Food policy reforms

8/11/2019 Food policy reforms

1/55

Food Policy Reforms: A Rapid

Tour of PossibilitiesBharat Ramaswami

IGIDR Silver Jubilee InternationalConference

December 1-3, 2012


2/55

Collaborators

Milind Murugkar

Ashok Kotwal

Pulapre Balakrishnan


3/55

A Rare Moment?

Food policy institutions have enjoyed stability and continuity fordecadesdespite changes in scale and objectives

The public distribution system (PDS): origins in WW II rationingsystems.

The Food Corporation of India (FCI)the Central government

agency responsible for procurement and storage was set up in themid-60s.

Practice of offering support prices to rice and wheat also dates fromthat period.

At this moment, though, Indias food policy is in a state of flux.

Real possibility that Indias food policy institutions may look quitedifferent in a decade.


4/55

Pressures on Food Policy

Stunning GDP growth but only modest dent inpoverty.

Contradiction hard to ignore politically

The National Food Security Bill at the Centrallevel

Several food policy reforms at the Stategovernment level.

Civil Society activismamong other things theyhave demanded and obtained judicial oversightof the States food intervention.


5/55

The neo-classical economics case for

food policy interventions

Price supports and procurement

Absent risk markets, price supports can be Paretoimproving. Producers gain from insurance against lowprice outcomes. Consumers gain too: the supplyresponse to price insurance lowers food prices (Innes,QJE, 1990)

Subsidised food distribution to the poor

Traditionally, justified with reference to equityobjective.

A small literature now on growth impacts of safetynets (Alderman and Yemetsov, 2012)


6/55

Other instruments: open market sales,

public stocks?

In the Indian context, it can be argued that theseare the outcomes of price supports andsubsidized food distribution.

Open market sales occur fitfully and almostalways to dispose off excess stocks. There is noannounced protocol for these sales.

Although there are announced norms for public

stocks, these are driven mostly by the needs ofthe public distribution rather than marketstabilization.


7/55

Price supports in Practice

Price supports supplanted by administeredprices and procurement.

The farm lobby and its hold

Counter-moves by the government to reduce thecost of procurement by restrictions on exportsand other private sector activity

Has bolstered the profitability of the grain sector

distorting crop choice and diversification Unsustainable environmentally (e.g., paddy

production in Punjab)


8/55

Subsidised food distribution in practice

Identifying the poor has been difficult Drawing a line in the sand!

Massive exclusion of the poorboth exogenous andendogenous.

Illegal arbitrage between the PDS and the marketRent seeking and political patronage of PDS dealers

Unviable government marketing chain (PDS)volumesinsufficient to justify the costs.

This has encouraged illegal diversion, limited andunpredictable service timings and customer unfriendlypractices.


9/55

Facts are never enough!

Practically no disagreement about how

procurement and subsidized food distribution

work in practice.

But there are very different views about what

to do next.


10/55

The Tower of Babel: what do we do

next?

Policy advise from economists, multi-lateralinstitutions: target subsidies, make stateagencies open to competition, include privatesector

Pressure from activists, NGOs: Make subsidiesuniversal, no private sector, empowercommunities and enact laws to make the stateaccountable.

Government speaks in many voices: axe of fiscalconsolidation, social justice, farmer rights,consumer interests


11/55


12/55

PDS Forever? (Kotwal, Murugkar and

Ramaswami, 2011)

Balance between politics and markets.

Food subsidies ought to be near universal.Targeting is hard to do when there are so many

just above the poverty line. Exclusion errorsbound to happen.

Therefore, tolerate the leakage of resources tothe non-poor.

Among other things, use markets in the form ofcash transfers to reduce diversions and otherwaste.

No reason to tolerate such leakage of resources.


13/55

In the remainder of this talk.

Will not pursue further the issues of balance

between politics and markets with respect to

the grand design of food policy.

Use this perspective to assess the prospects of

`incremental reform in (a) storage and

procurement and (b) distribution.


