Jump to ContentJump to Main Navigation
Agricultural Input SubsidiesThe Recent Malawi Experience$

Ephraim Chirwa and Andrew Dorward

Print publication date: 2013

Print ISBN-13: 9780199683529

Published to Oxford Scholarship Online: January 2014

DOI: 10.1093/acprof:oso/9780199683529.001.0001

Show Summary Details
Page of

PRINTED FROM OXFORD SCHOLARSHIP ONLINE (www.oxfordscholarship.com). (c) Copyright Oxford University Press, 2018. All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a monograph in OSO for personal use (for details see http://www.oxfordscholarship.com/page/privacy-policy).date: 19 August 2018

FISP activities and achievements

FISP activities and achievements

Chapter:
(p.87) 5 FISP activities and achievements
Source:
Agricultural Input Subsidies
Author(s):

Ephraim Chirwa

Andrew Dorward

Publisher:
Oxford University Press
DOI:10.1093/acprof:oso/9780199683529.003.0005

Abstract and Keywords

The implementation of the FISP as a national programme is complex, involving several activities and interaction and coordination of various stakeholders at different levels. Drawing on programme records and household surveys, this chapter provides detailed information on the changes that have taken place in the design and implementation of the programme from 2005/6 to 2010/11. It is argued that implementation of the programme has largely been successful, with lesson learning and adaption allowing improvements in design and implementation. Cross party political commitment has been an important feature in the implementation of the programme. Nonetheless, major challenges remain in terms of targeting, timeliness of input delivery to farmers, the involvement of the private sector, payment of ‘tips’ by smallholder farmers, and cost control.

Keywords:   input subsidies, targeting processes, coupon distribution, programme costs, Malawi, fertilizers, seeds

5.1. Introduction

This chapter describes the implementation of the Malawi FISP from its inception in 2005/6 to 2011/12. Our aims in this chapter are to

  • set out the major activities involved in implementing the programme, and the core determinants of programme outputs which in turn determine programme impacts, as discussed in Chapters 6 and 7;

  • describe the major features of the programme with regard to design and implementation features identified at the end of Chapter 2 as critical for programme evaluation (although some aspects of targeting, rationing, and private sector input supply are addressed in later chapters); and

  • consider some of the political and other influences that have affected programme design and implementation.

We begin the chapter by introducing the core elements of the programme and major changes from 2005/6 to 2011/12. The bulk of the chapter then provides more detail on these, on implementation achievements and outputs of the programme, and on some of the determinants of changing practices. We go into considerable detail on some topics both to document these topics and to provide readers with information that is needed to understand issues discussed in later chapters. We conclude by discussing how consideration of the implementation of the Malawi subsidy programme fits within, extends, and/or challenges understandings of theory, practice, and policy set out in Chapters 2 to 4. We first, however, conclude this introduction with an overview of the major tasks and core activities involved in programme implementation and generation of its core outputs. A simplified summary of these is provided in Figure 5.1.

There are a number of points to note from Figure 5.1. First, following on from key lessons from the reviews in Chapters 2 and 3, programme outputs (p.88)

FISP activities and achievements

Figure 5.1. Major tasks in programme implementation

Note: see list of abbreviations for description of stakeholders.

Source: adapted from Dorward and Chirwa (2011c).

are defined as timely coupon redemption and incremental input sales. The activities and tasks required to deliver these outputs are then divided into two broad groups: input management and coupon management, with further division into input purchases and market opening on the one hand and coupon supply and beneficiary identification on the other. Across these divisions, however, there are major planning interactions, as coupon allocations to different areas depend upon overall planned input purchases and input distribution to different areas depends upon coupon allocations to those areas. There are also major implementation interactions between input and coupon management as regards timing of separate activities and of the exercise as a whole. Starting at the bottom of the figure, before coupon redemption can start in an area the markets must be stocked and open and beneficiaries must have been issued with their coupons. Similarly, input distribution to depots and markets requires the prior allocation of inputs to the areas in which markets are located. Allocations to areas, however, must follow initial decisions on the total volumes of inputs to be sold under the programme.

Finally, the figure also provides some indication of the way that the programme’s complexity and scale pose major challenges in programme planning and implementation. These activities are carried out for several different sets of inputs (for example, different fertilizers and seeds), over the whole (p.89) country in rural areas that are often poorly served by roads and services, serving dispersed farmers (many of whom have very low levels of literacy), without access to advanced information technology systems. There is also a wide range of stakeholders in the programme, often with different understandings of the objectives of the programme and of how it is supposed to be implemented, and different personal and professional interests in it, with some of these interests more and less ‘legitimate’. Fraud and theft are also major temptations and threats (criminals are not included as stakeholders, but their influence and effects on the programme also have to be recognized). All of this occurs on a very large scale. For example, in 2008/9 more than 1.5 million fertilizer coupon beneficiaries were selected from over 2.5 million farm households, 3.5 million fertilizer coupons and 2.5 million seed coupons were printed and distributed, and 3.5 million bags of four kinds of fertilizer were purchased and distributed, together with 2.5 million bags of two different kinds of maize seed and three kinds of legume seed. These subsidized commodities were together worth around US$275 million,1 with each fertilizer coupon’s value greater than 10% of annual household income for more than around 40% of the population. Discussion of the programme’s implementation, achievements, and weaknesses should take into account these considerations.

We now consider major implementation activities in turn, considering each of the activities in Figure 5.1 in terms of their structures, volume, and outputs over the period 2005/6 to 2011/12. We begin, however, with a brief overview of core programme features and of their evolution and of the context of that evolution from 2005/6 to 2011/12.

5.2. Programme design and evolution

The core stated objective of the FISP has consistently been to improve resource-poor smallholder famers’ access to improved agricultural inputs in order to achieve their and national food self-sufficiency and to raise these famers’ incomes through increased food and cash crop production. Later years of the programme have given greater emphasis to concerns for vulnerable farm households. Throughout this, however, there has been an emphasis on programme beneficiaries as farmers and producers, with little emphasis on beneficiaries as consumers (beyond the recognition of the programme’s contribution to improving household food self-sufficiency) and there has been no emphasis on the indirect programme effects on maize prices and hence consumers.

The core of the programme has been the use of vouchers (or coupons) to target approximately 50% of farmers in the country to receive fertilizers and (p.90) improved seeds for staple food (maize) production. ‘Maize fertilizers’ have been provided in a package of one voucher for a 50 kg bag of 23:21:0 +4S basal fertilizer (NPK) and one voucher for a 50 kg bag of urea for top dressing. Improved maize seeds subsidized under the programme were initially open pollinated varieties (OPVs) but there has subsequently been much greater emphasis on hybrid maize varieties. The seed and fertilizer packages drew on longstanding Ministry of Agriculture and Food Security (MoAFS) crop production recommendations. In the early years there were further vouchers for tobacco fertilizers, and in later years support for tobacco inputs was withdrawn and more emphasis given to provision of legume seeds. Grain storage chemicals were also provided from 2007/8 and cotton seed and chemicals provided in some years, but the disbursement and reporting channels and methods for these have been different from those for fertilizers and for maize and legume seed, and we do not report on them in any detail. The subsidization of complementary inputs (fertilizers, maize and legume seed, and maize storage chemicals) have been the main form of complementary investment in or associated with the programme, but there has also been some provision of extension messages on input use, and under the Agricultural Sector Wide Programme some investment in a nationwide set of on-farm trials of different integrated soil fertility management technologies.

Throughout the FISP, the government and other stakeholders have worked with varying success and agreement on innovations to address difficulties, improve programme performance, respond to changing political and economic conditions, and broaden impact. These changes have emerged from formal and informal discussions, reviews and lesson learning within government and with other stakeholders, and from changing policy concerns in a changing economic and political environment. Bilateral and multilateral donors have engaged substantially with the government over the programme from the 2006/7 season, with specific funding of some activities. This has involved consistent support for seed subsidies, logistics, monitoring and evaluation, and some fertilizer transport. They have also supported ‘buy back’ arrangements in some years (allowing the government to carry forward unused stocks from one year to the next with the costs of these stocks in the budget and accounts for the year of their use rather than purchase) and assisted with costs of auditing, anti-corruption measures, coupon printing, and exceptionally high fertilizer prices in 2008/9.

The major modifications in different years are shown in Tables 5.1 and 5.2 and involved changes in

  • volumes of subsidized fertilizer and seed sales (both maize and legumes) and of grain storage chemicals and cotton chemicals and seed, with (p.91) fertilizer volumes rising and then falling but seed volumes rising apart from a fall in maize seeds in 2011/12;

  • reliance on private sector imports to supply parastatal fertilizer sales (generally rising) and private sector involvement in retail sales of subsidized fertilizer only in 2006/7 and 2008/9 but in seed sales for all years except 2005/6;

  • programme objectives and beneficiary targeting criteria and systems placing increasing emphasis on vulnerable beneficiaries;

  • beneficiary registration and beneficiary selection, voucher distribution, and market monitoring systems;

  • redemption prices falling; and

  • coupon design and security features and processes increasing fitfully.

Further details and explanations are provided on these changes in the following sections.

