.72 216.37 232.91 258.53 306.08 293.92 7.98
Rice
Economic cost 457.52 497.04 585.27 665.1 694.71 762.82 847.69 940.4 980.36 9.73
Sales Realisation 330.02 365.58 442.4 500.42 600.76 613.34 610.57 610.8 601.18 7.98
Subsidy 127.5 131.46 - 142.87 164.68 93.95 149.48 237.12 329.6 379.18 13.39
Total costs other than
procurementcosts 120.07 125.59 152.35 165.49 159.32 182.73 221.71 254.98 237.67 9.51
Note: RE stands for revised estimates.

A-122 Economic and Political Weekly December 25, 1999

This content downloaded from 125.20.82.164 on Thu, 12 Sep 2013 04:39:44 AM

All use subject to JSTOR Terms and Conditions
 FIGURE 1: COMPONENTSOF ECONOMICCOST Figure 1. The procurementprice is the
pricepaidby the FCIto producersfor the
purchaseof grain.Procurementcosts are
Economic Cost the initial costs incurred during the
procurement of grainsat 'mandis'(market
yards) or other procurement centres.
Procurementcosts or incidentals are
reportedunderthefollowingsubheadings:
statutorycharges,labourcharges,amount
paid to state agencies for establishment,
storageand interestfor stocks procured,
portclearancecosts(inthecaseof imports)
andothercosts. Distributioncosts arethe
costs involvedin storingandtransporting
Procurement Procurement Distribution grainto the final distributionpoints.The
Costs Price Costs components of distribution costs are
freight,handlingexpenses,storagecharges,
interestcharges,transitshortages(orlosses
duringtransit),storageshortages(losses
duringstorage)and establishmentcosts.
Furtheranalysis in this section is based
entirelyon currentor nominalprices,as
Fe Handling Storage Interest Losses (during Establishment we areprimarilyinterestedin the relative
charges charges charges transport& storage) charges increase in different components of
economiccost.ll Note also thatsome of
the observationsmade below are hypo-
thesesbasedon a preliminaryanalysisof
costs that require furtherresearchand
substantiation.
The first featureof the basic data on
Incidentaland Labourand Chargespaidto
economiccosts, salesrealisationandsub-
Statutory Transport StateGovt. for sidyis thegrowinggapbetweeneconomic
costs andsales realisation,particularlyin
charges charges storageandinterest
thelasttwo years(Table1).First,the eco-
nomiccosts of procuringanddistributing
thenineyears1990-91to 1998-99.9Inthis Subsidy = Economic Cost - Sales bothrice andwheathaverisensteadilyin
section, the cost data are analysed to Realisation. the1990s.Theeconomiccostsof procuring
identifythe relativecontributionof dif- The economiccost, in turn,is defined anddistributing wheatrosefromRs356.50
ferentcomponentsto totalcosts andcom- as the sumof procurementprice,procure- per quintalin 1990-91 to Rs 807.95 in
ponentwisegrowth of costs during the ment-relatedcosts anddistributioncosts, 1998-99, reflectingan annualgrowthof
1990s.10In the FCI budget,all costs are as follows: 10 per cent (Table 1).12The economic
reportedin perquintaltermsfor rice and Subsidy = (Procurement Price + costs of procuringand distributingrice
wheatseparately.The subsidyis defined ProcurementCosts + DistributionCosts) grew at 9.7 percent a yearover the same
as the.difference
betweeneconomiccosts -Sales Realisation period, from Rs 457.52 to Rs 980.36.
andthepriceobtainedfromsales (orsales A diagrammaticsketch of the com- Secondly,the averagesales realisation
realisation),that is, ponents of economic cost is shown in (ortheaveragepriceobtainedby theFCI)
TABLE 3A: SHARE OF VARIOUS INCIDENTALCOSTS IN TOTAL INCIDENTALCOSTS FOR WHEAT

1990-91 1991-92 1992-93 1993-94 1994-95 1995-96 1996-97 1997-98 (RE) 1998-99 (RE) 1990-98

Statutorycharges 56.81 47.16 49.53 56.91 57.72 58.97 59.03 61.16 58.65 56.22
Labourand transportcharges 13.16 13.47 16.37 15.48 12.51 12.43 13.20 11.87 11.92 13.38
Storage and interest charges to state govt 20.84 20.20 25.49 20.95 21.18 20.71 19.85 18.50 22.23 21.11
Additonalcharges to state govt 9.19 19.16 7.82 6.41 6.55 7.18 7.56 7.06 5.99 8.55
Total charges on state government 30.04 39.36 33.31 27.35 27.72 27.89 27.41 25.55 28.22 29.65
Total procurementincidentals 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00

TABLE
3B: SHARE
OFVARIOUS
INCIDENTAL
COSTS
INTOTAL
INCIDENTAL
COSTS
OFRICE
1990-91 1991-92 1992-93 1993-94 1994-95 1995-96 1996-97 1997-98(RE) 1998-99 (RE) 1990-98

Statutorycharges 71.96 68.37 67.30 70.53 57.29 67.23 58.59 63.68 62.73 65.30
Labourand transportcharges 15.85 16.37 17.79 16.45 14.30 14.64 13.61 16.76 17.94 15.97
Storage and interestcharges to state govt 4.38 4.77 7.36 6.41 6.64 8.00 6.71 7.22 7.21 6.52
Additionalcharges to state govt 7.81 10.49 7.55 6.61 21.78 10.13 21.09 12.35 12.12 12.21
Total charges to state govt 12.19 15.26 14.91 13.02 28.42 18.13 27.80 19.56 19.33 18.74
Total procurementincidentals 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00

Economic and Political Weekly December 25, 1999 A-123

This content downloaded from 125.20.82.164 on Thu, 12 Sep 2013 04:39:44 AM

All use subject to JSTOR Terms and Conditions
 GRAPH 1: EXPENDITUREON FOOD SUBSIDY, CURRENT AND 1980-81 PRICES, RS CRORE

10000
9000
8000
7000
6000
5000
4000 /
3000 t O ' ? nO ? *t
2000

' ,
,6 64 e~ , ,
O O 00 00 00 00 00 0\ 0 0
%O
so oo
O0 ro0 C( b oo oo oo oo oo oC ON o
1, N N N N 00 00 00 00 00
0oo 0 0
0% 0% 0 0 0 %\ OO
x0 0\ O0 \ ON
0% O
ON Ch

- - Food subsidy currentprices - - Food subsidy 1980-81 prices

GRAPH 2: FOOD SUBSIDY AS A PROPORTIONOF GDP AND TOTAL GOVERNMENTEXPENDITURE

5.5
5
4.5

3.5
3
2.5
2
1.5

0.5

e en
Nr'-
r- 0% - N O% - N -^ N
\O O r-N N - 00 00 00 00 0t% 00 %
_0 Cs

O\ r- N- rN N 00 00 00 00 00
oo o Os ON O
aO ON as 0 ON as 0o Oa N lO O ON ON O0 ON ON