14/55

Storage

Since 2010, the problem of insufficient storage

capacity has attracted both political and

media attention.

Article1-578444.aspx.htm

142399.html

31386742_1_lakh-tonnes-fci-and-state-

central-pool.htm

So how bad is the shortfall in capacity?
http://localhost/var/www/apps/conversion/tmp/scratch_8/Article1-578444.aspx.htmhttp://localhost/var/www/apps/conversion/tmp/scratch_8/142399.htmlhttp://localhost/var/www/apps/conversion/tmp/scratch_8/31386742_1_lakh-tonnes-fci-and-state-central-pool.htmhttp://localhost/var/www/apps/conversion/tmp/scratch_8/31386742_1_lakh-tonnes-fci-and-state-central-pool.htmhttp://localhost/var/www/apps/conversion/tmp/scratch_8/31386742_1_lakh-tonnes-fci-and-state-central-pool.htmhttp://localhost/var/www/apps/conversion/tmp/scratch_8/31386742_1_lakh-tonnes-fci-and-state-central-pool.htmhttp://localhost/var/www/apps/conversion/tmp/scratch_8/31386742_1_lakh-tonnes-fci-and-state-central-pool.htmhttp://localhost/var/www/apps/conversion/tmp/scratch_8/31386742_1_lakh-tonnes-fci-and-state-central-pool.htmhttp://localhost/var/www/apps/conversion/tmp/scratch_8/31386742_1_lakh-tonnes-fci-and-state-central-pool.htmhttp://localhost/var/www/apps/conversion/tmp/scratch_8/31386742_1_lakh-tonnes-fci-and-state-central-pool.htmhttp://localhost/var/www/apps/conversion/tmp/scratch_8/31386742_1_lakh-tonnes-fci-and-state-central-pool.htmhttp://localhost/var/www/apps/conversion/tmp/scratch_8/31386742_1_lakh-tonnes-fci-and-state-central-pool.htmhttp://localhost/var/www/apps/conversion/tmp/scratch_8/31386742_1_lakh-tonnes-fci-and-state-central-pool.htmhttp://localhost/var/www/apps/conversion/tmp/scratch_8/31386742_1_lakh-tonnes-fci-and-state-central-pool.htmhttp://localhost/var/www/apps/conversion/tmp/scratch_8/31386742_1_lakh-tonnes-fci-and-state-central-pool.htmhttp://localhost/var/www/apps/conversion/tmp/scratch_8/31386742_1_lakh-tonnes-fci-and-state-central-pool.htmhttp://localhost/var/www/apps/conversion/tmp/scratch_8/31386742_1_lakh-tonnes-fci-and-state-central-pool.htmhttp://localhost/var/www/apps/conversion/tmp/scratch_8/142399.htmlhttp://localhost/var/www/apps/conversion/tmp/scratch_8/Article1-578444.aspx.htmhttp://localhost/var/www/apps/conversion/tmp/scratch_8/Article1-578444.aspx.htmhttp://localhost/var/www/apps/conversion/tmp/scratch_8/Article1-578444.aspx.htm


15/55

Seasonal storage

Crop harvests occur at finite discrete points

(once or twice during a crop year) while

consumption is continuous.

Hence the crop needs to be carried from

harvest to the other months when there is no

harvest. This is the demand for seasonal

storage.


16/55

No seasonal pattern in grain

consumption

Quarter All grain Rice Wheat All grain Rice Wheat

Per capita & per month,Kg

Index with July-September = 100

July-September 11.7 6 4.32 100 100 100

October-December 11.55 5.9 4.16 96.29 96.3 96.3

January-March 11.58 5.99 4.08 94.27 94.3 94.3

April-June 11.44 5.94 4.38 101.35 101 101


17/55

Seasonal demand for storage

Principle: Grain must be equally allocated

over time.