Table 5.1. Principal programme features, 2005/6 to 2011/12

2005/6

2006/7

2007/8

2008/9

2009/10

2010/11

2011/12

Fertilizer voucher distribution (MT equivalent)

166,156

200,128

216,000

195,369

160,000

160,000

140,000

Total subsidized fertilizer sales (MT)

Planned

Actual

137,006

131,388

150,000

174,688

170,000

216,553

170,000

197,498

160,000

159,585

160,000

160,531

140,000

139,901

Fertilizer voucher value, approx. (MK/bag)

1,750

2,480

3,299

7,951

3,841

5,237

6,536

Redemption price (MK/bag)

950*

950

900

800

500

500

500

Subsidy % (approx.)

64%

72%

79%

91%

88%

91%

93%

Subsidized maize seed (MT)

n/a

4,524

5,541

5,365

8,652

10,650

8,245

% Hybrid seed

0%

61%

53%

84%

88%

80%

68%

Legume seed (MT)

24

1

1551

2726

2,562

Cotton seed/chemicals

No

No

Yes

Yes

No

No

Yes

Total programme cost (MK million)

Planned

Actual

5,100

4,480

7,500

10,346

11,500

13,362

19,480

33,922

21,908

15,526

19,700

21,868

21,586

23,455

(*) 950MK per bag for ‘maize fertilizers’ and 1450MK per bag for ‘tobacco fertilizers’.

(p.92)

Table 5.2. Evolving programme implementation features, 2005/6 to 2011/12

Subsidized inputs

Voucher distribution system

Voucher redemption systems

Other system features

2005/6

Maize & tobacco fertilizers, Maize seed (OPV)

First round district allocation by maize areas, distribution through Traditional Authorities (TAs), subsequent rounds less transparent

Only through SFFRFM & ADMARC

Fertilizer coupons not specific to fertilizer type.

2006/7 changes

Hybrid & OPV maize seed

Distribution through varied stakeholders

Fertilizers also through major retailers; flexible maize seed vouchers through a wide range of seed retailers

Coupons specific to fertilizer type. Fertilizer buy back system. Logistics unit involvement

2007/8 changes

Very limited legume seed; cotton seed & chemicals

District allocation by farm hh & areas, distribution through MoAFS/Village Development Committees (VDCs)

Cotton inputs through Agricultural Development Divisions (ADDs)

Reduced copies of coupons. Remote Extension Planning Area (EPA) premium

2008/9 changes

Tea & coffee fertilizers, maize storage chemicals

Use of farm household register, open meetings for allocation & disbursement led by MoAFS. Additional ‘flexi-vouchers’ for maize or legume seed allocated separately

Fertilizers only through ADMARC & SFFRFM; Grain storage chemicals through ADDs

Extra coupon security features & market monitoring; No remote EPA premium; ADMARC computers for voucher processing

2009/10 changes

No cotton chemicals. No tobacco fertilizers, Increased legume seed, more maize seed (hybrid & OPV) per pack

Elimination of 2nd and 3rd round distributions. Use of voter identification for registration, receipt & redemption.

Variable top up for maize seed max 100MK

Complex extra coupon security features in centre & north. Features (e.g. numbering) varied

2010/11 changes

No major changes

No major changes except ‘flexi-vouchers’ replaced by legume vouchers added to standard package

No major changes

No major changes

2011/12 changes

Major foreign exchange & fuel problems; cotton seed & chemicals

No major changes

Cotton inputs through ADDs

Coupons printed outside Malawi with packs straight to Districts

(p.93)

5.3. Input purchases and distribution

This section describes processes, achievements, and performance in the procurement and distribution of fertilizers and of maize and legume seeds.

5.3.1. Procurement and distribution systems

The programme has used two fertilizer procurement and distribution systems in different years. The dominant system, with distribution and retailing through two parastatal corporations (ADMARC and SFFRFM—Smallholder Farmers Fertiliser Revolving Fund of Malawi), was the only system used in 2005/6 and from 2008/9, while in 2006/7 and 2007/8 this was augmented by parallel private sector retailing of subsidized fertilizers. With parastatal distribution, the government calls for tenders for the supply of specified fertilizers to three SFFRFM depots (in the southern, central, and northern regions). After a formal process of tender analysis and awards, successful bidders (both private importers and SFFRFM and ADMARC) then deliver contracted quantities to the specified depots, which also hold any programme stocks bought forward from excess supplies in the previous year. The government (working through the programme’s Logistics Unit from 2006/7) then contracts private transporters to ‘uplift’ fertilizers from the regional depots to ADMARC and SFFRFM local area markets which act as the retail outlets for subsidized fertilizer supplies. Timely availability of fertilizers for redemption by beneficiaries (p.94) requires timely award of tenders and tender deliveries, timely uplifting of supplies from depots to markets (to supply the markets and to free storage space for deliveries from importers to depots), and timely sales from markets (to supply beneficiary demand and to free storage space for deliveries from depots to markets).

Timely sales of fertilizers before or immediately on commencement of the rains are critical for a good fertilizer response, particularly for ‘basal’ NPK fertilizers applied at planting. However limited storage space at depots, limited transport capacity, poor feeder roads once the rainy season has started, and limited storage space at markets pose major logistical challenges in maintaining stocks for selling large quantities of fertilizers quickly after the start of the rains. Absence of such stocks leads to shortages at markets, which in turn leads to farmers wasting time and money searching and waiting for stocks. Shortages also give corrupt sales agents and others opportunities to extort extra payments from beneficiaries.

As mentioned above, a parallel fertilizer procurement and distribution system using private sector retail outlets was implemented for two years, 2006/7 and 2007/8. Under this system six input suppliers with rural chains of retail outlets were contracted to procure, store, and distribute the fertilizers that they sold under the programme, with stocks of ‘subsidy fertilizers’ not distinguished from stocks of fertilizers for unsubsidized sales (they were only treated differently at the point of sale, and in the subsequent submission of vouchers to the MoAFS Logistics Unit for redemption at previously agreed prices). An important part of this system was the funding by DFID of a financing agreement with Stanbic Bank to address government concerns that ADMARC and SFFRFM might be left holding unused stocks if the private sector’s subsidized sales were larger than expected. Under this agreement, Stanbic bought unsold fertilizer stocks from the government at the end of the season for resale to the government at the start of the next season at the same price (plus storage and financing costs). In 2006/7 prices varied by district to reflect varying transport costs to different districts, but in 2007/8 a further ‘remote areas premium’ was introduced to promote companies extending their distribution network into more remote zones. Private sector sales of subsidized fertilizer were cancelled by the government at short notice in the 2008/9 season when contract negotiations were at an advanced stage, on the basis of (as far as we know) unsubstantiated reports that some private companies had been misusing fertilizer coupons and accepting them in exchange for other sales of other products.

With regards to maize and legume seeds, from 2006/7 onwards these have been supplied and distributed by private seed suppliers and sold through both private and parastatal retail outlets. A small number of seed companies negotiate annual coupon redemption prices for different seed types (p.95) and packs (with the ability to charge a further ‘top up’ price to farmers in most years). Retailers (retail chains, small scale agro-dealers, and parastatal outlets) are then supplied with seed for sale under the programme, and have to return seed coupons redeemed by farmers back to the seed companies, who then present them to the MoAFS Logistics Unit for redemption payments.

5.3.2. Fertilizer procurement and distribution

Procurement of fertilizers for subsidized sale by ADMARC and SFFRFM is a major and critical activity in programme implementation, affecting beneficiaries’ access to fertilizers (in terms of timing, volumes, and access costs), programme impacts, and programme costs (with fertilizers accounting for around 75% of net programme costs from 2009/10 to 2011/12). While procurement of fertilizers and seeds for sale through private channels is also important, this is coordinated and implemented by private companies, with critical government involvement restricted to tendering at the beginning and payment at the end. We report on these later when we consider coupon redemption and seed sales under the subsidy, and beneficiary information about delivery performance. In this section we focus on government and parastatal procurement and distribution achievements and performance for fertilizers. We discuss more on this in Chapter 8.

Figure 5.2 shows fertilizer supplies to the programme by source from 2005/6 to 2011/12 (excluding unutilized supplies carried forward). The graph demonstrates well

  • the increasing volume of subsidized fertilizer from 2005/6 to 2008/9 and its subsequent decline;

  • the general trend for increasing reliance on private sector supplies, but highly variable volumes of private sector supplies between years (rising year on year from 2005/6 to 2008/9, then falling back markedly in 2009/10);2

  • government distribution of 125,000 MT or more in every year of the programme.

Given these large volumes, performance in government procurement and distribution of fertilizers is very important to programme performance as a whole. This can be considered in terms of timing and costs. (p.96)

FISP activities and achievements

Figure 5.2. Fertilizer supplies by source, 2005/6 to 2011/12

Sources: Logistics Unit annual reports.