--- Food subsidy/GDP -U- Food subsidy/Total expenditure

for rice and wheat did not rise as fast as by a totalof 185percentforcommonrice to have led to the observeddeclinein the
their respective economic costs. Sales and 172 per cent for wheat.Priceswere averagepriceobtainedon salesbytheFCI.
realisationfrom wheat grew at 7.2 per cent then held constanttill 1997. In 1997, a Nevertheless,the sales realisationfrom
a year and that from rice at 8 per cent a dualpriceregimewas introducedas part rice grew slightlyfasterthanfromwheat.
year. Average sales realisation from wheat of the TargetedPDS. In this scheme,the As a consequence,the unit subsidy on
fell in nominal terms from 1996-97 populationis dividedinto 'abovepoverty wheat grew more rapidlythan the unit
onwards, and in the most recent year, line'(APL)and'belowpovertyline'(BPL) subsidyon rice andexceededthe subsidy
1998-99, for rice. After 1996-97, the sales groupsandpricesdifferforthetwogroups. on rice in absolutetermsin most years.
realisation from wheat fell even below the Priceswereraisedfor the APLgroupand In 1990-91, the subsidypaid to the FCI
procurementprice. In the case of rice too, loweredfortheBPLgroup.As partof the was Rs 116.55 per quintalof wheatand
the average sales realisation was lower TargetedPDS, rationentitlementswere Rs 127.5 per quintalof rice. In 1998-99,
than the procurementprice after 1995-96. reduced,particularlyfor purchasesat the however,the unitsubsidywas Rs 419.62
If we exclude the last two years of data, lower BPL prices. In recentyears, there for wheat and Rs 379.18 for rice. From
sales realisation is observed to grow much has been a sharpdeclinein the quantities 1991-92onwards,withtheexceptionof one
faster, at around 11 per cent annually for sold throughthePDSanda corresponding year,the ratioof the wheatsubsidyto the
both rice and wheat. rise in stocks.13The fall in averagesale rice subsidy has been greaterthanone.
In addition to sales to the PDS and other priceis partlydue to the fall in quantities In short,a growingdivergencebetween
social security programmes, the FCI also sold. Also, from 1997,due to restrictions economic costs and sales realisationled
engages in open market sales and exports. on openmarketsales,theFCIhasnotbeen to a majorrise in unit subsidyduringthe
There may be several explanations for the ableto selloff surplusstocksin themarket. 1990s. The divergencehas been pheno-
recent decline in sales realisation. Central Damage to stocks or purchaseof poor menalin the last two years.If we exclude
issue prices were raised steeply and on quality grain could also lower the sale datafor 1997 and 1998, the growthrates
several occasions between 1991 and 1994, price.All thesefactorstogetherarelikely of subsidyareloweredsignificantly(halved

A-124 Economic and Political Weekly December 25, 1999

This content downloaded from 125.20.82.164 on Thu, 12 Sep 2013 04:39:44 AM

All use subject to JSTOR Terms and Conditions
for bothrice andwheatfrom 13 per cent wheat [BICP 1991]. A detailed discussion defined as a percentage of the procurement
to 6 per cent). of each component of procurement price, they have, not surprisingly, grown
incidentals and distribution costs can be in proportion to the procurement price.
CHANGES IN COMPONENTSOF ECONOMIC
found in the Appendix.16 Some of the For wheat, labour and transport costs
COSTS salient features are discussed here. involved during and after procurement
In termsof broadcategories,one of the rose in absolute terms but at the relatively
COSTS
PROCUREMENT
fastest growing components of the
economiccost of wheatwas the procure- Of total economic costs, procurement TABLE 5: RATIO OF ECONOMICCOST TO
mentprice,whichgrewat 11.51per cent incidentals were around 15 per cent for PROCUREMENTPRICES, RICE AND WHEAT,
1980-81 TO1998-99
annuallyover this decade (Table 2 and wheat and 7 per cent for rice. Procurement
Graph3).14Procurement costsgrewat 8.9 costs accounted for a bigger share of total Year Wheat Rice
percenta yearanddistribution costs grew costs for wheat; they have also risen more
1980-81 143.0 137.2
at6.8 percenta year.As a result,theshare rapidly for wheat. An importantreason for 1985-86 154.0 143.2
of procurementprice in total economic lower procurement costs for rice may be 1988-89 171.0 139.4
cost of wheatincreasedfrom57 per cent that it is obtained directly from millers. 1990-91 174.0 135.2
in 1990-91to 64 percent in 1998-99.The Changesin the componentsof procurement 1991-92 185.0 133.4
shareof distributioncostsintotaleconomic incidentals are plotted in Graphs 5 and 6 1992-93 170.5 134.9
163.5 132.9
costs fell correspondinglyfrom 24.5 per for wheat and rice respectively (see, also, 1993-94
1994-95 164.6 129.1
centto 19.2 per cent. In otherwords,the Table 2). 1995-96 166.3 131.2
rapidrise in procurementprices was the Statutory charges were the most 1996-97 163.8 134.9
most importantfactorin the rise of total importantcomponent of total procurement 1997-98 161.0 136.9
economiccosts of wheat.15 incidentals for rice and wheat. For wheat, 1998-99 156.2 131.7
In the case of rice too, procurement statutory charges accounted for 8.6 per Notes: Figures for the 1980s are from Tyagi
pricesrose over this period,at an annual cent of the economic cost in 1990-91, and (1990).
rateof 9.8 percent.Graph4 showsclearly grew at a rate of 11 per cent a year. Of TABLE 6: RATIO OF UNIT SUBSIDY TO
that the procurementprice of rice has total procurement incidentals, statutory PROCUREMENTPRICE, RICE AND WHEAT,
increased steadily.Procurement incidentals charges were the most important 1980-81 TO1998-99
grew slightly faster at 10.9 per cent component (Table 3A). Further, these
Year Wheat Rice
annually,anddistributioncosts rose at 9 charges amounted to one-fifth of total
percentayear.Inthecaseof rice,procure- costs of the FCI after excluding the pro- 1980-81 34.3 21.6
mentpriceshave alwaysaccountedfor a curement price of grain. For rice, statutory 1985-86 44.3 36.3
substantialshareof totaleconomiccosts charges were lower, accounting for around 1988-89 45.2 34.0
andtheshareincreasedfurtherinthe 1990s. 4 per cent of economic cost. Statutory 1990-91 56.9 37.6
1991-92 65.8 35.2
The shareof procurementprices in total charges grew at a rateof 9 percent annually 1992-93 76.0 32.9
economiccostsof ricewas73.9 percentin between 1990-91 and 1998-99. Statutory 1993-94 54.1 32.9
1990-91,77.4percentin 1994-95and75.9 charges, however, were the largest 1994-95 42.7 17.4
per cent in 1998-99.While procurement component of procurement incidentals 1995-96 49.0 25.7
incidentals haveincreasedsteadilyforboth (e g, as high as 7 1.9 percent of procurement 1996-97 52.9 37.7
wheatand rice, the patternof change is incidentals in 1990-91, Table 3B) and 1997-98 81.4 47.9
81.1 50.9
quitedifferentfromthatduringthe 1980s accounted for about 17 per cent of costs 1998-99
whenprocurement-related costsrosefaster other than the procurement price of rice Notes: Figures for the 1980s are from Tyagi
thanprocurement pricesfor bothrice and in 1998-99. Since statutory charges are (1990).
TABLE 4A: SHARE OF VARIOUS COSTS IN TOTAL DISTRIBUTIONCOSTS OF WHEAT

1990-91 1991-92 1992-93 1993-94 1994-95 1995-96 1996-97 1997-98 (RE) 1998-99 (RE) 1990-98

Frieght 41.24 31.94 26.86 42.76 44.95 39.26 38.30 37.93 39.92 38.13
Handlingexpenses 9.57 10.40 9.31 9.11 11.03 12.58 15.01 15.66 15.09 11.97
Storagecharges 7.84 8.50 10.44 6.91 7.53 9.74 9.94 9.96 10.47 9.04
Interestcharges 23.88 31.88 33.29 28.15 23.91 23.87 21.71 23.63 21.81 25.79
Transitshortages 7.04 7.29 6.61 4.28 4.20 4.11 4.91 3.06 2.70 4.91
Storage shortages 1.22 1.94 1.85 -0.54 -0.35 -0.29 -0.31 -0.35 -0.26 0.32
Transitand storage shortages 8.26 9.23 8.46 3.75 3.85 3.82 4.60 2.70 2.43 5.23
Establishmentcharges 9.20 8.05 11.65 9.32 8.73 10.72 10.44 10.11 10.28 9.83
Total distributioncosts 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00

