ODI November 2007 1

THE 2006/7 AGRICULTURAL INPUT SUPPLY PROGRAMME,

MALAWI

Andrew Dorward

School of Oriental & African Studies,University of London

2

Outline

� Background

� Programme development & implementation

� Programme impacts

� Work in progress, preliminary conclusions

DisclaimerThe results presented in this presentation are preliminary results in an

ongoing evaluation of the agricultural input subsidy programme of the Malawi Government. They have been developed in joint work involving SOAS,

Wadonda Consult, Michigan State University, and the Overseas Development Institute, and funded by DFID, USAID, and the Future Agricultures

Consortium. Findings are subject to revision. The views expressed are those of the author and should not be attributed to the Malawi Government, DFID,

USAID, FAC, SOAS, MSU, Wadonda Consult, or ODI.

3

Background: Key Issues

� Most farming households are poor and own food production lasts up to September – November, and so they are net consumers

� Highly volatile maize prices

� Force poor consumers to grow maize even with very low yields

� Inhibit investment in maize production

� The rural economy is trapped in low productivity maize cultivation

� History of

� subsidised fertilisers to the 80’s & 90s

� universal and targeted inputs from late 90’s

� maize and fertiliser politics

� contested and changing ‘solutions’ & policies

4

Background: Maize yields & rainfall

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

1.8

2.0

1990

/91

1991

/92

1992

/93

1993

/94

1994

/95

1995

/96

1996

/97

1997

/98

1998

/99

1999

/00

2000

/01

2001

/02

2002

/03

2003

/04

2004

/05

2005

/06

2006

/07

MK/kg (1990)

Hybridyield

MMP!

Localyield

5

Background: Maize yields & real prices

0

0.2

0.4

0.6

0.8

1

1.2

1.4

1.6

1.8

1990

/91

1991

/92

1992

/93

1993

/94

1994

/95

1995

/96

1996

/97

1997

/98

1998

/99

1999

/00

2000

/01

2001

/02

2002

/03

2003

/04

2004

/05

2005

/06

2006

/07

MK/kg (1990)

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

1.8

2.0

tonnes/ha

Peakprice

Minprice

Hybridyield

6

Background: National fertiliser sales, 1972-2007

National Fertiliser sales

0

50,000

100,000

150,000

200,000

250,000

300,000

350,000

1972

/73

1975

/76

1978

/79

1981

/82

1984

/85

1987

/88

1990

/91

1993

/94

1996

/97

1999

/00

2002

/03

2005

/06

Years

'000 tonnes Total

Compound

Amm. Sulph.

CAN

Urea

7

Background: National fertiliser imports breakdown, 1998-2007

Importation

0

50

100

150

200

250

300

350

400

1997/9

8 199

8/99

1999/0

0 200

0/01

2001/0

2 200

2/03

2003/0

4 200

4/05

2005/6

2006/7

'000 tonnes

ADMARC.SFFRM Private sector

Total

8

Background: National fertiliser sales breakdown, 1998-2007

Sales

0

50

100

150

200

250

300

350

1997/9

8 199

8/99

1999/0

0 200

0/01

2001/0

2 200

2/03

2003/0

4 200

4/05

2005/6

2006/7

'000 tonnes

ADMARC/SFFRM Private sector **

Total

9

Background: National fertiliser purchases breakdown, 1998-2007

Purchases

0

50

100

150

200

250

300

350

1997

/98

1998

/99

1999

/00

2000

/01

2001

/02

2002

/03

2003

/04

2004

/05

2005

/620

06/7

'000 tonnes

Subsidised* Unsubsidised

Total Predicted

10

Background: 2005/6 subsidy

� Poor harvests & withdrawal of universal starter pack & food crises in early 2000’s

� (different) fertliser subsidies advocated by both main parties in 2004 election

11

Background: 2005/6 subsidy

� 2004/5 uncertainty about subsidy – late targeted input programme, low production & food crisis

� 2005/6 government funded subsidy� Mixed donor views, no direct donor support (general budget support)

� 147,000MT target, ‘maize fertilisers’ coupons (NPK & urea) redeemable for 950MK; ‘tobacco fertilisers’coupons ( D compound & CAN) redeemable for MK1450

� 2 million ‘base’ coupons, 1.1 million supplementary coupons printed; 0.58 million issued; 2.6 million total redeemed: fertiliser stock constraint

� Coupon targeting criteria highly variable; 2, 1 or 0 coupons per household

� 70,000 tonnes out of 130,000 imported by private sector

� Late distribution

12

Background: 2005/6 subsidy Impact

� 131,000 tonnes fertiliser sold, all through ADMARC/SFFRM (stock not coupon constraints)

� MK7.2 billion ($51mill) cost (MK5.1 billion ($36 mill) budget) US$

� Large drop in commercial sales, due to subsidy & price rise?