Procurement activities have generally been completed a little earlier each year, and this has been associated with substantial improvements in end of October and end of November deliveries in 2009/10 and 2010/11 (there were also greater carry-forward stocks in 2009/10). Despite these improvements, however, end of October deliveries have still been less than 80% of total supplies even in the best years, and have frequently been below 40%. End of November deliveries were as low as 70% of total supplies in 2008/9, and in only two years have all supplies been delivered to depots by the end of December. The limited impact of earlier tendering on deliveries is due partly to lack of firm delivery dates and late delivery penalty clauses in tender contracts, partly to the failure of some suppliers to deliver at all (with consequent late contracts with alternative suppliers), and partly to constraints on storage space at SFFRFM regional depots. This is affected by difficulties in uplifting stocks from the depots to rural markets, to which we now turn.

Earlier deliveries to depots have allowed improvements in uplifting stocks from the depots to rural markets (see Figure 5.3), with uplifts by the end of November rising steadily from a little over 64% to 86% from 2006/7 to 2010/11. However this should be 100%, indeed all supplies should be available for sale by mid November, whereas even in 2010/11 only 70% was uplifted by the end of October and 85% by end November. Although uplifts are constrained by the availability of stock in depots, they are also constrained by transport capacity, poor roads, and availability of storage space in markets, with the latter determined by the timing of seasonal markets’ opening and by (p.97)

FISP activities and achievements

Figure 5.3. Montly uplifts to rural markets (‘000 MT & % parastatal sales by end of month)

Source: Logistics Unit weekly and annual reports.

sales disruptions to stock flows. Sales are determined not only by stock availability, but also by beneficiaries being in possession of redeemable coupons. We discuss coupon distribution systems determining this later, in Section 5.4. First, however, we briefly examine another aspect of procurement performance: prices paid for inputs.

Figure 5.4 compares average fertilizer costs per metric ton (MT) for NPK (23:21:0 +4S) and urea delivered to the three depots with international prices for DAP and urea over the period 2006 to 2011 (with good data on procurement prices available for the last three years but weaker data for the previous three years). Data are shown each year from March to November, with March to April generally being the time at which tenders have been submitted (and hence the basis for tender pricing, although some contracts were established later in each season) with delivery from August to December: actual timing of different suppliers’ purchases is not known, nor are the prices paid. The figure shows

  • broad patterns of change in prices and in difference between delivered and international prices;

  • sharp rises and falls in international prices, particularly for phosphate fertilizers in 2008, with tenders and being placed when prices were rising to their peak;

  • sharp rises in urea prices from March to June 2011, with a peak in September;

  • more generally, the very high volatility in fertilizer prices in recent years;

  • variations in price differentials between international and landed prices and between urea and NPK prices.

(p.98)

The variability in international prices, with foreign exchange and other risks, and consequent exposure of suppliers to large potential losses and gains must lead to suppliers building substantial risk margins into their tenders. Measures that allow shorter bid validity periods and faster tender processing and awards should lead to lower prices. These measures would also allow for earlier delivery of supplies. Costs might also be reduced if the government took on more of the foreign exchange risks and was able to manage these risks more effectively than private suppliers.

Further insights into fertilizer prices paid by the government can be gleaned by comparing cost per MT delivered to markets with normal commercial (unsubsidized) sales prices, as shown in Figure 5.5. Here market prices include landed costs (as in Figure 5.4) and transport costs, but exclude overhead costs of ADMARC and SFFRFM. Unsubsidized prices are for November each year averaged across major centres in Malawi, and for 2008 and 2009 there is also data collected from markets around the country by the MoAFS. The figure suggests that if overhead costs were added on to the subsidized costs then unsubsidized market prices might be closer to the costs incurred by the programme through parastatal delivery (although if ADMARC is reaching more remote areas then higher average costs might also be incurred by private suppliers in reaching these).

FISP activities and achievements

Figure 5.4. Landed and international fertilizer prices, 2006–11

Notes: 2007/8 & 2008/9 information for only part of purchases, 2009/10 excludes ADMARC b/fwd. Urea at US$1,254/MT, 2011/12 excludes ADMARC supplies of Urea at US$864/MT & NPK at US$904/MT.

Sources: Logistics Unit reports and data, World Bank commodity price data (World Bank, 2012).

(p.99)
FISP activities and achievements

Figure 5.5. Subsidized fertilizer market costs and unsubsidized market prices, 2006–11

Notes: 2011 FAM prices are sensitive to the choice of exchange rate converting MK to US$. Prices shown are calculated with the official exchange rate, but if the widely used unofficial exchange rate in late 2011 were applied then the FAM price would be lower.

Sources: Logistics Unit reports and data, Fertiliser Association of Malawi (pers. comm.), MoAFS market reports.

The effectiveness of the programme in obtaining low-priced fertilizer supplies may also be investigated by examining the patterns of tender prices. Table 5.3 presents the range of prices paid from 2008/9 to 2011/12. Price ranges are high, although they have generally been falling and in some years (such as 2009/10) cannot be explained by wide variation in international prices during the buying period. Some suppliers appear to have been paid consistently higher prices (for example, Simama and, notably for a large urea consignment in 2009/10, ADMARC). On the other hand there is no evidence of higher prices being paid to SFFRFM (indeed the low outlier prices for NPK and urea in 2008/9 were both for supplies from SFFRFM). Price ranges remain high even after removal of outliers, but again have generally been falling since 2008/9.

5.4. Coupon distribution

We now examine the activities involved in coupon distribution, which, as shown earlier in Figure 5.1, run in parallel with input procurement and distribution activities. We again first provide an overview of systems and then consider (p.100)

Table 5.3. Fertilizer procurement prices before and after removal of high outliers

Year

Fert.

Price ($/MT)

After removal of high outliers

Outlier price

% supply

Mean

Range ($/MT)

Mean price

Range ($/MT)

% change mean

Suppliers removed

2008/9

NPK

1,453

407

1,450

348

0.2%

Mulli Bros

1,649

1.3%

Urea

1,121

313

1,090

239

0.4%

Simama

1,246

3.0%

2009/10

NPK

744

390

736

287

3.1%

Simama, Coin

886

5.1%

Urea

715

670

647

221

8.4%

ADMARC

1,246

9.2%

2010/11

NPK

791

202

779

192

1.5%

Various inc Simama

870

13.0%

Urea

650

236

645

103

0.7%

Masina Investments

830

2.5%

2011/12

NPK

824

124

817

51

0.7%

ADMARC

904

7.1%

Urea

776

200

775

135

0.2%

ADMARC

864

1.4%

Source: Author calculations from Logistics Unit Annual reports

achievements and performance over the life of the programme, drawing on information from programme records and from farm household surveys.

5.4.1. Systems

Coupon distribution to beneficiaries involves a number of activities, as set out in Figure 5.1: allocations to different areas, beneficiary identification, and then actual distribution of coupons. These activities are conditional on the quantity of subsidized inputs set in the national budget, the registration of all farm families, the printing of coupons, and then their distribution to districts and communities.

The formal process and criteria for determining budgeted quantities of subsidized inputs are not clearly documented. Starting from 2005/6 there has been an aspiration to provide enough inputs to provide a core maize fertilizer package for roughly 50% of Malawian smallholder farmers. Budget decisions for subsequent years are likely to have been made on an incremental rather than zero budgeting basis, with increases and decreases determined by political objectives, funding availability, competition for funds between ministries, and budgeted prices and costs.

Formal area allocations of coupons then involve the division of the national budget for subsidized inputs between Extension Planning Areas (EPAs) within districts. From 2005/6 to 2008/9 allocations of the total across EPAs were made in two or more rounds. In 2005/6 and 2006/7 the first allocation was reported to be proportionate to the previous year’s MoAFS estimates of maize and tobacco hectarage in each EPA, but from 2007/8 onwards EPA allocations were increasingly, and from 2009/10 exclusively, reported to be proportional (p.101) to MoAFS estimates of the number of farm families per EPA, with, again from 2009/10, annual registration of farm families by villages within each EPA.3 Within EPAs, processes and formal stakeholder roles in coupon allocations between and within villages have varied between years and areas and have also differed from actual, informal roles. Overall, they have involved Traditional Authorities (TAs), local government and MoAFS staff, Village Development Committees (VDCs), and local stakeholders identifying beneficiaries to receive coupons for redemption for different inputs at very reduced cash prices.

Printing of coupons has been funded and managed by the MoAFS and carried out in Malawi in all programme years except 2011/12 (when DFID funded printing outside Malawi). A requirement that coupons be clearly identified with unique beneficiary serial numbers by district and (in some years) EPA, has meant that printing has been conditional on district and EPA allocations. Once printed, with security features that have varied from year to year but have generally become more stringent with time, packs of coupons are bundled and distributed to each EPA, via District Agriculture Development Officers (DADOs). Coupon distribution to beneficiaries has involved the same parties as allocation, with the addition of a police presence to keep law and order and to safeguard the valuable packs of coupons.

A constant feature of programme design has been that intended recipients of maize fertilizers should each receive two coupons, to allow subsidized purchase of one 50 kg bag of basal fertilizer (23:21:0 +4S) and one 50 kg bag of urea. However, there has also been considerable variation over time and between areas in allocation and distribution processes and criteria determining selection of beneficiaries and the numbers of coupons of different types actually received per recipient.