TABLE4B: SHAREOFVARIOUS
COSTSINTOTALDISTRIBUTION
COSTSOFRICE

1990-91 1991-92 1992-93 1993-94 1994-95 1995-96 1996-971997-98 (RE) 1998-99 (RE) 1990-98

Freight 41.24 31.94 26.86 33.60 29.25 33.49 35.26 35.20 36.24 33.68
Handlingexpenses 9.57 10.40 9.31 8.59 11.78 11.13 11.51 14.51 13.67 11.16
Storage charges 7.84 8.50 10.44 6.48 8.04 8.61 9.13 9.29 9.46 8.64
Interestcharges 23.88 31.88 33.29 34.52 34.18 28.18 26.66 26.33 26.08 29.44
Transitshortages 7.04 7.29 6.61 5.42 4.99 5.78 5.81 3.27 3.32 5.50
Storage shortages 1.22 1.94 1.85 2.60 2.45 3.33 3.63 2.01 1.97 2.33
Transitand storage shortages 8.26 9.23 8.46 8.01 7.44 9.11 9.44 5.28 5.29 7.83
Establishmentcharges 9.20 8.05 11.65 8.79 9.32 9.48 8.00 9.39 9.27 9.24
Total distributioncosts 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00

Economic and Political Weekly December 25, 1999 A-125

This content downloaded from 125.20.82.164 on Thu, 12 Sep 2013 04:39:44 AM

All use subject to JSTOR Terms and Conditions
low rate of 6.5 per cent a year and the GRAPH 3: COMPONENTSOF ECONOMICCOST OF WHEAT, 1990-91 TO 1998-99
contribution of these costs to total costs 900
fell over time. In the case of rice, however,
these costs rose quite rapidly, at 10.8 per 800-
cent annually. In some cases, agencies
owned by state governments make 700- -- Procurement
purchases of grain and then hand them price of
over to the central pool maintained by the 600- gain
, -"- Total
FCI. The FCI pays the state agencies for -- Procurement
500
this service. These costs increased very .' '--w;--?- \ talincidentals
' --,- Total
rapidly for rice, at 17.9 per cent annually, 400- distribution
during the 1990s; they also showed large costs
Economic
year-to-year variations (Graph 6). 300 . . .
cost
It is clear that the patterns of increase
-r
in total procurementcosts and in statutory 200 4 -- _____

charges are similar. Statutory charges are

a weighty component of total costs of the
FCI, and should be part of the debate on ~ ~-
~~~~~~~~~200
1000 ------- -
0
reduction of the costs of procurement. By
definition, statutorycharges are not under
the control of the FCI. Statutory charges
differ across states, and some state govern-
ments use this as a way of raising re-
sources.17 So a reduction of these costs GRAPH4: COMPONENTS
OFECONOMIC
COSTOFRICE,1990-91 TO 1998-99
involves other policy changes such as the 1200
rationalisation of taxes on food across oo~Total
states.18
1000 -
DISTRIBUTION
COSTS
- ~ctacosts - ? Procurement
In the early 1990s, the FCI reported 800 - price of gain
uniform distribution costs for rice and
wheat. The distribution costs for rice and I ~c~' I-_- Total
'3 600 -( procurement
wheat differed only after 1993-94. From .
^^^^r^
~~~~~
S^~~ -^-^^^^ ~ ~ ~~
- ,! incidentals
then on, in nominal terms, distribution
costs have generally been higher for rice.
• 400-
^^^^^ <--Total
|
distribution
In 1998-99, for example, the distribution costs
Economic
cost was Rs 170.8 per quintal of rice and
200 - ,... cost
Rs 154.9 per quintal of wheat. This is not ~7

surprising,as the quantityof rice distributed

~
. r~-~-~_ -"-"_--.....
........n __-'_ ~ _,I ~ ~_ .
_,--A
"^
..
...

in the PDS is much higher than that of

wheat. In 1997-98, 10.8 million tonnes of U\ -\ I I I I 00
rice and 6.9 million tonnes of wheat were oc, ,r cr o ,
,.~ ~ e~ cr' Cz
??'?
supplied through the PDS. It is thus likely
that rice is transported longer distances
than wheat.
Although distribution costs grew lessFreightchargeshavegrownmorerapidly banksand loans fromthe governmentof
for ricethanfor wheat(Table2). Forrice,
rapidly than procurement incidentals for India.Interestcharges,on average,were
Graph8 demonstratesvividly the steep
both rice and wheat, in absolute terms they 25.8percentof distributioncostsforwheat
are higher than procurement incidentalsrise in the costs of freightfrom 1994-95. and 29.4 per cent for rice during this
Any strategyof cost reductionwill have
for both rice and wheat. Distribution costs period. Interestcharges,however, grew
to examinefreightcostscarefully.Therise
havejj mped in the latterhalf of the period, less rapidly than other componentsof
in transportcosts in on accountof a rise
froni 1995-96 for rice and in 1997-98 for distributioncostssuchasfreight.Handling
in rail freights as well as a rise in the
wheat, to very high levels. Secondly, they expensesorthecostsof labourforhandling
volume of movement.19The evaluation
account for a sizeable proportion of total graincomprisea smallcomponentof total
undertakenby the BICPin the late 1980s
costs excluding the procurement price for costsbutgrewveryrapidly.Establishment
indicatedthatbetterplanningof movement
both crops and particularlyfor rice (52.76 costs or theadministrativeoverheadcosts
of grains could help reduce costs of
per cent for wheat and 72.15 per cent for of wheatgrewfasterthantotaldistribution
transport[BICP 1991]. In particular,it
rice on average). I now turn to the com- costs but this was not so in the case of
ponents of the costs of distribution wasnotedthatthequantityof grainsmoved rice. Storagechargesaccountedfor about
(Tables 4A, 4B and Graphs 7 and 8). was greaterthan the quantityprocured 9 percentof distributioncostson average.
Freight charges are the single most indicatingthat old stocks were moved a Storage costs rose quite rapidly in the
second or thirdtime as well.20
importantcomponent of distributioncosts, 1990s.Thereis probablyscopeforreducing
accounting for 38.13 per cent of distri- Interestcharges comprise the second thecostsof storageby,amongotherthings,
butioncostsin thecaseof wheatand33.68 most importantcost item,andreferto the betterplanning,utilisationandmodernis-
percent forrice(averagefor 1990-1998). interestpaidby theFCIonoverdraftsfrom ation of storagecapacity.21

A-126 Economic and Political Weekly December 25, 1999

This content downloaded from 125.20.82.164 on Thu, 12 Sep 2013 04:39:44 AM

All use subject to JSTOR Terms and Conditions
 The final componentof distributioncosts GRAPH5: COMPONENTS OFPROCUREMENT INCIDENTALS
FORWHEAT,1990-91 TO 1998-99
was losses due to transitandstorage.Losses 140 - .
account for a significant proportion of
costs. However, the share of losses in total - - Statutory
distribution costs has fallen sharply for charges
wheat over this period. For rice, the share 100 I
of losses in total costs rose in the mid- 80'a Labourand
1990s but it has fallen in the last two years
(Graph 8). Nevertheless, the ~~~~~~~~~~~80
decline in .JQ -'~ ~transport
losses over time should not detract from 0I 2--ccharges
60
the very existence and size of these losses.
(-d~~~~~~~~2 ^ ---Total
While some losses may be due to unfore- 40 -_ _ chargeson
seen factors like bad weather, a large part State
is on account of bad management and 20- - J government
~~ ,^ Total
corruption.
COMPARING 1990s TOEARLIERPERIODS O I , ,p, , , , ,
-p
The behaviour of different components _o\ ~~ ... _
of costs during the 1990s has been unlike ,h ,o\ o\ t55 o crLbou and