� Contraction in private sector distribution network� Good rains, record maize production, 2.6 million tonnes� Low maize prices and higher wages benefit poor rural

consumers

rev

13

Conventional subsidies

� ‘Conventional’ policy objectives� Promote technical change & productivity, & economic & technically efficient use of inputs

� Support farm incomes� Pan territorial� Provide delivery systems

� Modalities� General, quota or channel price reduction

� Problems� Cost control� Exits� Targeting and diversion� Inefficient over-use due to low price� Regressive (benefit larger farmers)� Crowding out of private sector

14

Rethinking subsidies

� New objectives?� Private input market development (crowding in)� Replenishing soil fertility� Social protection to targeted recipients with market externalities

� Supply chain development thresholds (kick-starting markets)� Lower staple food prices and raise wages� Structural livelihood and economy changes

� New modalities?� ‘Smart’ subsidies (vouchers) for cost control & private sector market development

� Longer time frame� Complementary policies

� Output price policies� Roads & communications infrastructure investment� Agricultural services – use of inputs, diversification� Private sector development & Non-farm diversification� Social protection policies

� Political context

15

2006/7 evaluation

� DFID funded (DFID Malawi, Future Agricultures) & USAID� Imperial College / SOAS, Wadonda Consult, Michigan State, ODI

� Information sources:� Implementers: Min of Ag HQ, Donors, Logistics Unit, Key informants – 3 districts

� Input supplier study: HQs, Branches / agrodealers (3 districts)

� Recipients / non-recipients: Focus group discussions & case histories (3 districts), national household survey (NSO)

� Household and rural economy modelling� March preliminary report on implementation� December final report

� Note:� Extremely complex issues, multiple & diverse stakeholders, data difficulties, many uncertainties

� Presentation in December in Lilongwe

16

1. Input subsidy implementation

Scale, cost, modalities, timing,

‘targeting’

3. Effects on Input Supply System

Private sector, ADMARC/ SFFRM

Profits, cash flow, confidence, volumes, prices,

investment, innovations, other services

4. Effects on Macro economy

Fiscal balanceForeign exchange balance

Health, education, infrastructure spending

2. Effects on Rural Households

2a Effects on recipients

(different hholds & hhold members) Farm & non-farm

activities & productivityLabour hire in/out

Crop purchases / salesIncome

Food securityWelfare

2b Effects on non-recipients

(different hholds & hhold members)

Farm & non-farm activities & productivity

Labour hire in/outCrop purchases / sales

IncomeFood security

Welfare

Local & national, market & non-

market relations: (maize, labour, cash,

land, etc prices & flows)

Other macro-

economic management

Political & policy

processes

Global & regional prices

Maize price policies

Previous season(s)’events & outcomes

Weather

Disease (HIV/AIDS,

malaria, etc)

Other rural economic activities

Other Social Protection Measures

17

Elements of system� Coupon allocation & distribution� Input procurement & distribution & sales� Coupon redemption/ Input purchase� ….Input use� ….Impacts

Inputs:� ‘Maize fertilisers’:

� 23:20:0 NPK� Urea

� ‘Tobacco fertilisers’: � Compound D� CAN

� Maize seeds: � Hybrid� OPV

18

Development of the programme

� Constant change from April to October

� Fertiliser procurement tenders

� Roles, terms & private sector players in in fertiliser distribution pushed by….

� Donor funding

• for seed with fixed value vouchers;

• logistics unit;printing of coupons; communications; transport;

• financing facility for unsold government fertiliser stocks

� Development of new systems, opportunities & relationships

� but …. uncertainty & delay for all stakeholders

19

Coupon allocation

� 3 million fertiliser coupons (50kg fertiliser bags)

� 2.6 million NPK and urea coupons base allocation; 0.4 million D compound & CAN

� Four types of fertiliser redeemable for MK950

� 2 million seed coupons: No farmer payment, each coupon worth MK400 (2kg hybrid, 3kg and 4kgOPV)

� Security printed coupons, serial numbers, triplicate

� Base allocation proportional to maize & tobacco hectaragebetween regions & between EPAs; adjustments between districts; finalized end September