Supplementary rounds of coupon allocations to areas and beneficiaries from 2006/7 to 2008/9 were much more opaque as regards systems, criteria, and numbers of coupons distributed. These were nominally intended to address problems of unmet demand in the first round of distribution but were widely considered to allow opportunities for often politically driven allocations outside formal systems.

5.4.2. Total coupon receipts

Two sources of information on coupon distribution and receipt can be compared: programme records on coupon disbursement, and farm household survey estimates of coupon receipt. These can also be compared with input sales records on coupon redemption. There is, however, a major difficulty (p.102) with this as the estimation of total coupon receipts from farm household survey data requires the multiplication of mean receipts per household (estimated from the survey) by the total number of farm households—but there are major discrepancies between the estimates of the number of rural households based on the National Statistical Office (NSO) 2008 census on the one hand and MoAFS estimates of farm families on the other. The extent of these discrepancies is shown in Figure 5.6.

These differences may have a number of causes:

  1. a. the definitions of farm families and of rural households may be different, with a larger number of smaller-farm families;

  2. b. MoAFS processes may lead to households splitting for the purposes of registration, to increase the number of coupons that they may be able to receive;

  3. c. ‘ghost’ families and villages may have been created by corrupt field staff or traditional leaders to enable them to obtain coupons;

  4. d. the NSO may have missed some households in the census;

  5. e. the NSO rural households listing excludes all households residing in cities or in urban areas, but many urban households are also engaged in agriculture and may receive coupons.

It is possible that the discrepancies in Figure 5.6 result from all these causes. Examination of the figure certainly suggests that there has been some ‘inflation’ in MoAFS farm family estimates. The central region appears to have had a remarkable growth in farm families from 2005/6 to 2009/10 (reaching 20% in 2009/10), with numbers then flattening off and actually declining from 2010/11 to 2011/12. Combined with a spurt in growth in the southern region from 2008/9 to 2009/10, this resulted in an estimated increase of 14% in Malawian farm families in 2009/10. Northern region

FISP activities and achievements

Figure 5.6. MoAFS farm families and NSO rural households

(p.103) growth rates were also high from 2005/6 to 2007/8 and then flattened off. These regional and temporal variations pose questions regarding the reliability of the farm family estimates, and suggest that the NSO estimates may be more reliable—although they may also exclude some urban farm households.4

Table 5.4 compares estimates of coupon issues and receipts from these different sources. Higher MoAFS farm family estimates lead to higher estimates of total coupon receipts, exceeding MoAFS formal allocations in 2008/9 and 2010/11 (although there may have been substantial informal supplementary allocations in 2008/9, this was not the case in 2010/11 when the excess is greatest). Lower NSO-based rural household estimates, however, lead to lower estimates of total coupon receipts, considerably below MoAFS formal allocations in all three years.

Following the discussion above of Figure 5.6 and the use of NSO rural household definitions in the farm household estimates of coupons received per

Table 5.4. Estimates of fertilizer coupon issues and receipts from different sources5

2006/7

2008/9

2010/11

A. Households receiving one or more coupons

54%

67%

79%

B. Coupons received per recipient household

1.7

1.5

1.4

C. Coupons received per household (all households)

1.01

1.12

1.13

D. Estimate of total coupons received, NSO based rural household estimates (millions)

2.51

2.79

2.73

E. Estimate of total coupons received, MoAFS farm family estimates (millions)

3.32

4.11

4.42

F. MoAFS voucher allocation (millions)

3.48

3.91

3.20

Discrepancy as % of missing coupons as % MoAFS allocations

G. Using NSO rural household estimates (D as % of F)

28%

28%

15%

H. Using MoAFS farm family estimates (E as % of F)

5%

−5%

−38%

Sources: 2006/7, 2008/9, and 2010/11 household surveys; Logistic Unit reports; author calculations from National Statistical Office (2008a) and data supplied by MoAFS.

(p.104) household, we may expect more reliable estimates of total coupon receipts using the NSO rural household estimates. This suggests that a total of 3.3 rising to 4.4 million fertilizer coupons were distributed to rural households in the 2006/7, 2008/9, and 2010/11 seasons. We discuss later the implications of this for estimates of diversion and fraud in the programme.

Comparable analysis for seed coupons is difficult, as a result of changing systems and quantities of seed coupon allocations. However, 2008/9 and 2010/11 survey estimates of coupon receipts per household multiplied by estimated rural household numbers give total coupon receipt estimates that are close to MoAFS records of total allocations.

5.4.3. Coupon distribution processes and performance

Section 5.4.1 has described broad processes for coupon distribution. In this section we examine these in a little more detail, comparing formal procedures adopted by the programme with those actually reported in 2006/7, 2008/9, and 2010/11 farm household surveys and focus group discussions. We consider how procedures and criteria for coupon allocation and distribution have changed, the timing of these activities, their outcomes in terms of the numbers of coupons distributed, and perceptions of these processes.

Formal instructions for coupon allocation and distribution have specified who should be involved and procedures to be used. These instructions have changed over time in response to both difficulties and complaints on the one hand and examples of success on the other. Thus, in 2005/6 Traditional Authorities (TAs) were tasked with allocating coupons between villages within EPAs, and Village Development Committees (VDC) with allocating coupons between households within villages. Complaints that some TAs and VDCs were subverting allocations led to instructions in 2006/7 that allocations be made by district, area, and village committees in accordance with standard local government structures, with detailed terms of reference and membership for each committee. Beneficiaries, were supposed to be ‘full time smallholder farmers who cannot afford to purchase one or two bags of fertilizer at prevailing commercial prices, as determined by local leaders in their areas’. Guidelines also specified that coupons should be issued to beneficiaries ‘just before they go to a market point to purchase inputs, to minimize chances of abusing them’. (p.105)

In 2007/8 programme objectives and formal targeting criteria were amended to give greater emphasis to concerns for vulnerable households. Systems for allocation and distribution of coupons within districts were also modified to give less power to TAs and more responsibility to MoAFS staff, following support from communities to disburse vouchers through MoAFS staff following the 2006/7 programme experience. ‘Flexi-seed’ vouchers were also introduced, allowing households who were not given a set of vouchers for the maize seed and fertilizer package to nevertheless benefit from subsidized maize, legume, or cotton seed. This was not, however, particularly popular as there were very limited supplies of legume seed, cotton seed was wanted by relatively few farmers, and many farmers saw it as a poor substitute for receipt of fertilizer coupons. Flexi-seed coupons were discontinued in 2009/10 and replaced by specific legume seed coupons.

In 2008/9 all farm families were registered and the list of registered farm families was used in subsequent identification of beneficiaries in ‘open village meetings’ led by teams involving MoAFS staff, local government staff, religious leaders, VDC members, and civil society representatives. Beneficiary lists were compiled by village and EPA and district. These lists were then checked by the Logistics Unit against allocations before distribution registers were printed with beneficiary names by village and delivered to the MoAFS. Similar procedures were followed from 2009/10 to 2011/12 except that beneficiaries were also supposed to have voter ID cards from the 2009 elections and there were further increases in emphasis on beneficiary targeting to focus on more vulnerable households—emphasizing child-headed, female-headed, or orphan-headed households, those infected or affected with HIV and AIDS, and guardians or carers of vulnerable people (all of whom should be also be Malawians owning land).

Key informant interviews, information from community questionnaires, and anecdotal evidence suggest that a wider variety of coupon allocation and distribution procedures has been followed than would be expected from consistent implementation of the formal instructions. Thus, in 2006/7 the involvement of the DDC (District Development Committee) and VDC was almost, but not quite, universal with, for example, in one surveyed district only one committee level under the DDC organizing bulk transport of fertilizers for farmers. The extent and nature of involvement of the TAs in the process varied widely, and was strongest in the central region where there were reports of TAs subverting the process in various ways. Targeting criteria reported also varied markedly between districts with, for example, different emphases on farmers’ inability to otherwise afford fertilizer purchase.

Wide variations in practice were also associated with the introduction of open meetings from 2008/9. First the extent of their adoption varied, with 81 (p.106) and 96% of respondents reporting open meetings for allocation and distribution of fertilizer coupons in 2008/9, and similar figures for 2010/11. Where open meetings were adopted, some FGDs described them as meetings where pre-determined beneficiary lists were announced, while others reported actual participatory selection of beneficiaries. Some respondents recognized their value in informing people about the programme and consequently helping to reduce struggles and conflicts, but perceptions also appear to be conditional on perceptions about changing numbers of available coupons compared to the previous year.

As regards the second, supplementary allocation of coupons, a large proportion of respondents reported that these were distributed based on the choices of traditional leaders and, in other cases, by politicians mainly targeting their supporters and party sympathizers. However, Chinsinga (2012b) documents reports of some more general political interference in coupon allocation.

A significant and widely practiced variation from official procedures has involved ‘redistribution’ of coupons. Coupon allocations should lead to all beneficiaries getting two coupons for ‘maize fertilizer’ (one for NPK and one for urea). However, substantial numbers of respondents report receipt of one coupon (or even the proceeds of a part of a coupon). This is also reported by Holden and Lunduka (2010b) and we discuss this further with regard to targeting outcomes in Chapter 10.