that during the 1980s. Excluding the last 0 0

two years, 1997-98 and 1998-99, there has
been an improvement in the operations of
the FCI.22Two indicators used to identify
operational efficiency of the FCI are the
ratio of economic costs to procurement 701-. _ . O statOe to govt.
price and the ratio of subsidy to procure-
ment price [Tyagi 1990].
The ratio of economic cost to procure-
ment price increased during the 1970s and
1980s, particularlyfor wheat (ibid). Data = 1 4 charges
for the 1990s show that after reaching a
peak of 185 in the 1991-92, the ratio ~~~~~~~~~_--60
--e- --+--- ?-
declined in the case of wheat (Table 5).
For rice, the ratio was more or less un- al20-l-, -Total
changed during the 1990s but lower than procurement
the peak of 143 reached in the mid-1980s. incidentals
10 _ _ c-- ----
In short, the costs of acquisition and
distributionrelativeto the price of purchase -\C
020 \ 0 0
_- l l- -n - a Labour
have fallen for wheat in the 1990s and
00 charges
have been relatively constant for rice. By
this criterion, the FCI has improved its 0 I 0~ 0% 0%
operational efficiency in the 1990s in the
case of wheat. In the case of rice, there
is clearly no worsening of operational
efficiency by this criterion. SUMMING UP curementcosts, grew at the same rateas
The totalsubsidydepends on the quantity procurementprices. Turning to distri-
distributedand the rate of subsidy or per ThedatafromFCIperformance budgets bution,the majorcontributorsto the cost
unit subsidy. From the mid-1970s to the show clearly that the increase in pro- escalationwerefreight,storageandhand-
late 1980s, the ratio of unit subsidy to curementpricewas a criticalfactorin the ling expenses. It is clearly importantto
procurementpriceincreasedforboth wheat increase in economic costs of rice and improvethemanagementof transportand
and rice: the ratio went from 30 to 45 in wheat.The shareof procurement pricein storagein orderto containthese costs.
the case of wheat andjumped from around total economic costs rose between 1990 Despite the absoluteincreasein many
1 to 34 in the case of rice [Tyagi 1990]. and 1998 for both wheat and rice. The componentsof costs, it is importantto
Turningto the 1990s, we find that the ratio procurement price,however,is a variable note that the ratio of economic cost to
of unit subsidy to procurement price of overwhichtheFCIhasno control.Setting procurement pricedeclinedfor wheatand
wheat rose to a peak of 76 in 1992-93 and theprocurement priceis apoliticaldecision remained constant for rice during the
then fell steadily till 1996-97 (Table 6). inIndia's parliamentary system;thecentral 1990s.Thisindicatesimprovedoperational
It has risen very sharply again in the last governmentsets the procurementprice efficiency of the FCI.
two years. For rice, the ratiopeaked at 37.6 based on the recommendationsof the
in 1990-91 and declined till 1994-95. For Commissionon AgriculturalCosts and IV
rice too, the ratio has risen in the last few Prices(CACP).Anothersteadilygrowing Efficiency of FCI versus Private
Trade: Some Price Comparisons
years. By the second measure, the per- componentof costs was statutorycharges
formance of the FCI improved in the early leviedby stategovernments;thesetoo are In the debate on food subsidies, it is
andmid-1990s buthas deterioratedsharply outside the controlof the FCI. Statutory often statedthat the FCI is 'inefficient',
in the last few years. charges,a significantcomponentof pro- and one of the major reasons for the

Economic and Political Weekly December 25, 1999 A-127

This content downloaded from 125.20.82.164 on Thu, 12 Sep 2013 04:39:44 AM

All use subject to JSTOR Terms and Conditions
growing size of the food subsidy is said GRAPH 7: COMPONENTS OFDISTRIBUTION COSTS FORWHEAT, 1990-91 TO1998-99
to be this 'inefficiency' of the FCI. It is 70 + Freight
argued that the "financial imbalance (of
the FCI) is largely due to the excessive cost 60 -
of its operations" [Sharma et al 1997]. / --Handling
Specifically, it is argued that the unit costs 50 /^- / expenses
of the FCI's operations "have been sub-
stantially higher than those of private -1~ 3.g:~~ 40 ,I~/ .
~Storage S ~chargs
traders"(ibid).Inarguingtheircase, Sharma
et al (1997) attempt to compare itemwise 30 - '
marketing costs of private traders with &B / ^ Interestcharges
those of the FCI. While systematic data 20
are available for the FCI, their discussion
of the costs of private traders is based on ~~~10 - - ~~~~~~~~-~~~~.--~ Transit and

assumptions and assertions. For example, storage

,01 , . . .
they simply state that"FCIstorage charges 0 shortats
are higher than those of private traders" >o- *Establishment
o o, o, ' ox r.: .
O
* F T
1$ W , cSl
charges
(ibid, p 24) or that "compared to private 2 S S | J
ON
traders,the administrativeexpenses of FCI 0t Q\
00
0%
are high" (ibid, p 25). There are no data
on private tradersto support these claims.
How do we judge the efficiency of GRAPH 8: COMPONENTS OFDISTRIBUTION COSTS FORRICE,1990-91 TO1998-99
in this 70- ... .. ...
marketing? Marketing efficiency
context has been defined as "themovement
of goods from producers to consumers at 60 -
the lowest possible cost, consistent with
/ Freight
theprovisionof services desiredby the con-
50 -
sumer"[Acharyaand Agarwal 1987:309].
Efficiency can be measuredatthe aggregate ---- handling
level or at the level of individual marketing 3 40- / expenses
functions (such as transport, storage). At t 30-^ ^.~ \\. g/ <~ X ' Storage
the aggregate level, the following are two 330- charges
measures used commonly in the literature Interest
, .'--'-
(ibid). 20 - . charges
(i) Marketing efficiency (E) is defined
as the ratio of the value of outputs to the ire^ - '(---i^ i - -Total transit
10 -- ^ --and storage
value of inputs. Assuming that market ==:-^^ .J^::::r:^::::s
shortages
prices are not distorted, the difference --Establishment
between consumer and producer prices is 0 - I charges
- C' eel ~"
taken as a proxy for 'output' and the costs 0% . 0% . 0%

of marketing as the 'inputs'. The higher

the value of E, the more efficient the
system of marketing.
(ii) The net margin (M), that is, the price distributesgrainin all partsof thecountry of the FCI with wholesalemarketprices.
spread (consumer price -producer price) andinregionswhereprivatetradeis limited Privatetraders,it is assumed,will set the
as a ratio of the final consumer price, is or even non-existent.23 Thirdly,thereare wholesalepriceat a ratethatcoverstheir
anothermeasure of efficiency. The higher differencesin the qualityof productsand costs and in this way we can examineif
the price spread, or the higher the value these are not easily accounted for in privatetraderssupplygrainto consumers
of M. the lower the efficiency. aggregatemeasures.Fourthly,any com- ata lowerprice.A systematicanddetailed
There are several conceptual and empi- parisonof averagecosts missesthe actual comparison of marketpricesandeconomic
ricalproblemsin computing such measures variabilityof costs and pricesover time costs of the FCI,by stateandregion,was
and using them to compare the efficiency and acrossagents.Fifthly,thereis a big undertakenby Jharwal(1998).24He ex-
of the private and public sectors. First, as problemof dataavailabilitywith respect aminedpricesforrice andwheatin major
noted in the previous section, the FCI to privatetrade.Thereis verylittlepublic states,forruralandurbanareasseparately,
serves many objectives, and being more informationon the costs of operationsof fortheyears1981-82to 1991-92.Inurban
'inefficient' thanthe privatesectorin terms private traders. For all these reasons, areas,the comparisonswere betweenthe
of these two criteria does not invalidate researchers haveoftenusedmeasuressuch economic costs of the FCI and the local
the efforts of the FCI. Secondly, since no as a comparisonof finalprices(see below) wholesaleprice. For urbanareas,I have
private traderengages in tradeon the scale to compare the relative 'efficiency' of extendedthecomparisonof relativeprices
undertakenby the FCI,simple comparisons privateandinstitutional channelsof trade. for the periodup to 1996-97.25Broadly,
between private and public trade may be the patternsare similarto thoseobserved
PRICE COMPARISONS
inappropriate. On the one hand, FCI can duringthe 1980s [Jharwal1998].
be expected to reap economies of scale Itis possibletoappraisetheperformance For rice, in a large numberof states,
from its huge operations. On the other of the FCI in relationto that of private wholesalepriceswerehigherthan theFCI's
hand, the FCI incurs extra costs as it tradersby comparingthe economiccosts economic costs (Table 7).26 In Andhra