� 1.003 million supplementary coupons issued in December in response to demand – and more

rev

20

Coupon allocations per ha & per hhold

0.630.611.060.320.82National

0.520.300.980.300.68South

0.750.621.120.140.98Centre

0.701.221.860.960.90North

------------------------per hhold -----------------------

1.412.722.530.711.83National

1.433.162.690.831.86South

1.392.472.070.261.80Centre

1.393.303.711.911.80North

SeedfertsTotalSupp.Base

MaizeTobaccoMaize fertilisers

rev

21rev

2006/7 Maize fertiliser coupon allocations to hectarage by district

0.00

1.00

2.00

3.00

4.00

5.00

6.00

7.00

8.00

Chi

tipa

Karon

ga

Nkh

atab

ayM

zim

baR

umph

iN

tchi

siD

owa

Mch

inji

Kasun

guSal

ima

Nkh

otak

ota

Lilo

ngw

eD

edza

Ntc

heu

Balak

a

Man

gochi

Mac

hinga

Zomba

Chi

radz

ulu

Mw

anza

/Nen

oThy

olo

Phalo

mbe

Mul

anje

Blant

yre

Chi

kwaw

aN

sanj

e

District

Supp.

Base

22rev

2006/7 Maize fertiliser coupon allocations to grower hholds by

district

0.00

0.50

1.00

1.50

2.00

2.50

3.00

Chi

tipa

Karon

ga

Nkh

atab

ayM

zim

baR

umph

iN

tchi

siD

owa

Mch

inji

Kasun

guSal

ima

Nkh

otak

ota

Lilo

ngw

eD

edza

Ntc

heu

Balak

a

Man

gochi

Mac

hinga

Zomba

Chi

radz

ulu

Mw

anza

/Nen

oThy

olo

Phalo

mbe

Mul

anje

Blant

yre

Chi

kwaw

aN

sanj

e

District

Urea + NPK voucher allocations per maize hh

Supp.

Base

23

Coupon distribution

� Distributed through districts, ADC’s/ EPAS

� Variable involvement of TAs (chiefs/ village heads)

� People unaware of numbers of coupons per village, often disappointed & suspicious of misappropriation, but difficult to determine its extent

� Coupons & funding to districts in Oct/(Nov)

� Coupon distribution often late due to late arrival in districts and lack of funds for distribution

� Coupon targeting criteria & systems highly variable

� 56% households received coupons (64% in northern region, 58% in Centre, 53% in South, livelihood/ district variations)

� 2, 1 or 0 coupons per household (24%, 26%, 44%)

24

Coupon distribution per household

0 1 2 >2

North 36% 15% 42% 7%

Centre 42% 28% 23% 6%

South 47% 28% 21% 4%

Non-poor 43% 26% 24% 7%

Poor 44% 26% 25% 4%

All 44% 26% 24% 6%

Fertiliser

25

Fertiliser sales

� Parastatal channels:

� Imports to central depots, deliveries to markets, sales

� Delayed imports (late tenders, mixed performance, tender changes)

� Delayed market openings

� Delayed sales

� Private channels

� Delayed agreements

� Delayed sales

26

NPK Urea CAN D Comp Total Seed

ADMARC/SFFRM sales (from stock records)

NR 8.17 8.10 2.32 2.12 20.71

CR 21.98 25.30 3.19 4.47 54.94

SR 23.80 24.17 0.95 0.87 49.79

Total 53.96 57.56 6.46 7.45 125.44

Retail sales

NR 4.26 5.12 0.67 0.64 10.69

CR 12.56 11.10 3.37 2.11 29.13

SR 4.15 4.28 0.52 0.48 9.43

Total 20.97 20.50 4.55 3.23 49.25

Total sales

NR 12.43 13.22 2.99 2.75 31.40 0.54

CR 34.54 36.39 6.56 6.57 84.07 2.03

SR 27.95 28.45 1.47 1.35 59.22 1.93

Total 74.93 78.06 11.02 10.68 174.69 4.52

Subsidy Sales (‘000MT)

27

Inputs Sales

� Cumbersome & slow redemptions

� Inclusion of private sector improved input access, especially seeds due to participation of agro-dealers

But….

� Most rural areas not served by private traders

� Long queues, limited stocks and slow replenishment, particularly at ADMARC markets

� Some mismatches between stock allocations, coupons, and demand

� Lack of commercial fertilizers at ADMARC and SFFRFM outlets

� Cumbersome & slow payments process for private suppliers

28

� Varied across districts� Press (media) reports tended to pick isolated bad stories –some confirmed

� Sale of coupons�Most were suspicions / rumours, prompted by opaque systems, lack of info, inadequate coupons (? fewer than received in 2005/06 ?)