We now turn from a description of coupon allocation and distribution processes to consideration of performance in these activities as regards timing and respondents’ perceptions of effectiveness. Discussion of diversion and fraud in coupon allocation is deferred to its own section later in this chapter, while targeting is considered later in Chapter 10.

Information on the timing of beneficiaries’ receipt of coupons is available from the 2006/7, 2008/9, and 2010/11 household surveys. This is summarized across all regions in Figure 5.7 and shows a clear improving trend over the three surveys, but with only 40% of beneficiaries receiving their coupons by the end of October in 2010/11 (70% in the south, 11% in the centre, and none in the north) there is still need for considerable improvement. Community survey respondents reported similar patterns.

Survey respondents have also been asked to rate different aspects of programme implementation from very good to very bad. In Table 5.5 scorings on timing show small but steady annual improvements in perceptions of timing since the start of the programme, with overall ratings moving from ‘bad’ to ‘not good, not bad’ from 2005/6 to 2010/11. Ratings of methods of coupon distribution and of criteria for coupon distribution have not changed much, and are broadly ‘not good, not bad’. Perceptions of numbers

FISP activities and achievements

Figure 5.7. Percentage of beneficiaries reporting receipt of maize fertilizer coupons by end of each month

(p.107) of coupons distributed have been declining. This may result from changes in the average number of coupons received per recipient (see Table 5.4) as a result of the increasing redistribution of coupons discussed earlier. Holden and Lunduka (2010b) find that 28% of respondents in 2008/9 reported ‘insufficient coupons’ as the main problem with the programme, while 42% considered corruption to be the main problem. However Holden and Lunduka (2010a) also report that 31% of respondents from the same survey suggested that more or enough coupons should be supplied, with 20% suggesting that the coupon distribution system need to be improved, 10% each suggesting providing more inputs and introducing a general fertilizer subsidy, and 9% suggesting that chiefs be removed from involvement in coupon distribution.

The final issue to be discussed on coupon distribution is the extent to which farmers had to pay for coupons. We may expect some under-reporting of this. Five per cent of fertilizer coupons were reported as being obtained with some payment in both 2006/7 and 2008/9 (this is lower than the estimate of 14% for 2008/9 survey reported by Holden and Lunduka (2010b)). Reported sources of such fertilizer included TAs and headmen, agricultural staff, VDC members, traders, and fellow farmers. A much lower figure of 2% was found in 2010/11. Reported prices varied dramatically, with medians of 600MK per coupon in 2006/7, 2000MK in 2008/9 (when prices for unsubsidized fertilizer were very high) and 1000MK in 2010/11 (Holden and Lunduka (2010b) reported median prices of 1500MK for 2007/8 and 2500MK for 2008/9). (p.108)

Table 5.5. Scoring on different programme elements by year

2005/6

2006/7

2007/8

2008/9

2009/10

2010/11

Timing of distribution

2.73

2.73

2.56

2.01

2.20

1.93

Methods of coupon distribution

2.89

2.94

2.97

2.81

2.69

2.74

Criteria for coupon allocation

2.95

2.92

2.83

2.79

2.83

Number of coupons

2.96

3.02

3.07

3.16

3.20

3.36

Scores: 1 = very good; 2 = good; 3 = not good not bad; 4 = bad; 5 = very bad.

5.5. Coupon redemption

As with input and coupon distribution, we consider coupon redemption with an overview of systems and then consider achievements and performance over the life of the programme.

5.5.1. Redemption systems

Redemption of coupons for subsidized inputs essentially involves the presentation of a valid coupon and any necessary payment to an authorized retailer who, after checking the validity of the coupon, issues the specified input. Three main systems have been used over the life of the programme: redemption of fertilizer coupons by parastatals (SFFRFM and ADMARC), redemption of fertilizer coupons by approved private retailers in the 2006/7 and 2007/8 seasons, and redemption of seed coupons by parastatals or private retailers from 2006/7.6 We describe core features of each of these systems: a description of more detailed changes between years is provided in Section 5.5.3.

Under the parastatal system for fertilizer redemption, fertilizer stocks are distributed to ADMARC and SFFRFM markets (some of which are specially opened during the input supply season) as described earlier. Beneficiaries then present their coupons with the required farmer payment and the parastatal issues the fertilizer. ADMARC and SFFRFM were then supposed to return to the government both the money and coupons received from farmers. The (p.109) extent to which coupons have been returned has varied between years, while information on reimbursement of farmer payments is not available. From 2006/7 onwards fertilizer subsidy sales information has been reported by the Logistics Unit on the basis of stock reconciliations and, from 2006/7 to 2010/11 on the basis of weekly market reports.

Redemption of fertilizer coupons by approved private retailers in the 2006/7 and 2007/8 seasons was identical to redemption of coupons by parastatals as regards transactions with farmers. However, reimbursement for sales (of the retailer’s own stocks) was obtained by submission of invoices to the MoAFS supported by redeemed coupons, with reimbursement per coupon at a previously contracted price covering subsidy costs (with retailers keeping farmer payments as the unsubsidized part of the overall payment).

The seed coupon system has been based on annual agreements between the MoAFS and seed supply companies in the Seed Traders Association of Malawi (STAM). Under these agreements individual companies contract with different retailers to supply them with seed approved for subsidized sale, and then to receive from these retailers the details of subsidized sales, with supporting coupons. The seed companies subsequently invoice the MoAFS for each seed coupon returned to them as a result of its use in purchasing their seed. The distribution between retailers and seed supply companies of MoAFS payments and of any farmer payments are a matter for negotiation between retailers and seed supply companies and are not a matter for MoAFS.

An important issue in coupon redemption is the setting of redemption fees to be paid by farmers. As indicated earlier in Table 5.1, while fertilizer prices have risen over the life of the programme, farmer contributions through the redemption price have fallen by almost 50% in nominal terms (more in real terms), from 950MK per bag to 500MK/bag for ‘maize fertilizers’, and the subsidy has therefore increased from around 65% to over 90%. Fertilizer redemption prices have been politically determined, generally announced by the president in political statements or at rallies. There is little evidence of any technocratic involvement in the setting of these prices, which are very low: when a price of 500MK/bag was announced for the 2012/13 season it was widely criticized as a contributor to high and unsustainable and distortionary programme costs undermining budgets for other MoAFS activities. Politics has generally played a smaller role in the setting of seed prices, which have been negotiated by seed companies and the MoAFS. However in 2007/8 agreement that farmer redemption payments (‘top ups’) of up to 100MK could be charged per 2 kg pack of hybrid seed had to be abandoned following a political statement that all subsidized maize seed was free. Seed redemption prices are less important than fertilizer prices, both politically and in their impact on the programme budget. (p.110)

FISP activities and achievements

Figure 5.8. Subsidized fertilizer and seed sales by year

Source: Calculations from Logistics Unit (2011) and earlier reports.

5.5.2. Total redemption

Figure 5.8 shows the total volumes of fertilizer and seed redemptions by year. Subsidized fertilizer sales increased from 2005/6 to 2007/8, and were considerably over budget from 2006/7 to 2008/9 (with budgeted fertilizer sales of 150,000 MT in 2006/7 and 170,000MT in 2007/8 and 2008/9). The lower sales from 2009/10 were almost exactly on budget. Maize seed sales increased from 2006/7 to 2010/11 but fell back in 2011/12. Hybrid maize seed sales increased dramatically from 2007/8 to 2009/10, with reduced sales of OPV seed, but OPV sales then grew again in 2010/11 and 2011/12—with hybrid seed sales hardly growing in 2010/11 and then declining in 2011/12. There were also dramatic increases in legume seed sales in 2009/10 and 2010/11, with 2011/12 sales a little lower than in 2010/11.

5.5.3. Coupon redemption performance

Apart from the introduction of private sector seed and fertilizer sales in 2006/7 and the withdrawal of private sector fertilizer sales in 2009/10, the only major changes in redemption processes between years have been variation in redemption payments required from farmers. These are set out in Table 5.6. Variation in pack sizes for hybrid and OPV seed packs and in farmer top up payments required for different varieties led to some competition and farmer choice, where different seed stockists accessible to beneficiaries carried different or a range of stocks.

Performance in coupon redemption can be considered in terms of availability of inputs for redemption, and timing and costs of redemption. These are all related. (p.111)

Table 5.6. Coupon redemption parameters, 2005/6 to 2011/12

Fertilizers top-up payment (MK per 50 kg)

Maize seed pack

Legume seed

Size (kg)

Top up farmer payment (MK)

Redemption value (MK)

Size (no farmer top up)

Redemption value (MK)

2005/6

*

OPV pack size

n/a*

n/a*

-

-

2006/7

950

2 kg hybrid seed or 3 or 4 kg OPV seed

0

400

-

-

2007/8

900

2 kg hybrid seed or 4 kg OPV seed

0**

400

2 kg

400

2008/9

800

2 kg hybrid seed or 4 kg OPV seed

0

680

2 kg

680

2009/10

500

5 kg hybrid seed or 10 kg OPV seed

〈=100MK

1500

2 kg

350

2010/11

500

5 kg hybrid seed or 7.5 kg OPV seed

〈=100MK

1650

2 kg

740

2011/12

500

5 kg hybrid seed or 10 kg OPV seed

〈=100MK

1815

0MK for 2 kg

815

(*) 950MK per bag for 23:21:0 and urea, 1450MK per bag for Compound D and CAN.