A-128 Economic and Political Weekly December 25, 1999

This content downloaded from 125.20.82.164 on Thu, 12 Sep 2013 04:39:44 AM

All use subject to JSTOR Terms and Conditions
Pradesh, Assam, Gujarat, Kerala, deficitstates.Formoreaccuratecompari- nomiccostinmoststates,particularly after
Karnataka,Tamil Nadu and Delhi, the sons, we wouldrequirestatewisedataon 1983.This sectionis basedon pricecom-
ratioof marketpriceto economiccost of economiccostsfromtheFCI.Inthisregard, parisonsreportedby Jharwal(1998) as
the FCIwas generallygreaterthan one. the recentrecommendation by the CACP betweenruralretailpricesandretailPDS
However,the ratio declined duringthe thatthe economiccosts of the FCIshould prices.30The marketretailpriceof rice was
1990sindicatingthat the FCI lost some be estimatedon a monthwiseand state- higher than the economiccost in Assam,
of its priceadvantagein recentyears.The wise basis should be approved[CACP BiharandWestBengal(Table9). In most
ratiowas less thanone in Bihar,Madhya 1998]. otherstates,includingKamataka, Madhya
Pradesh, Haryana, Himachal Pradesh, Turningto ruralareas,the marketprice Pradesh,Orissa, Tamil Nadu and Uttar
PunjabandUttarPradesh.Thepatternwas of rice appearsto be lower thanthe eco- Pradesh,marketprices were below eco-
mixed in Maharashtra and West Bengal TABLE 7: RATIO OF WHOLESALEPRICE TO ECONOMICCOST FOR RICE, URBAN AREAS
(see summaryin Table 11).This exercise
showsthatin termsof the final price,the 1991-92 1992-93 1993-94 1994-95 1995-96 1996-97
marketdoes not always offer consumers Andhra Pradesh 1.12 1.05 1.01 1.10 1.05 1.00
a betterdeal thanthe FCI in distributing Assam 1.13 1.02 0.95 1.04 1.04 1.05
rice in urbanareas. Bihar 1.08 0.98 0.89 0.92 0.93 0.87
Gujarat 1.33 1.11 0.97 1.15 1.23 1.15
Thepictureis slightlydifferentforwheat 0.93 0.86 0.87 0.96 0.96 0.99
Haryana
pricesinurbanareas(Table8). Innorthern Himachal Pradesh 0.96 0.97 0.89 0.90 na na
India,in thewheat-producing belt,market Kerala 1.28 1.19 1.08 1.25 1.19 1.22
priceswerelowerthantheaverageecono- Madhya Pradesh 1.01 0.95 0.85 0.88 0.82 0.79
miccostof theFCI.Thisincludesthestates Maharashtra 1.10 0.95 0.85 0.91 0.94 1.10
Karnataka 1.12 1.05 0.94 1.06 1.07 1.10
of Haryana,Punjab,Uttar Pradeshand Orissa 0.95 0.80 0.77 0.85 0.82 0.80
Rajasthan.27 Marketpriceswerealsobelow Punjab 0.92 0.87 0.93 0.94 0.90 0.83
economic costs in Delhi and Himachal Tamil Nadu 1.13 1.14 1.04 1.08 1.11 1.09
Pradesh.ThesituationwasvariedinBihar, Uttar Pradesh 0.97 0.87 0.85 0.91 0.90 0.92
West Bengal 1.13 0.96 0.96 1.08 1.02 0.95
MadhyaPradesh,Maharashtra and West Delhi 1.12 1.04 1.03 1.00 0.99 1.04
Bengal.InKarnataka, themarketpricewas
Notes: I The annualwholesale price is an averageof the monthlywholesale price of variousvarieties
consistentlyhigher than the economic of rice in different urbanmarketsof the state.
cost.28InGujarat, theratioof marketprice 2 The data for Himachal Pradesh for the year 1994-95 is an average of price in only two
to economiccost increasedbetween1991 successive months in that year.
and1996.Thesedataindicatethatthe market 3 The data for West Bengal for the year 1994-95 is an 8 month average.
provideswheatata lowerpricethantheFCI Source: Agricultural Situation in India, several issues.
in a largepartof northernIndia.29 TABLE 8: RATIO OF WHOLESALEPRICE TO ECONOMICCOST FOR WHEAT,URBAN AREAS
Thereis a commonpatternthatemerges
in the pricecomparisonsfor urbanareas, 1991-92 1992-93 1993-94 1994-95 1995-96 1996-97
andthiscouldberelatedto transport costs. Bihar 1.13 0.85 0.85 0.86 0.87 1.04
Theeconomiccost of the FCIis basedon Gujarat 1.19 0.99 0.98 1.09 1.09 1.28
anall-Indiaaveragecost(nostatewiseesti- Haryana 0.84 0.66 0.68 0.69 0.70 0.86
Himachal Pradesh 0.90 0.67 0.71 0.76 0.72 0.83
matesareavailable)andit standsto reason Karnataka 1.52 1.24 1.23 1.29 1.29 1.60
thatsome states will benefit more from Madhya Pradesh 1.08 0.90 0.88 0.87 0.79 1.04
the averagingof costs such as transport Maharashtra 1.20 0.96 0.93 0.96 1.00 1.20
costs. Transportingrice from Punjabto Punjab 0.82 0.65 0.68 0.68 0.68 0.85
Rajasthan 0.97 0.71 0.75 0.75 0.74 0.93
Keralais likely to be expensive but this Uttar Pradesh 0.98 0.70 0.74 0.73 0.73 0.94
is notcapturedin uniformeconomiccosts West Bengal 1.03 0.81 0.83 0.85 0.86 1.01
wheretransport costs areaveragedacross Delhi 0.97 0.75 0.77 0.77 0.78 0.98
allstates.Inotherwords,privatetransport Notes: 1 The annualwholesale price is an averageof the monthly wholesale price of variousvarieties
costswill be lowerthantheFCI'saverage of rice in different urbanmarketsof the state.
cost in the surplusstates where grainis 2 The wholesale price of wheat in Karnatakais for 'Bansi' variety of wheat.
procuredandvice versain the distantand Source: Agricultural Situation in India, several issues.
TABLE
9: RATIO
OFRETAIL
PRICE COSTOFRICEINRURAL
TOECONOMIC AREAS,1981-82 TO1991-92
Year Kerala Madhya Orissa Tamil Nadu Andhra Bihar Karnataka Assam Uttar West Bengal
Pradesh Pradesh Pradesh
1981-82 0.94 0.99 1.17 1.15 1.12 1.25 1.20 1.16 1.02 1.23
1982-83 1.09 1.03 1.23 1.44 1.02 11.36 1.31 1.24 1.20 1.26
1983-84 1.00 0.92 0.92 1.09 0.86 1.08 1.14 1.20 0.95 1.01
1984-85 0.89 0.86 0.86 0.91 0.83 0.90 1.02 1.03 0.79 0.93
1985-86 0.91 0.82 0.83 0.98 0.84 0.93 0.88 1.02 0.82 0.97
1986-87 0.84 0.81 0.83 0.91 0.84 0.98 0.83 1.04 0.94 0.96
1987-88 0.89 0.91 0.95 0.99 0.95 1.12 0.95 1.15 0.99 1.01
1988-89 1.05 0.94 0.87 0.99 0.87 1.05 0.97 1.18 0.88 0.98
1989-90 0.72 0.91 0.81 0.99 0.80 1.02 0.90 1.08 0.83 0.98
1990-91 0.92 0.91 0.94 0.86 0.89 1.08 0.99 1.16 0.99 1.09
1991-92 0.79 0.87 0.82 0.83 1.30 1.07 0.99 1.16 0.97 0.98
Source: Jharwal(1998), Table 5.23.