� Purchase of coupons by 5% of hh, 38% from traders, (600MK)� Fake coupons not widely reported � Allocation of coupons at village level not transparent in most districts (lack of clarity on eligibility)

� ‘Tips’ to buy fertilisers the most common irregularity�Most common in centre & south (monitored in north )�At both parastatal and private sector retail markets�MK50 – MK500, exacerbated by long queues (FGD)�Median 0, mean +100 (Urea) +160 (23:20) (survey)

� Exchange of coupons for other items/inputs reported ……

Irregularities

29

Beneficiary Access� Lack of clarity on targeting criteria led to variations in

inclusion and exclusion errors� Wide variations in targeting criteria� Different allocation philosophies & systems� More wealthy households more likely to get coupons & to get more

� High transport and transaction costs � Some households were too far from input markets and travelled as much as 30 km, Long queues (FGD)

� Mean & median travel/waiting time: 13 & 9 hours� Mean & median travel/waiting costs: MK247 & 150� Mean & median distance: 7 and 5 km

� Some households had difficulties in obtaining cash� Poor information (e.g. on 1st & 2nd round) (FGD)� Limited safety net opportunities cf. 2005/06 (FGD)� 3% financed from PWP, 4% loans, 4% gifts, 22% ganyu, 50% savings

30

Fertiliser acquisition by land holding size

43.647.847.239.3% buy subsidized fert.

29.83627.924.6% buy unsubsidized fert.

2006/07 crop season

38.831.841.942.2% buy subsidized fert.

34.750.234.620% buy unsubsidized fert.

2003/04 crop season

31.330.634.229% buy subsidized fert.

39.452.337.627.6% buy unsubsidized fert.

2002/03 crop season

All-20-1.39<0 .81

Landholding size

tercile (ha)

31

Programme Costs

� Total identified cost to government: US$64 million� US$13 million over the budget (which was already 47% of MoA budget)

� sale of nearly 25,000 tonnes of fertiliser above budget, caused by issuing of supplementary coupons.

� Total donor costs estimated at US$9.5 million� Cf MoA budget US$120mn� Cf national budget US$1.1bn� Limited operational support to implementing government

agencies� Operational budgets for ADMARC and SFFRFM not fully funded e.g. ADMARC only got 33%• Unable to recruit required staff in markets, long queues and kick backs

� Delayed and inadequate funding to districts to support sensitization and coupon distribution (had to use funds from normal operating budget)

32

Private / public delivery costs

�Major part of parastatal costs is supply to SFFRFM depots (average cost US$455/tonne)

�Total cost per ton sold is around $490 for both parastatal and private sector sales

�Difficult to compare ADMARC/SFFRFM and private sector cost and relate it to efficiency:

�ADMARC/SFFRM overhead costs difficult to establish

�Private sector minimum price not known

�ADMARC/SFFRFM served more remote areas …

…. but

�ADMARC/SFFRFM had higher % of sales in South than North

33

Evaluation

� Was the 2006/7 subsidy worthwhile?

� What can be done to improve the benefit cost ratio?

34

Programme Impacts

� Record national fertiliser sales: 296,000MT� Incremental fertiliser use? 90-100,000 MT?

�40% to 50% of subsidy displaces commercial sales?� Increased hybrid area: +20% on 2005/6?� With good rains, record crop estimate. � Reduced search for casual (ganyu) employment

�Higher wages�More farm production

� Lower prices for consumers …but …� Low returns & disincentives for surplus maize growers� Commercial input imports/sales better� Fertilizer sector confidence improved but still very cautious� Seed sector more enthusiastic about hybrid� Independent retailers (agrodealers) still big problems� Government confidence in private sector improving� Welfare & development benefits?

35

Was the 2006/7 subsidy worthwhile?

� ‘Simple’ cost benefit: value of incremental maize production greater than costs

1. Incremental maize production & its valuation?

1. Fertiliser impacts (diversion, displacement, allocation, yields)

2. Seed impacts

3. ‘cash transfer’ impacts

2. Incremental tobacco production?

3. Crop switching?

4. Other household level opportunity costs?

5. Total programme costs?

6. Add social protection/ welfare benefits?

7. Add developmental benefits?

8. Administration / overhead costs?

9. Subtract fiscal opportunity costs?

� Other economic & social benefits?

36

Estimating incremental maize production

� National crop estimates:

� 2003/4: 1.7 mill tonnes - TIP

� 2004/5: 1.2 mill tonnes – late TIP, mixed rains

� 2005/6: 2.6 mill tonnes – subsidy & good rains

� 2006/7: 3.2 mill tonnes – subsidy & good rains

� Regression estimates of 2006/7 seed & fertiliser impacts from 0.4 to 1.0 mill tonnes!

� Agronomic estimates:

� 0.5mill tonnes @15:1 grain:N, without seed impacts or ‘cash transfer’ effects

� depends on

• timing of fertiliser use,

• displacement

� Overall estimate? ?0.55 million tonnes?