(**) Farmer top ups agreed initially but subsequent political announcements led to confusion and inconsistent payments.

Lack or limited availability of inputs when needed has been a major difficulty that has frequently delayed purchase and use of subsidized inputs. We have discussed in earlier sections how for fertilizers this has been caused by late opening of seasonal markets, late distribution of inputs to markets, and late distribution of coupons to beneficiaries—and how these three factors interact—as late market opening and coupon distribution makes timely input distribution to markets more difficult, and hence fertilizer stock-outs more likely. Farmers and sellers have no flexibility in substituting between types of fertilizer since coupons are specific to each type of fertilizer. With seeds, however, there has been more flexibility, with opportunities to exchange maize seed coupons for different varieties of hybrid and OPV seed, if different stocks are available. (p.112)

Table 5.7. Reported distances to buy inputs, time spent buying inputs, and costs for transport and miscellaneous expenses

Hours travel & waiting

Transport and misc expenses (MK)

Distance to nearest ADMARC (km)

Distance to nearest private selling point (km)

Mean

Median

Mean

median

Mean

Median

Mean

Median

2006/7

13

7

247

150

7

5

7

5

2008/9

17

9

304

200

9

5

14

8

2010/11

23

12

270

200

5

4

8

6

Stock-outs have been relatively common in all years of the programme, have affected all inputs, and in 2006/7 were experienced by both private and parastatal retailers.7 However when they manage to get their inputs farmers are generally pleased with the fertilizers they get, but are sometimes disappointed by lack of availability of hybrid maize seed and specific types of legume seed. We discuss access to maize seed further in Chapter 8.

Stock-outs lead to increased time and cost for farmers travelling to and from and waiting at markets, and while it appears that hours travelling and waiting have increased over the life of the programme there is no clear evidence of changes in transport and other costs or in distances to markets (see Table 5.7).

Stock-outs also provide sales clerks with opportunities to demand extra payments for access to limited stocks. This is a problem that is widely reported in the media and in focus group discussions and key informant interviews. Table 5.8 provides survey estimates of the extent and nature of these extra demands.

Table 5.8 shows no clear trend in the frequency or size of extra payments. In 2006/7 the percentage of coupons requiring tips and mean redemption price were lower at ADMARC/SFFRFM than at all other distributors, but this masks considerable variation between different distributors, with some having a slightly lower reported incidence of tipping than ADMARC/SFFRFM. In 2008/9 and 2010/11 community survey respondents, focus group discussions, and key informant interviews all suggested that the payment of bribes to redeem inputs was more widespread than is suggested by the survey results, with focus groups suggesting that this was closely related to problems of queuing. Chinsinga (2009) also reported extra payments of 200MK to 800MK per bag of fertilizer being demanded by some ADMARC staff (with (p.113)

Table 5.8. Reported extra payments for coupon redemption (MK/bag fertilizer)

Year

Retailers

% coupons paid tips

Median price

Mean price

Mean extra payment

Median extra payment if paid

2006/7

Private

27%

950

1,223

273

n/a

2006/7

Parastatal

18%

950

983

33

n/a

2008/9

Parastatal

14%

800

827

27

200

2010/11

Parastatal

9%

500

536

36

250–500

those unable to pay being required to wait two or three days before they were served); extortion of cash from beneficiaries by criminal elements to ‘facilitate’ input acquisition; and organized theft through tricking farmers. The Farmers Union of Malawi (2011) report that 5% of their sample of registered beneficiaries reported being asked to pay bribes for input redemption although 42% considered it common or very common to be asked for such a bribe.8 However, only 20% of those asked for a bribe reported that they had paid it. The Farmers Union of Malawi (2011) FGDs also reported that women were particularly vulnerable to these demands

It is difficult to determine if extra payments made for hybrid seed are corrupt, as extra payments have been supposed to be made for some hybrid varieties in some years. However, no payments should have been required for OPV or legume seed, but 12% and 11% of respondents reported making some payment for OPV and legume seed in 2010/11, respectively.

5.6. Diversion and fraud

Diversion of subsidized inputs away from intended beneficiaries is an issue that has been raised in Chapters 2, 3, and 4. Diversion takes several forms, from inclusion errors in targeting (whereby inputs go to smallholder farmers who were not supposed to get them) through leakages (for example, as a result of sales of vouchers or subsidized inputs at low prices) to outright fraud and corruption, where people capture subsidized inputs as a result of direct criminal activity or as a result of political influence (and of course illegal political capture is itself criminal). We make a distinction here between mis-targeting within the smallholder sector and diversion outside the smallholder sector. (p.114) Targeting is the subject of Chapter 10. We consider here evidence regarding the extent of leakage or diversion outside the smallholder sector.9

Illegal activities are inevitably clandestine and difficult to trace. There have been numerous media reports of cases of theft and fraud regarding FISP coupons and inputs (Tambulasi, 2009), successful prosecutions, and on-going police, audit, and Anti-Corruption Bureau investigations. There are also frequent allegations of fraud by village officials and extension agents in focus group discussions and key informant interviews. It is, however, very difficult to obtain objective and comprehensive estimates of the extent of these activities. The lack of transparency in coupon allocation when combined with excess demand for coupons leads to perceptions of and complaints about corruption and diversion of coupons, and this may occur even in situations where these perceptions and complaints may not be warranted. In addition, while there may be many cases that are not discovered, acute interest in those cases that are discovered may lead to over-estimates of their occurrence and scale.

Fraud may arise in a number of ways: through voucher allocation to non-existent (‘ghost’) beneficiaries (or villages), with theft of the allocated vouchers by Ministry of Agriculture staff and/or Traditional Authorities, diversion of vouchers to people with power (government staff, traditional leaders, or politicians) or to criminals for sale, direct allocation to non-beneficiaries, and printing of extra or counterfeit vouchers within or outside the system. There may also be thefts of inputs. Another type of corruption is operated by some who act as gatekeepers, with village headmen or sales clerks demanding bribes (extra payments or services) in exchange for coupons or coupon redemption which beneficiaries are entitled to without such bribes. This last form of corruption was discussed in Sections 5.4.3 and 5.5.3.

Although there should be control systems that prevent or at least monitor these different types and sources of fraud, the existence and operation of these systems have not always been clear, or, as with some audits, their implementation has been slow and a full picture has still to emerge. The extent and investigation of diversion has also been affected by changing political conditions, as discussed in Chapter 4.

The divergence in estimated number of households between the NSO census and the MoAFS farm registry (discussed in Section 5.4.2) poses difficulties in using household surveys to investigate this issue. For example, with the NSO farm family estimate it appears that 2.7 million fertilizer coupons were received by smallholder farmers in 2010/11 against a recorded allocation of 3.2 million, leading to an estimate of 0.5 million ‘missing’ coupons (15% of those issued, compared to 28% in 2008/9). However, with the MoAFS farm (p.115) family estimate it appears that 4.4 million fertilizer coupons were received by smallholder farmers in 2010/11 against a recorded allocation of 3.2 million, leading to receipts exceeding issues by 38% (compared to 5% in 2008/9). These difficulties are compounded by the lack of a nationally representative sample for the 2010/11 FISP survey. However, they do suggest that if the NSO figures are taken as being closer to the to the true population with the NSO (as opposed to MoAFS) household definition (as suggested by earlier examination of the differences in regional changes in MoAFS figures and by the survey using NSO households for sampling and the NSO household definition) then high losses in 2008/9 have been substantially reduced in 2010/11.

We can, however, go further than this to make some very broad estimates of leakage and diversion of subsidy vouchers and inputs outside the smallholder sector. The basis of these is the diagram of possible voucher and subsidized input flows in Figure 5.9. The diagram shows on the left hand side the legitimate flows of vouchers and inputs within the smallholder sector, and on the right the illegitimate flows of vouchers and inputs outside the smallholder sector. Boxes in bold represent variables for which we have estimates from household surveys or administrative records (although estimates for smallholders’ cheap purchases are not available for 2006/7). This then allows some estimation of the scale of diversion first of vouchers (by comparison of estimates of voucher receipts and use by smallholders against programme voucher redemptions) and then of subsidized seed and fertilizers (by comparison of estimated input acquisition and use by smallholders against total programme supply). These estimations are prone to error, first as a result of uncertainty regarding the total number of farm households (as discussed earlier) and second as a result of making estimates by subtraction, which tends to magnify percentage errors. Nevertheless, these estimates may be taken to provide some broad indications of the changes in the scale of diversion in the programme. These are set out in Table 5.9 for the 2006/7, 2008/9, and 2010/11 years when survey data were available. As in Section 5.4.2, we take the NSO-based estimates of the number of rural households to be most reliable and consistent with the sample surveys.