Economic and Political Weekly December 25, 1999 A-129

This content downloaded from 125.20.82.164 on Thu, 12 Sep 2013 04:39:44 AM

All use subject to JSTOR Terms and Conditions
nomiccosts,especiallyafterthemid-1980s. increasedat around10 per cent a year have improved in the 1990s, particularly
The situationwas variablein Keralaand between 1990 and 1998. As the rate of for wheat.
AndhraPradesh.Inthe case of wheat,the inflationin the indexof wholesaleprices In a major evaluation of the FCI, the
averagemarketpricewas lower thanthe was 8.5 per cent over the same period, BICP committee made several useful
economiccost in states such as Gujarat, therehas been a small rise in real costs suggestions for streamlining operations
Punjab, Rajasthan,Haryanaand Uttar too. For bothrice and wheat,increasein and containing costs of distribution and
Pradeshwhereasthe FCIpricewas lower theprocurement pricewasthesinglemost these should be given serious consider-
inMaharashtra andKarataka(Table10).31 importantfactorin the rise in costs. Any ation. These suggestions include major
The picture was mixed in Bihar and substantialreductionin economic costs policy changes such as deferred pro-
MadhyaPradesh. requires control over the procurement curement and decentralised storage and
Thesecomparisonsrevealthe presence price,aparameter determinedbythecentral they need to be part of the debate on
of a rural-urban divide. In severalstates, governmentandnottheFCI.Thereis need improving food security. Implementing
ruralretail prices of rice, for example, to examine this tendencyof continuous some of these suggestions may requirenot
were lowerthanurbanwholesaleprices. and high rise in procurementprices. An only greater autonomy for the FCI but
This could be due to severalfactorsbut expansionof the geographicalcoverage major organisational changes as well.
it is also indicativeof imperfectmarkets. of procurementand promotion of de- There are many reports of corruption in
Also, as Ghosh(1998) observed,"evenin centralisedprocurement mayhelpcontain the FCI (and some of this is reflected in
deficit regions,ruralfood prices remain the rise in procurementprices.32 the 'losses' item of costs), and this must
10-20percent below urbanfood prices". Next, among procurementcosts of be checked by bringing in greater
In other words, privatemarketsare not wheat, incidentalstatutorychargesrose transparency and accountability in the
undertaking arbitrage satisfactorily steadily. For rice, a rapidly growing workings of the FCI.
betweenruralandurbanmarketsof even component of procurementcosts was Lastly, the paper attempted to compare
the same state. paymentsto state governments.Control the 'efficiency' of private traders and the
over both these componentsof procure- FCI. Given limited dataon marketingcosts
UP
SUMMING
mentcost requiresactionby stategovern- and prices of grain distributed in the two
TheFCI,as notedearlier,is theorganis- ments and bettercollaborationbetween channels, we attempted a comparison of
ation that implementsthe government's the FCI and state government-owned wholesale market prices and economic
policies on food and food security.The corporationsin procurementoperations. costs of the FCI. The price comparisons
FCI purchasesgrain from cultivatorsat Thecostsof transport, storageandhandling indicatedthatthe FCIcomparesfavourably
government-announced minimumprices. havealsogrownrapidly.Nevertheless,the with private traders in the distribution of
In yearsof badharvest,it providessome operationalefficiencyof theFCIseemsto rice in a large number of states. In the
insuranceby purchasinggrainat less than
normallystipulatedquality.The FCI has TABLE 10: RATIO OF RETAIL PRICETO ECONOMICCOST OF WHEAT IN RURAL AREAS, 1981-82 TO 1991-92
a hugedistributionnetworkthatsupplies Year MP Maharashtra Karnataka Punjab Rajasthan Gujarat Haryana Bihar UP
grain to consumersin all parts of the
country,includingin isolatedandremote 1981-82 0.99 1.07 1.21 0.76 0.95 0.92 0.74 0.98 0.75
locations.It also caters primarilyto the 1982-83 0.92 1.17 1.23 0.76 0.95 0.95 0.80 1.06 0.91
1983-84 0.96 1.21 1.33 0.74 0.85 0.99 0.73 1.05 0.77
lowerendof themarket.Giventhemultiple 1984-85 0.79 0.90 1.04 0.71 0.70 0.74 0.68 0.82 0.73
roles it performs,the materialreviewed 1985-86 0.97 1.09 1.26 0.74 0.77 0.99 0.70 0.93 0.78
aboveshowsthatthe FCIis not arguably 1986-87 0.92 0.96 1.20 0.67 0.71 0.80 0.65 0.90 0.71
worsethanthe privatesectorin termsof 1987-88 0.97 1.01 1.28 0.59 0.70 0.92 0.71 1.02 0.79
1988-89 1.02 1.05 1.38 0.80 0.80 1.02 0.76 1.00 0.88
marketingcosts. Pricecomparisonsindi- 1989-90 0.96 0.92 1.29 0.70 0.69 0.88 0.68 0.89 0.74
cated that privatetradersoffered lower 1990-91 1.03 1.00 1.28 0.77 0.78 0.88 0.75 1.02 0.81
pricesforwheat,particularly inthesurplus 1991-92 0.96 1.02 1.33 0.78 0.75 0.95 0.76 1.12 0.93
states(Table11). Whereasthe FCIcould Note: MP stands for Madhya Pradeshand UP for Uttar Pradesh.
providericemorecheaplythantheprivate Source: Jharwal(1992), Table 5.22.
sectorin manyareas,particularlyin the
urbanareasof deficitstates(suchas some TABLE 11: SUMMARY OF PRICE COMPARISONS
of thesouthernstates).Thepersistenceof Area/Grain FCIPriceLower MarketPrice Lower Mixed
largedifferencesin marketpriceswithin
states, across rural and urban regions, UrbanAreas
Rice AndhraPradesh,Assam, Bihar,MadhyaPradesh, Maharashtra,
signalsthe imperfectionsof privategrain
marketsandthelackof marketintegration. Gujarat,Kerala,Karnataka, Haryana,Punjab,UttarPradesh, West Bengal
Tamil Nadu, Delhi Orissa,HimachalPradesh
V Wheat Karnataka,Gujarat Punjab,Haryana,UttarPradesh, Bihar, Madhya
Concluding Remarks Rajasthan,Delhi, Himachal Pradesh,Maharashtra,
Pradesh West Bengal
To conclude,the debate on the food RuralAreas
subsidyneedsto shiftfroma narrowfocus Rice West Bengal, Assam, Bihar Tamil Nadu, UttarPradesh, Kerala,
on the size of the budgetaryallocationto Orissa,MadhyaPradesh, AndhraPradesh
all issues pertainingto food security.To Maharashtra,Karnataka
of thefood (after 1983)
providea betterunderstanding
this detailsof the Wheat Maharashtra,Karnataka Gujarat,Haryana,Punjab, Bihar,
subsidy, paperprovided
costs of the Food Corporationof India. Rajasthan,UttarPradesh MadhyaPradesh
Thispaperfoundthattheeconomiccosts Note: The classification of states for urbanareas is based on price data for the 1990s whereas the
of the FCI - for both rice and wheat - classification for ruralareas is based on data for the 1980s.