37

Direct Benefit : Cost sensitivity analysis

216%238%80.973.5175.0250700I

121%133%80.973.598.0140700H

95%105%80.973.577.0110700G

170%187%80.973.5137.5250550F

95%105%80.973.577.0140550E

75%82%80.973.560.5110550D

93%102%80.973.575.0250300C

52%57%80.973.542.0140300B

41%45%80.973.533.0110300A

+10%$mill$/MT000 MT

BABAValuePriceProd'n

Benefit:CostCostsMaizeMaizeMaize

Incremental

38

Critical determinants of B:C ratio

� Grain benefits from incremental N

� Timing of fertiliser delivery & use

� Hybrid seed

� Scope of programme (maize, cash crops)

� Incremental N per MT subsidised

� Diversion – administration systems

� Displacement – targeting

� Farmer expectations

� Costs

� Fertiliser prices

� Administration costs

� Other benefits

39

Further economic & social benefits …� Complex impacts

� household & market� short, medium, long term� vary with

• household characteristics, • access to coupons• use of coupons• area (socio-agro-economic)

� supply chain – crowding in our out� Wider development objectives affect design/ implementation

� Increase land & labour productivity, maintain soil fertility � Increased grain availability, lower maize prices, raise wages & real incomes, stimulate non-farm demand

� Domestic supply & demand stimuli to diversify out of maize to other crops & non-farm goods & services

� Major implications for beneficiary targeting, maize price stabilisation, policy stability & transitions, complementary programmes

40

Subsidy impacts – hhold & market, now & later,

Input Subsidy

Poorer households

Less- poor households

Resale

Incremental use

Displacement use

Y1 Increased real incomes

Y1 Increased production

Y2 Increased real incomes

Farm/ non farm investment

Y2 Reduced maize prices

Y1 Increased wages

Y2 Increased production

RURAL

ECONOMY

RURAL

HOUSEHOLDS

Y2 Increased wages

Farm/ non farm demand & investment

Input service demand & investment

41

Subsidy impacts – vary by hhold & area, pro-poor

KAS zone BOR zone

Base year 1 year 2 year 2 Base year 1 year 2 year 2

subsidy no

subsidy subsidy subsidy

no

subsidy subsidy

Wage change (%

base) +3% +3% +5% +3% +6% +9%

Maize price (%

base) 0% -7% -8% 0% -9% -8%

Total seed subsidy (MT)

1,482 0 1,482 754 0 754

Total fert subsidy

(MT N) 12,971 0 12,971 6,602 0 6,602

Total seed (MT) 2,599 3,306 3,683 3,655 781 1,402 1,563 1,623

Total fert. (MT N) 17,503 24,856 20,032 24,002 5,770 10,188 7,066 10,068

Displacement Seed 52% N/A 29% 18% N/A -12%

Fert 43% N/A 50% 33% N/A 35%

Reselling of fertilizer

16% N/A 5% 24% N/A 7%

Poor

hh 7% 8% 14% 6% 6% 10% Real net

income

change All

hh 4% 1% 4% 4% 2% 5%

42

Low producer investment

Unstable maize prices

Low maize & agric productivity

Consumer ‘lock in’to low productivity

maize

Low & vulnerable real incomes

Low demand for non-agric goods & services

UNSTABLE POLICIES

UNSTABLE WEATHER

SLOW PRIVATE SECTOR

DEVELOPMENT

POOR ROADS

SOCIAL PROTECTION

INPUT SUBSIDY

MAIZE PRICE & TRADE POLICY

CREDIT,RESEARCH, EXTENSION, CASH & OIL CROPS

ROADS

PRIVATE SECTOR,NON-FARM

43

What can be done to improve the benefit cost ratio?

�Objectives� Targeting - ?hhold, ?area, ?crop� Reduce displacement� Scale� Cost� Scope� Systems

� Coupons?• Type – fixed value?• Allocation –comprehensive?• Distribution• Redemption• Smart cards? ID cards?

� Roles & relationships of government & private sector?� Remote areas?

� Exits / transitions

�Timing�Communication�Consistency�Transparency�Trust�System adaptation/ transition�Policy coordination – maize trade & prices, social protection, etc

�Budgetary & political controls�Quality & probity audits

ODI November 2007 44

THE 2006/7 AGRICULTURAL INPUT SUPPLY PROGRAMME,

MALAWI

funded by

DFID USAID FutureAgricultures

(DFID)

Overseas Development Institute

School of Oriental & African Studies/ Imperial College London

Wadonda Consult

Michigan State University