The results of this analysis are presented in Table 5.9 (capital letters in each row refer to the framework in Figure 5.9). This suggests that there were very high leakages or diversions outside the smallholder sector in 2006/7 and 2008/9, with smallholders receiving only 63 and 68% respectively of recorded coupon issues. Despite their smaller share of disbursed coupons in 2006/7, it appears that smallholders redeemed a higher percentage of their coupons as compared with others, and as a result bought a greater share of the subsidized fertilizers than in 2008/9. Of the fertilizers bought by others under the subsidy in 2008/9, however, a large proportion (about 50%) appears to have been sold to smallholders at lower than unsubsidized prices, (p.116)

FISP activities and achievements

Figure 5.9. Flows of coupons and subsidized inputs

so that the smallholder sector ended up using about 80% of the subsidized supplies—though a little over 25% of this was not purchased with the full subsidy. This is broadly comparable with Holden and Lunduka (2010b) who estimate that in 2008/9 21% of total household fertilizer use was obtained through cheap purchases of fertilizer (and 25% through purchased coupons). In 2010/11, however, smallholder farmers received a higher, but still not high enough, proportion of voucher issues (85%) and it is also estimated that in addition almost all the diverted subsidy fertilizer sales were then resold to smallholders (again at prices between full and zero subsidy). If smallholders are assumed to have received around 50% of the subsidy when buying originally subsidized fertilizers from other sellers then the proportion of fertilizer subsidy captured by others fell from a little under 30% of the subsidy in 2008/9 to a little over 10% in 2010/12. Since the subsidy volume was approximately 25% higher in 2008/9, and the price of fertilizers was approximately 65% higher, the estimated loss through diversion fell by just under 80% from 2008/9 to 2010/11.10 (p.117)

Table 5.9. Estimated shares of coupon and subsidized fertilizer receipts by smallholders and others1

2006/7

2008/9

2010/11

Coupons % recorded issues

Recorded issues (from MoAFS and Logistics Unit)—A

100%

100%

100%

Received by smallholders, exc. purchases—B

63%

68%

85%

Received by others (by subtraction)—A-B = C-X

37%

32%

15%

Redemptions

Total (from Logistics Unit)—I + J

87%

102%

100%

Smallholders—I

63%

69%

78%

Others (by subtraction)—J

25%

34%

22%

Counterfeits/extras redeemed (min.)

-13%

2%

0%

Fertilizers % subsidy sales

Total subsidy sales (from Logistics Unit)—I + J

100%

100%

100%

Smallholder redemption & use—N

72%

58%

78%

Others’ redemption—J by subtraction

28%

42%

22%

Smallholder low price purchases—O

n/a

21%

21%

Total smallholder use—N + O

n/a

79%

99%

Others’ low price/redemption use—P + Q

n/a

21%

1%

(1) Smallholder low price purchases (O) are estimated as all purchases below a cut-off price. In 2008/9, there was a much greater range of fertilizer prices reported by respondents, and the cut-off point for ‘low price’ purchases was set at 150MK/kg, around 70% of the standard commercial price. For 2010/11 examination of the data suggested a cut-off point of 85MK/kg, around 85% of the standard commercial price. However some of the higher price purchases could be for stock from the subsidy if some traders bought subsidized fertilizers and sold it as if it were commercial, particularly when selling in smaller quantities. Allowance for this might suggest somewhat higher reselling of subsidized purchases than estimated above, in which case estimates of ‘others low price redemption and use (P + Q)’ would fall. However, the broader conclusions of this section would not be significantly changed—indeed some would be strengthened.

Source: Author calculations from survey data.

Despite the apparently precise estimates presented here, these should be taken as illustrative—sampling error will affect survey estimates and, as discussed earlier, there is uncertainty regarding the number of farm families and the subtractions in the calculations may magnify errors. Nevertheless, the broad conclusion that there were substantial reductions in the extent and value of diversion of subsidized fertilizer from 2008/9 to 2010/11 is likely to be robust. While it appears that there is still substantial diversion of fertilizer subsidy benefits away from smallholders, it is instructive to compare this with, for example, the UK Income Support Allowance, which in 2009/10 suffered from estimates of over-payments (estimates which are unlikely to be complete) of a little over 5% as a result of fraud and error (Department for Work and Pensions, 2012).

These conclusions regarding levels and changes in diversion and fraud are compatible with the continued reports of variable but widespread diversion of fertilizer coupons in rural areas (by government staff, TAs, headmen, and VDC members). The apparent reduction in diversion of coupons and fertilizers that never reach rural areas (‘Others’ low price/redemption use—P+Q’ (p.118) in Table 5.9 is dramatically reduced in 2010/11) is also compatible with a reduced and changed political significance and role of the programme after the 2008/9 election, as mentioned earlier in Chapter 4. Further information on these matters may become available in the future as ongoing audit investigations continue (but these tend to be slow with, for example, media reports in 2012 regarding audit investigations of the 2007/8 and 2008/9 programmes) and with the change in government in 2012 opening up the political space for investigations of prior fraud.

5.7. Farmer support

Proper use of subsidized seed and fertilizer is an important determinant of the impact of the FISP, and this would suggest that access to extension advice on use of fertilizer and improved seeds should be an important component of the programme. The household surveys in 2006/7, 2008/9, and 2010/11 do not show any consistent changes by year but between 15% and 20% of farmers (beneficiaries and non-beneficiaries) received extension advice on fertilizers and maize varieties, with advice on average being considered moderately useful. A greater proportion of farmers in the north received advice than in the other two regions, while a somewhat lower proportion of female-headed households and poorer households receiving advice (as compared with male-headed households and less-poor households).

5.8. Programme finance and costs

The final major implementation issue that we address in this chapter is the cost of the programme. We consider overall programme cost to provide a measure of fiscal resources used by the programme and of its sources. Examination of the cost breakdown and comparison of actual and budgeted costs provides further information on resource use and financial management.

Constructing a consistent set of financial information about the programme over the period 2005/6 to 2011/12 faces a number of difficulties:

  1. 1. There are difficulties with the treatment of farmer payments for fertilizer sold though ADMARC and SFFRFM. For inputs supplied through the private sector, the government reimburses the supplier at an agreed rate which is calculated as the cost of input supply less farmer payments. However, for fertilizers supplied through ADMARC and SFFRFM, the government bears the whole cost of supply and requires reimbursement of farmer payments from ADMARC and SFFRFM. There are no publicly available accounts of this, though anecdotal information suggests that (p.119) SFFRFM has been repaying these monies in full for a number of years while ADMARC only began to increase its payments in later years.

  2. 2. There are substantial gaps in reported costs in different years, leaving analysis with decisions to make about estimating these in some years but not others, and with doubts about their completeness when they are included.

  3. 3. Fertilizers may be bought one year and carried over to the next as stock. These costs have been handled in different ways in different years, and when there are large price rises and falls between years these pose difficulties for allocating costs between years.

  4. 4. There are substantial programme costs that do not appear in any accounts—for example, the costs of extension staff time allocated to the programme, for at least three months of the year, are not included in any financial costs for the programme.

  5. 5. Finally, and expanding on the previous two points, there are differences between annual fiscal accounts of the programme on the one hand (important for consideration of the programme’s fiscal impacts) and economic analysis of the programme’s opportunity costs. While this difference may be conceptually clear, this is often not the case in practice.

As a result of these practical and conceptual difficulties in recording programme costs, different statements of aggregate costs may be found in different contexts.

Our best estimates of programme costs from 2005/6 to 2011/12 are set out in Table A5.1. This provides some detail on different information available and unavailable in different years and is the basis for information presented in Figure 5.10.

There are a number of points of interest in Figure 5.10. First, we note how programme costs appeared to be increasing exponentially from 2005/6 to 2008/9, while the programme budget was supposed to rise more slowly and steadily. This led to an increase in estimated expenditure from US$74 to US$250 million from 2006/7 to 2008/9, with the programme accounting for 68% and 16% of the MoAFS and national budgets and 6.6% of GDP in 2008/9. However, while the 2009/10 budget continued to rise, 2009/10 actual programme costs fell dramatically below the budget, and although costs rose again above the budget in 2010/11 and 2011/12 they remained below half of the 2008/9 levels in relation to the national budget and GDP (at or below 8% and 3% respectively). The major cost item causing programme cost changes is clearly the cost of fertilizer (which here includes procurement and transport to markets), as seed and other costs are responsible for a small proportion of programme costs and show a steady increase (p.120)

FISP activities and achievements

Figure 5.10. Budgeted and actual programme costs

Sources: see Table A5.1.

over the life of the programme (due to increases in seed volumes and prices and greater reporting of other cost items). Fertilizer costs are then the product of volumes and prices, as shown earlier in Figures 5.4,, 5.5, and 5.8, respectively. Figure 5.8 shows that fertilizer volumes increased dramatically from 2005/6 to 2007/8, and then fell back somewhat in 2008/9 (though they were still above 2006/7 volumes) before being rigorously cut and held to budget from 2009/10 onwards. The major cause of the price spike in 2008/9 was therefore high international fertilizer prices, as clearly shown in Figures 5.4 and 5.5. The pattern of rising fertilizer volumes from 2005/6 to 2008/9, the commitment to these volumes despite very high prices and costs in 2008/9, and then the rigorous cut in volumes after 2008/9 have to be seen in the context of the changing political circumstances before and after the May 2009 presidential and parliamentary elections, as discussed in the previous chapter.