18/55

Application

1. Compute marketed surplus

2. Assume that the portion consumed on-farmdoes not require commercial storage

3. No carry-overs of grain across marketing years(only seasonal storage considered).

4. In the harvest periods, consumption demand isinstantaneously met without any storage. Ofcourse some temporary storage is requiredbutthat could be in shops, transit or out in theopen.


19/55

Quarter-wise demand for storage

Rice Wheat Total

July 1 12.5 30 42.5

Oct 1 0 20 20

Jan 1 38.5 10 48.5

April 1 25 0 25

Marketed surplus of rice =50 mill tons

Marketed surplus of wheat= 40 mill tons

Oct 1, Kharif marketingyear: start with zero ricestocks

Jan 1 = 3/4th

of 50 = 38.5 April 1 = of 50 = 25

July 1 = 1/4thof 50 = 12.5


20/55


21/55

Demand for Public Stocks under the

National Food Security Bill

Scenari

os TotalProcurem

ent

Riceprocurem

ent

Wheatprocurem

ent

Ricestorage

on Jan 1

Wheat

storageon Jan

1

Total

storageon Jan

1

1 64 38.4 25.6 28.8 6.4 35.2

2 74 44.4 29.6 33.3 7.4 40.7


22/55

Implications

Calculations suggest that peak seasonalstorage demand is of the order of 41 milliontons in the immediate future.

As rice procurement takes place across bothOct-Dec and Jan-March quarter, aboveestimate is an upper demand.

To this add, the emergency reserverequirement for annual storage = 47 milliontons.


23/55

Supply-Demand gap

Policies emphasize the urgency of creating

more storage.

The gap between supply (32 mt) and

estimated demand (47 mt) is about 15 million

tons.

Yet even 47 mt is not sufficient today when

peak stocks top 70 mt. So what is wrong with

out calculations?


24/55

Supply-demand gapII

Our calculations assumed that procurement

would match the PDS commitments to

distribution.

This may not happen: procurement may

outstrip requirements as has been the case for

nearly 2 decades.


25/55

Procurement larger than PDS sales


26/55

Why excess procurement?

Procurement larger than PDS sales. No stabilization

Why? Farm lobby and coalition politics??

Other reasons: Suppose the central government onlywants to buy enough to meet PDS requirements. Then

the problem is: what is the right procurement pricethat elicits the required quantity?

Politicians and bureaucrats fear the embarrassment ofunder-supplying the PDS but receive no penalty forexcess stocks and high prices. Works to strengthen thefarm price lobby.

Bias in favour of higher than necessary procurementprices and therefore large procurement.


27/55


28/55

What can be done?

Open market sales

Basu (2010) proposed a mechanism of selling grain insmall batches to many traders and consumers tomaximise the impact of open market sales on price.

Basusproposal was made in the context of marketstabilizing intervention where procurement variesaccording to available supplies.

But as we have seen, Indian intervention has beensystematically biased towards subtracting supplies.

So the first best policy is to reduce procurement.


29/55

Policy options

Reform procurement from being open-ended to closed-ended.

This may be politically difficult.

Incremental reform proposalcreate a new agency under

the CACP called the Risk Management Agency (RMA). Let the FCIs liability be limited to the grain purchased for

PDS.

Stocks in excess will be transferred to the books of RMA.

This will (a) make excess stocks visible and (b) force the

office of CACP to take this into account in recommendingprocurement prices!


30/55

Part IIDistribution Reforms

The distribution of food subsidies happens withina federal structure.

Central government: largely responsible forfunding, procurement and transport of grain to

the States States: responsible for implementation and

delivery of food subsidies

Distribution reforms have to be understood withreference to initiatives at the Centre as well aswith the States.

We report on some state-level reforms


31/55

Principal elements of distribution

reforms

1. Computerizing the data base of beneficiaries

2. New listing of beneficiaries

3. Issue of new ration cardsincorporating bar-coding and biometric id

4. Authentication of transactions by smart cards

and/or biometric id.

5. Recording of transactions in real time or

near-real time through IT systems.