A-130 Economic and Political Weekly December 25, 1999

This content downloaded from 125.20.82.164 on Thu, 12 Sep 2013 04:39:44 AM

All use subject to JSTOR Terms and Conditions
states of northern and central India, over to the central pool maintained by the banks led by the State Bank of India.37
however, private traders had lower prices FCI. The FCI pays the state agencies for Interestcharges also depend on the level
for grain.The presence of systematic rural- this service. For wheat, these charges of stocks and the requirementsof working
urban price differences indicated the accounted for about 4.5 per cent of total capital. While the discussion of optimal
imperfections of grain markets. economic costs, on average, and grew at stock levels is beyond the sccpe of this
To end, given its special role in the 5.7 percent annually.These costs increased paper, it is worth pointing out that actual
implementation of food policy, and the very rapidly for rice, at 17.9 per cent levels of both operational stocks and grain
operational constraints that follow, the annually, but did not account for as large required for buffer stocks have risen in
FCI is not unambiguously worse - as has a share of total procurement incidentals recent years. Actual stocks held in the
been suggested in the literature - than as for wheat (Tables 3A and 3B). In the centralpool, forexample, have been higher
private trade in terms of costs of distri- case of rice, note the large year-to-year than minimum desirable stocks from 1994
bution. This is not to deny the scope for variations in the charges paid to state onwards.38In the case of rice, actualstocks
improvement in the functioning of the governments. The share of payments to were significantly greater than desired
FCI.I wish to arguethatit is worth building state governments in total procurement norms (e g, the ratio of actual to required
on the strengths of the FCI (such as its costs of rice increased from 12.2 per cent stocks was 235 percent in 1995). To under-
impressive distribution network) and in 1990-91 to 19.3 per cent in 1998-99. stand interest costs, we would also need
improving its performance rather than Distribution Costs
to examine how costs are attributed to
dismantling the FCI and leaving distri- distribution as against maintenance of
bution of grain to private traders. The sharesof specific components of the buffer stocks.
costs of distributionfor the years 1990-91 to (iii) Handling expenses are the costs of
Appendix 1998-99 are shown in Tables 4A and 4B. labour for handling grain, and include
COMPONENTSOF PROCUREMENTCOSTS AND (i) Freight charges refer to the costs of contract labour and own or departmental
DISTRIBUTION
COSTS,1990-91 TO1998-99 transportinggrain to the final distribution labour. Handling costs accounted, on
centres. As grain is usually transportedby average, for around 11 per cent of total
Procurement incidentals
rail, the costs depend primarily on rail distribution costs for rice and wheat.
The changing shares of the different freights. These costs include payment to Handling expenses increased at about 14
components of procurement incidentals the railways for demurrage,diversion and per cent annually for rice and wheat.
are shown in Tables 3A and 3B for wheat detention. They also include lorry charges (iv) Next in importanceis establishment
and rice, respectively.33 and steamer freight charges for the grain costs or the administrative overhead costs
(i) The first component is reportedunder transportedby road and steamer. Freight that include all salary and staff related
the headingof statutorycharges. These are charges are the single most important costs. Establishment costs accounted for
obligatory payments and include mandi component of distributioncosts, account- 9.2 per cent of distribution costs of wheat
fees, mandi cess, auction fees, commis- ing for 38.13 per cent of distributioncosts in 1990-91 and 10.2 per cent in 1998-99.
sion paid to 'Kutcha Arhatiya's' (licensed in the case of wheat and 33.68 per cent Overhead costs of wheat grew faster than
farmer'sagents) and purchaseor sales tax. for rice (average for 1990-1998). Freight total distribution costs but this was not so
The commission agent's fee is fixed by charges have grown more rapidly (9.6 per in the case of rice, where establishment
a stategovernment,as aretheotherstatutory cent annually) for rice than for wheat (8.5 costs accounted for 9.2 per cent of distri-
costs. Statutorychargesgrew at 11 per cent per cent annually). bution costs in 1990 and in 1998.
a year for wheat and 9.1 per cent a year (ii) Interest charges, the second most (v) Storage charges accounted for about
for rice. important cost item, refer to the interest 9 per cent of distributioncosts on average.
(ii) Labour and transportcharges refer paid by the FCI on overdrafts from banks There refer to the costs of storing grain
to the costs of mandi labour and the costs and loans from the government of India. by the FCIeitherin its own storagegodowns
of internal movement such as from the Interest charges were 25.8 per cent of or in hired godowns. The costs depend on
mandito the storagecentre. These services distribution costs for wheat and 29.4 per the type of storage used. Storage costs rose
are obtained by the FCI on contract from cent for rice.35 Interest charges, however, quite rapidly in the 1990s. In the case of
open tenders.The movement cost depends grew less rapidly than other components rice, for example, costs of storage grew at
on the location of mandis and distance of distributioncosts such as freight.Interest 10.8 per cent a year between 1990-91 and
from the nearest rail head or storage charges rose in the early 1990s, fell and 1998-99, and the share of storage costs in
depot.34 Labour and transport costs of then rose again and peaked in 1997-98. total distribution costs rose from 7.8 per
wheat rose in absolute terms but at the re- As part of a rising trend, interest charges cent in 1990-91 to 9.46 per cent in 1998-99.
latively low rate of 6.5 per cent a year and have fluctuated from year to year. In the case of wheat too, the contribution
the contributionof these costs to total costs The interest costs depend primarily on of storage costs to total distribution costs
fell overtime.They accountedfor 13.16 per the interest charged by commercial banks increased during the 1990s (Table 4A).
cent of procurement incidentals in 1990 for food credit. Up to 1994-95, the Reserve (vi) Losses due to transitand storage are
and 11.9 per cent in 1998-99 (Table 3A). Bank of India set a ceiling on the interest the final component of distribution costs.
In the case of rice, however, these costs that commercial banks could charge on Grain losses occur at several stages and
rose quite rapidly, at 10.77 per cent food credit or credit for food procurement. places. They include losses due to lost
annually. They accounted for 15.85 per This interest rate was high in the early railway wagons and shortages in despatch.
cent of total procurementcosts in 1990-91 1990s (a rise in interest rates was part of Losses are suffered on account of damage
and 17.9 per cent in 1998-99 (Table 3B). the process of stabilisation and monetary during storage. From time to time, quality
(iii) Amount paid to state agencies for tightening).36 After 1994-95, the Reserve specifications are waived or eased and
establishment, storage and interest for Bank, as part of the strategy of financial result in purchase of damaged crops that
stocks procured. In some cases, agencies liberalisation, stopped fixing a ceiling are left unsold and eventually turned into
owned by state governments make interest rate. The interest on food credit losses.39 Shortages also occur in deliveries
purchases of grain and then hand them is now set by a consortium of commercial from millers ortransporters.Inthe accounts