5.9. Summary

This chapter has set out the main features of the evolving design and implementation of the Farm Input Subsidy Programme from 2005/6 to 2011/12. This has emphasized the scale and complexity of the programme, and ways (p.121) in which design and implementation have evolved in response to lessons learnt and, importantly, changing political conditions. Key programme elements as set out in Chapter 2 concern its stated food security and productivity objectives, its focus on producers of staple crops, the importance of political commitment to its establishment and operation, and the consequent effects of that on targeting, rationing, and diversion. Other ongoing challenges are noted, with regard to timeliness of input delivery to farmers and involvement of the private sector in the fertilizer supply system, although cost comparisons for private and public sector distribution are not clear cut. The changing fiscal costs of the programme illustrate both the political pressures for programme continuation and expansion (and resource diversion), and the potential for scaling the programme back when political conditions and strategies change. Implementation of the programme has clearly been affected by continuation of the political considerations that led to its establishment, as discussed in Chapter 4. Chinsinga (2012b) reports considerable frustration among senior MoAFS staff at the way that the president’s political interest in the programme inhibited discussion and implementation of technical suggestions for programme improvement. These sensitivities also affected relations between donors and programme management within MoAFS and engagement with monitoring and evaluation activities. The high costs in 2008/9 also highlight the vulnerability of the programme to international fertilizer prices (a vulnerability that is again likely to be evident in the 2012/13 season, following the major devaluation of the Malawi Kwacha in 2012).

The programme shares many of the design and implementation features of other recent subsidy programmes in Africa, as reviewed in Chapter 3. As regards objectives, there is a focus on producer gains and on household and national food self-sufficiency, with a lack of regard for consumer interests and wider growth processes. There is also limited attention to improving input efficiency, soil fertility replenishment, and fertilizer supply system development. This focus is consistent with switching attention to monitoring and evaluation and auditing over the life of the programme and associated with the political management of the programme by the president. It is also consistent with a relative lack of integrated attention paid both to complementary investments that will raise programme efficiency and to ‘graduation’, the process by which beneficiaries and the economies in which they are embedded may grow out of the need for agricultural input subsidies.11 As regards implementation, the programme involves heavy reliance on private sector (p.122) fertilizer imports and a substantial targeted and rationed subsidy distributed using vouchers, but limited success in household targeting. It also faces the widespread problem of late delivery of inputs to beneficiaries.

These issues are the subject of Chapters 6 to 11 which consider beneficiary, wider and private sector impacts of the programme, targeting, and graduation.

Table A5.1. Estimated programme costs, 2005/6 to 2011/12 (US$ million)

2005/6

2006/7

2007/8

2008/9

2009/10

2010/11

2011/12

Exchange rate, MK/US$

140.00

140.00

140.00

140.00

141.31

151.55

166.71

Recorded costs

Seeds—flexi/legumes

0.00

0.00

1.89

5.24

2.83

6.66

6.26

Seeds—maize

0.00

5.23

4.58

7.33

17.13

21.64

15.12

Cotton chemicals

0.00

0.00

0.24

n/a

0.00

0.00

n/a

Fertilizer b/f from y-1

0.00

0.00

11.82

24.88

35.17

0.00

0.00

Fertilizer—new supply

51.62

61.16

77.60

237.63

57.18

115.28

112.63

Fertilizer—private retail

0.00

17.43

24.53

0.00

0.00

0.00

0.00

Transport costs

n/a

4.76

5.99

9.24

6.33

5.95

5.54

Logistics Unit operations

n/a

0.37

0.42

0.24

0.21

0.34

0.28

ADMARC operations

n/a

n/a

0.00

0.06

1.06

2.24

1.57

SFFRFM operations

n/a

0.75

1.41

n/a

n/a

2.05

0.98

District financing

n/a

0.19

n/a

n/a

n/a

n/a

n/a

Coupon production

n/a

0.11

0.09

n/a

n/a

n/a

n/a

Communications

n/a

0.20

n/a

n/a

n/a

n/a

n/a

Input quality monitoring

n/a

0.05

n/a

n/a

n/a

n/a

n/a

M&E

n/a

0.29

n/a

n/a

n/a

n/a

n/a

Buyback finance fees

0.00

0.39

n/a

0.00

0.00

0.00

0.00

Total recorded costs

51.62

90.92

128.58

284.63

119.92

154.16

142.38

Less: Farmer redemption due

19.62

17.02

21.32

23.12

11.43

10.59

8.39

Unused stock (exc. buyback) 0.0

0.00

0.00

19.83

0.00

0.00

0.00

Net recorded Costs

32.00

73.90

107.26

241.68

108.49

143.57

133.91

Estimated other costs

Brought forward stocks

0.00

0.00

n/a

0.62

1.76

0.00

0.00

MoAFS operations

n/a

n/a

n/a

7.86

7.78

7.26

6.60

ADMARC/SFFRFM

n/a

n/a

n/a

1.26

n/a

n/a

n/a

Voucher printing

n/a

n/a

n/a

0.14

0.14

0.13

0.00

Other agencies’ costs

n/a

n/a

n/a

0.23

0.23

0.21

0.19

Total est. other costs

n/a

n/a

10.11

9.91

7.60

6.79

Total net costs, recorded & estimated

n/a

n/a

n/a

251.79

118.40

151.17

140.70

Total costs, recorded & estimated exc. cf stock

n/a

n/a

n/a

274.91

129.83

161.76

149.09

Programme budget

36.43

53.57

82.14

139.14

155.04

129.99

129.48

Funding

Direct donor support

0.00

9.51

7.13

37.75

17.48

22.05

44.85

Balance: Malawi Govt.

n/a

64.39

100.13

214.04

100.92

129.12

95.84

Cost, % MoAFS budget

n/a

46.8%

57.2%

67.6%

52.7%

60.1%

48.9%

Cost, % national budget

n/a

6.8%

8.2%

16.2%

6.5%

8.0%

7.1%

Cost, % GDP

n/a

2.5%

3.1%

6.6%

2.5%

3.0%

n/a

Notes: Farmer redemption due on ADMARC and SFFRFM fertilizer sales. Malawi Government contribution based on recorded and estimated costs shown.

Sources: Author calculations from Logistics Unit Annual reports, Nakhumwa (2006), School of Oriental and African Studies et al. (2008), Dorward and Chirwa (2009, 2011a, 2012a), Dorward et al. (2010b), Malawi Government Annual Budget Statements.

(p.123)

Notes:

(1) This is inclusive of transport costs and farmer redemption payments.

(2) Total private sector supplies to government in 2008/9 were greater than shown in Figure 5.2, as this excludes government purchases for winter production and purchases that were not used and then carried forward, contributing to the 2009/10 drop in new private sector supplies.

(3) Initial allocations based on crop areas or farm families were commonly modified for a variety of generally undocumented technical, administrative, political, and other reasons.

(4) The small increase in the NSO rural household growth rates from 2008/9 is due to post-2008 estimates relying on 2008 rather than 1997 census data.

(5) Estimates of the percentage of households receiving coupons are in broad agreement with survey results for six districts in central and southern Malawi reported by Holden and Lunduka (2010b) with 68 and 75% of households receiving coupons in 2007/8 and 2008/9 respectively, and 1.4 and 1.5 coupons per recipient household in the same years. Chibwana et al. (2010) estimate that 77% of households received fertilizer coupons in 2008/9 in samples from two districts in central and southern Malawi.

(6) Some subsidized OPV maize seed was supplied through ADMARC and SFFRFM under the 2005/6 programme without coupons with a 70% subsidy, but there is no further documentation of this.

(7) Holden and Lunduka (2010b) find that only 7% of respondents in 2008/9 reported ‘insufficient inputs’ as the main problem with the programme. However data from the same survey indicate that 19% of households considered fertilizer to arrive too late, 6% considered it was on time but insufficient, and 8% considered it both late and insufficient (equivalent figures for 2007/8 were 37%, 6%, and 6%, respectively) (Holden and Lunduka, 2010a).

(8) This divergence between perceived frequency and reported experience is interesting and may inform interpretation of FGD information—perhaps suggesting that the incidence of these problems is overstated in FGDs.

(9) It should be noted that some apparent targeting errors within the smallholder sector also involve the abuse of power and influence by people who are smallholder farmers.

(10) A similar analysis for maize seeds in 2010/11 (Dorward and Chirwa, 2011) and administrative records suggests substantial numbers of counterfeit maize seed coupons that year. These have less national and political significance and less individual and aggregate value than fertilizer coupons, but nevertheless offer substantial income opportunities for criminal activities.

(11) Some of these issues are addressed in the 2011–16 Medium Term plan for the programme (Government of Malawi, 2011) which has as its goal ‘to increase food security at household level through agricultural output growth’ and its purpose ‘to increase agricultural productivity and input market development’.