32/55

Use of IT systems in recording data is a

major distribution reform

Automation of retail transactions leads to real timeinformation on supply gaps at each retail outlet.

Hence, it is possible to connect this module with aback-end module of inventory management system

(stocks and grain movement between different storagedepots) resulting in automated supply and movement.

This reduces paperwork and increases the timelinessand predictability of supplies.

This is the major reform of PDS in the state ofChhatisgarh


33/55

Authentication of transaction is a

major distribution reform

Illegal diversion of grain: arbitraged grain isrecorded as sold at the issue price in governmentrecords.

This is possible to do when the sales are to

fictitious consumers. Multiple ration cards maybe held by a single consumer or ration cards maybe `bogus.

This can only be stopped if the retail transaction

is authenticated in a fool-proof manner. This is the major distribution reform being

attempted in Madhya Pradesh and Gujarat


34/55

Authentication by Smart cards

With or without biometric id (like bank cards with anumeric code id).

Requires the use of smart card readers at the retaillevel connected to a central server.

Connectivity at all FPS may be a problem. Pilot project in Chandigarh where the infrastructure is

reasonable.

However, record is mixed because of failure of smart

card readers. Smart card based authentication proposed for big

urban centres of Chhatisgarh.


35/55

Biometric ID

If employed at the retail level, it is subject to

the same limitations as smart cards (without

biometrics)namelyconnectivity and the

possibility of `engineered device failure.

Intermediate system: Use biometric id at

select offices to obtain food coupons which

are then redeemed at the FPS.

Connectivity is not required at all retail points.


36/55

Costs of distribution reform

Reported costs are often incomplete because

of `zero-price transactions between

government agencies.

MP model: all activities out-sourced to a

private consortium for 5 years.

Cost = Rs. 4611 million or Rs. 461 crores.


37/55

Individual State Experiences:

Chhattisgarh

Computerization of Procurement system under

the decentralized procurement scheme.

Timely management of supplies because of

computerization and control over supplies (notdependent on FCI for grain movement to state).

`Door-step delivery

No transaction authentication mechanismsmart card based solution (without biometrics

proposed for urban areas).


38/55

Chhattisgarh

Extended coverage: 70% of population underBPL/AAY. Low exclusion errors.

Lower BPL rates: Rs. 2 for rice and Rs 1 for AAY

State spends significant resourcesover Rs.1000 croresin addition to Central subsidy

De-privatization of FPS: Shops are run by

community organizations: self-help groups,panchayats and coops. Such experimentshave not worked elsewhere.


39/55

Chhattisgarh

Low prices, extended coverage and well

publicised timely supply have worked to

create public consciousness about the right to

receive PDS entitlements. This is claimed tohave checked illegal diversions.


40/55

Chhattisgarh: Implications?

Supporters claim that (near) universal coverageand de-privatization of FPS is responsible forsuccess and can be replicated elsewhere.

Several unique features Willingness to spend out of its resourceshigh

political commitment

Bureaucracy is unusually pro-active in monitoring thesupply chain. This is essential because the incentives

for arbitrage continue to be present. Neither can be taken for granted elsewhere because

of entrenched interests in existing PDS


41/55

Gujarat

Encompasses all 4 components of distribution

reform.

Pilot project of reform: one FPS in each taluka

of 22 districts are participating.


42/55

Gujarat Model: A Food Coupon Model

All households to re-register to obtain bar-

coded ration cards. All household particulars

digitised and biometrics recorded.

Enrollment in this process requires an

electoral photo ID.

Using the bar-coded ration cards, beneficiary

visits an E-kiosk (in gram panchayat duringpilot).


43/55

Gujarat Model 2

Computer operator uses a bar code reader toenter beneficiary details. On verification ofbiometrics, bar coded food coupons issued.

Biometric verification requires real timeconnectivity.

Beneficiary redeems coupons at designatedFPS.