Economic and Political Weekly December 25, 1999 A-131

This content downloaded from 125.20.82.164 on Thu, 12 Sep 2013 04:39:44 AM

All use subject to JSTOR Terms and Conditions
provided,FCI has classified losses into in 1997-98 (Railway Yearbook).
20 Ultimately,anoptimalfood systemwouldseek
specificationfor purchaseby the FCI [Swami
1998].
storagelosses andtransitlosses butI have to disperse production. 40 Transit shortages are defined as losses that
groupedthemtogetherhere.40The share 21 See BICP (1991). occur during internaltransportand handling,
of losses in total distributioncosts has 22 For both rice and wheat sales realisationfell andstoragelosses arethosethatoccurinstorage.
fallen sharplyfor wheatover this period, sharply in the last two years. An important The methodof dividing total losses into these
caveatis thatthese areestimatesandnot actual two heads,however,can be questioned[BICP
from8.2 percentin 1990-91to 2.4 percent values but even then the sharp fall requires 1991].
in 1998-99.Forrice,theshareof losses in some explanation.Loweringof issue pricesas
Referemces
totalcostsrosefrom8.26percentin 1990- part of the targetedPDS and restrictionson
91 to 9.4 percentin 1996-97andthenfell. open market sales are possible explanatory Acharya,S S andN L Agarwal(1987):Agricultural
factors. Marketingin India, Oxford and IBH, New
In the last two years, there has been a 23 As in the tribal areas of north-easternIndia Delhi.
notablefall in losses on the rice account. (Julius Sen, pers comm). Bhagwati,J and T N Srinivasan(1993): India's
24 An all-India level comparisonof wholesale Economic Reforms, Government of India,
Notes prices of rice and wheat with economic costs Ministry of Finance.
of the FCI is reportedby Sharmaet al (1997). Bureauof IndustrialCostsandPrices(BICP)(1991):
[I wish to thankthe officials of the Food Corpora- Wholesalepricesvaryacrossstatesandacross Report on the Operations of the Food
tion of India for making available data from ruralandurbanareasandover months,and an Corporationof India, New Delhi.
their performance budgets. I am grateful to all-Indiaaveragefor wholesalepriceshas little Commission on Agricultural Costs and Prices
V K Ramachandranfor his suggestions on the meaning. (CACP)(1998): Reportof the Commissionfor
argumentspresentedhere, and to Jesim Pais for 25 The figure for wholesale prices was obtained AgriculturalCostsand Prices on Price Policy
assistancein preparingthe tables and graphs]. by averagingoverdifferentcentresin eachstate for Kharif Crops of 1998-99 Season, New
in each monthof the selectedyear as reported Delhi.
1 See, for example, Bhagwati and Srinivasan in differentvolumes of AgriculturalSituation Ghosh, J (1998): 'Assessing Poverty Alleviation
(1993), GoI (1997) and differentissues of the in India. All the selected centres were urban Strategiesfor their Impacton Poor Women',
EconomicSurvey after 1991. centres. UNRISD Discussion Paper 97, October.
2 It is suggested that the FCI can operateas an 26 One limitationof this approachis thatquality Governmentof India, AgriculturalSituation in
autonomousbody engaged in open market differencesarenottakenintoaccount.However, India, Ministryof Agriculture,Directorateof
purchasesand sales, andcompete with private the FCI sells threetypes of rice (common,fine EconomicsandStatistics,New Delhi, different
traders.See Sharmaet al (1997) and Jha and and superfine)and the marketdata also refer volumes.
Srinivasan (1998). to differenttypes of rice (excluding the very Governmentof India,EconomicSurvey,Ministry
4 For a discussionof the impactof the PDS on superiorbasmatirice). So, the marketprice is of Finance, differentyears.
consumers, see Swaminathan(1996). unlikelyto be higheronly on accountof better Governmentof India,AnnualReport,Ministryof
5 Theincidenceofbenefitsfromthesugarsubsidy quality. Civil Supplies, different years.
is likely to be verydifferentfromthe incidence 27 As mentionedabove, averagingof transport Governmentof India(1997):GovernmentSubsidies
of benefitsof the cerealsubsidy as the pattern costs by the FCIimpliesthatstateslike Punjab in India, Ministryof Finance, Departmentof
of expenditureon these commoditiesis going andHaryanaaresubsidisingmoredistantstates. Economic Affairs, Discussion Paper.
to differ across income classes. 28 This may be becausethe wholesaleprice data Gulati, I S and T N Krishnan(1975): 'Public
6 I thereforedo not discuss the activities of the reportedwas for only one superiorvarietyof Distributionand Procurementof Foodgrains:
FCI in respectof other commodities such as wheat. A Proposal',Economicand Political Weekly,
sugar and oilseeds. 29 DifferentoutcomesarereportedbyJayatiGhosh 10, 21, May 24, pp 829-42.
7 On the methodof apportioningcosts and the (1998). She comparedthe economic costs of Jayawardena,L, A Maasland and P N Radha-
anomalies therein,see BICP (1991). theFCIwiththeimplicitconsumerpricesemer- krishnan(1988): Stabilisationand Adjustment
8 There can even be situations of conflict in ging from the roundof the National Sample Policies and Programmes:CountryStudy15,
certainoperations,as for example, when state Surveyon consumerexpenditurefor 1993-94, Sri Lanka, WIDER.
governmentsdeterminethe 'mandis'in which and foundthatthe economic cost of FCI grain Jharwal,S M (1998): Public DistributionSystem
the FCI can operate. remainsbelow the marketpricein almostevery in IndiaReassessed,ManakPublications,New
9 Thefiguresfor 1997-98and1998-99arerevised urban area [Ghosh 1998]. Delhi.
estimates;all other figures are actual. 30 The retail PDS price differs across states, as Lustig,N (1992):Mexico:TheRemakingofan Eco-
10 Themajordetailedstudyof costs by component state issue prices are not uniform. nomy,BrookingsInstitution,Washington,DC.
is that undertakenby the BICP (1991). The 31 As mentioned the data on market prices of Pinstrup-Andersen, P, M JaramilloandF S Stewart
BICP study, however, pertainsto the 1980s. wheat in Karnatakaare for a superiorvariety (1991): 'The Impact on GovernmentExpen-
11 If the price indices for differentcomponents andthatmayexplainthehighratioin Karnataka. diture' in G A Cornia,R Jolly and F Stewart
of costshavegrowndifferently,then,of course, 32 In thiscontext,see GulatiandKrishnan(1975). (eds), Adjustmentwith a HumanFace, Vol 1,
changes in nominal values of different com- 33 A categorythatis not discussedhere is thatof ClarendonPress, Oxford.
ponentsarenot comparable.However,as such portclearancecosts.Thiscomprisesstevedoring Sharma,P, A GulatiandS Kahkonen(1997): 'The
disaggregatedprice indices are not available, charges,bagging and stitchinglevies, harbour Food Corporationof India: Successes and
this remains a hypotheticalquestion. dues, transitcharges etc that are incurredin Failuresin IndianFoodgrainMarketing',paper
12 Growth estimates are based on a log-linear the case of imports.As importshave declined, presented at the Conference on Agriculture
regression of all observations. these costs are not relevant for the 1990s. and Rural Reform Institutionsand Economic
13 The ratioof foodgraindistributedor 'lifted' to 34 Note thatthe FCIcannotchoose mandiswhere Policy,January21-22,Calcutta(IndianInstitute
for
quantityallocated, example, fell from 86 per it makespurchases;mandisareallocatedto the of Management,Calcutta).
cent in 1991-92 to 74 per cent in 1997-98. FCI by the state governments. It has been Subbarao,K, A U Ahmed and T Teklu (1997):
14 Thefastestcomponentwas handlingcosts (part arguedthatuneconomicsmallmandisareoften SafetyNet Programsand PovertyReduction:
of the costs of distribution)but it is a small allocated to the FCI while the largermandis LessonsfromCountryExperience,WorldBank.
componentof total costs (see Appendix). are allocated to state government-owned Swami, P (1998): 'Punjab'sProblem',Frontline,
15 The procurementcost reportedby the FCI is corporations[BICP 1991]. November 20.
not identicalto the Minimum SupportPrice 35 Note that the interest charges paid to state Swaminathan,M (1996): 'StructuralAdjustment,
announced by the government. The price governmentsfor theirprocurementoperations Food Security and the System of Public
announcedbythegovernmentis forfairaverage were included in procurement incidentals. Distributionof Food', Economicand Political
qualitygrain.TheFCImaybe procuringcertain Together,total costs due to interestpayments Weekly,June 29, pp 1665-27.
grainsof higher/lowerqualityatdifferentprices, are thus higher than reportedhere. Tuck, L and K Lindert(1996): 'From Universal
and the reportedprice is the average price of 36 The interestceiling on food creditwas 16 per Food Subsidiesto a Self-TargetedProgram:A
all grain procured. cent in 1990-91 and 1991-92, and 15 per cent Case Study in TunisianReform',WorldBank
16 For definitionsof various cost items, I have in 1994-95 (see Statement63 of the Reporton Discussion Paper No 351, WashingtonD C.
relied on BICP (1991). Currencyand Finance 1996-97). Tyagi, D S (1990): Managing India's Food
17 In the states of Haryana,Punjab and Uttar 37 Interestrateon food creditpeakedat 16.5 per Economy, Sage Publications,New Delhi.
Pradesh,themajorstatesforwheatprocurement, cent in 1996, and then fell to 14.9 in 1997 and World Bank (1996): 'India's Public Distribution
thereare severaltaxes on wheat procurement. 14.0 in 1998. System: A National and InternationalPer-
See BICP (1991) on statewise tax rates. 38 See Economic Survey, different years. spective', Poverty and Social Policy Depart-
18 This issue is unfortunatelynot being discussed 39 Most recently, after damage to the standing ment,November(Reprintedas R Radhakrishna
in the debate on tax reform. crop duringthe monsoonof 1998, the govern- andK Subbarao,WorldBankDiscussionPaper
19 Forexample,railfreightsincreasedquitesharply ment announced a lowering of the quality No 380, 1997).

A-132 Economic and Political Weekly December 25, 1999

This content downloaded from 125.20.82.164 on Thu, 12 Sep 2013 04:39:44 AM

All use subject to JSTOR Terms and Conditions