FPS retailer submits these coupons at E-Kioskto be read into an electronic sales register.


44/55

Gujarat Model 3

Back-end inventory management system

linked with distribution network is in the

works.

IT solutions developed by NIC and in-house

team.

Modest capital costs of Rs. 800 million and

recurring costs of Rs. 250 million. However,this does not include NIC costs.


45/55

Gujarat model: Assessment

Transaction authentication is the focus and thestrength of the model.

Weakness Requires consumers to make 2 visits monthly the E-

kiosks are often more distant. Internet connectivity is not yet good enoughthe two

visits could stretch to more

Problem could be less acute if coupons were issued

annually or bi-annually. Requirement of electoral ID is bound to exclude some

of the poor.


46/55

MP Model

Similar to Gujarat in intent and scope.

But different in terms of design and execution.

Further MP is not at a pilot-stage but at a roll-

out stage.

Biometric id is at the heart of the MP model.

Designed to be compliant withAadhar

, thenation wide biometric id project.


47/55

MP Model 2

Aadharenrollment is a pre-requisite for PDS.

Camps organized in villages forAadharenrollment.

Enrollment used to create a newcomputerized data base of PDS beneficiariesand to the issue of newAadharbased rationcards.

Food coupons couriered annually tobeneficiaries.


48/55

MP Model 3

Biometric id verified on receipt with portable

devicies using GPRS connectivity of cell phone

networks.

Beneficiary redeems coupons at FPS. Redeemed coupons picked up and transported to a

central high speed scanning centre.

On coupon verification, electronic system generates

a report of transaction and sales which can be used

for allotment, supplies and movement


49/55

Execution

Execution outsourced to a private consortium.

No capital costs for government; pays Rs. 10.98per transaction.

Strengths of model Transaction authentication

Avoided the smart card route which is demanding ofinfrastructure and which is prone to sabotage.

Zero upfront costs for governmentall risks of projectimplementation with consortium.

Incentives of vendors aligned with customers.


50/55

Challenges to MP Model

Will enrollment leave out many of the eligible?

And how easy will be for them to subsequentlyenroll?

Reliance onAadhar: Issue of ids is not keepingpace with enrollment.

Will the real-time verification ofAadharid work?

MP model does not yet include computerizationof procurement and storage (unlikeChhattisgarh).

MP d G j t d l diff t f


51/55

MP and Gujarat model different from

the Direct subsidy model Direct subsidy model championed by the Task Force on

Direct transfers.

Here the grain (or the subsidised commodity) flows throughthe government marketing chain at market prices.

So no incentive for leakage.

Consumer buys from authorised retailer at market prices.

The retail transaction is subject to aadhar id verificationand is linked to a payments system.

This link transfers the subsidy directly to the beneficiarys

account. The direct subsidy model requires devices to capture

biometric id and transaction at the retail level while thecoupon model needs it only when the coupons are issued.


52/55

Summary Findings I

We are short of storage capacity

Extent of shortfall would be less if

procurement were to be in line with

distribution.

While this might be difficult to implement

straightaway, it should be possible to devise

new institutional structures to make `excessstocks visible.


53/55

Summary Findings II

Distribution reforms have enormous potential

because most states are starting at a high level

of inefficiency.

While these reforms have wide support, the

entreched interests in unreformed PDS are

strong and political commitment in the States

cannot be taken for granted even if it allowsreforms to be initiated.


54/55

Summary Findings III

Distribution reforms hold the promise of

accountability and transparency.

Computerizing the supply chain and digitising

records are low-hanging fruit.

Transaction authentication is more demanding

but with higher payoffs too.


55/55

Summary Findings IV

Smart card based systems are not practical at this point.

Intermediate systems such as food coupons based on

biometric id are more practicalperhaps even more so

than the direct subsidy transfer model of the Central

government.

It is imperative therefore to allow and experiment with

different models.

The question is how to design them without imposingadditional costs of access on poor consumers.

Food policy reforms

Documents