566–2004

ICRISAT-Patancheru(Headquarters)Patancheru 502 324Andhra Pradesh, IndiaTel +91 40 23296161Fax +91 40 23241239icrisat@cgiar.org

ICRISAT-Niamey(Regional hub WCA)BP 12404Niamey, Niger (Via Paris)Tel +227 722529, 722725Fax +227 734329icrisatsc@cgiar.org

ICRISAT-Nairobi(Regional hub ESA)PO Box 39063, Nairobi, KenyaTel +254 20 524555Fax +254 20 524001icrisat-nairobi@cgiar.org

ICRISAT-BulawayoMatopos Research StationPO Box 776,Bulawayo, ZimbabweTel +263 83 8311-15Fax +263 83 8253/8307icrisatzw@cgiar.org

ICRISAT-BamakoBP 320Bamako, MaliTel +223 2223375Fax +223 2228683icrisat-w-mali@cgiar.org

ICRISAT-Maputoc/o INIA, Av. das FPLM No 2698Caixa Postal 1906Maputo, MozambiqueTel +258-1-461657Fax +258-1-461581icrisatmoz@panintra.com

ICRISAT-LilongweChitedze Agricultural Research StationPO Box 1096Lilongwe, MalawiTel +265-1-707297/071/067/057Fax +265-1-707298icrisat-malawi@cgiar.org

The International Crops Research Institute for the Semi-Arid Tropics (ICRISAT) is a non-profit, non-political, international organization for science-based agricultural development. ICRISAT conductsresearch on sorghum, pearl millet, chickpea, pigeonpea and groundnut – crops that support the livelihoodsof the poorest of the poor in the semi-arid tropics encompassing 48 countries. ICRISAT also sharesinformation and knowledge through capacity building, publications and ICTs. Established in 1972, it is oneof 15 Centers supported by the Consultative Group on International Agricultural Research (CGIAR).

Contact information:

About ICRISAT

Visit us at www.icrisat.org

Rohrbach D, Charters R, and Nyagweta J. 2004. Guidelines for emergency reliefprojects in Zimbabwe: seed and fertilizer relief. PO Box 776, Bulawayo, Zimbabwe: ICRISAT; andRome, Italy: FAO. 68 pp.

Drought is endemic to southern Africa. Floods are common. In response to these disasters,agricultural inputs such as seed and fertilizer are distributed to one or another affected communityvirtually every year. Yet these relief programs are generally defined anew each year, and lessonsgained from previous experience are lost. Following the 2002/03 drought in Zimbabwe, the needwas felt for a set of guidelines to help NGOs and government agencies better plan reliefinterventions, and provide advice on strategies to help households and communities re-establishand strengthen their farming operations. The guidelines were drafted by a voluntary committee ofNGOs and international institutes.

This publication summarizes current advice on the design and implementation of agriculturalrelief programs in Zimbabwe. It covers various aspects – geographical and household targeting,what inputs should be provided, procurement and distribution methods, and monitoring andevaluation systems. The Guidelines primarily deal with the provision of crop inputs and technicaladvice, but livestock assistance is also discussed. They are written specifically for Zimbabwe, butmuch of the advice is relevant to neighboring countries, and thus suitable for broad application.

These Guidelines will continue to evolve and expand over the next few years, with improvedadvice on input packages, additional information on micro-irrigation systems and nutritiongardens, and suggestions for specialized assistance to households affected by HIV/AIDS. Readersare therefore encouraged to provide the authors with feedback for improving these Guidelines:corrections, lessons drawn ongoing relief programs, and suggestions of additional advice.

© 2004 by the International Crops Research Institute for the Semi-Arid Tropics (ICRISAT) and the Foodand Agriculture Organization of the United Nations (FAO).

All rights reserved. Except for quotations of short passages for the purposes or criticism or review, no part ofthis publication may be reproduced, stored in retrieval systems, or transmitted in any form or by any means,electronic, mechanical, photocopying, recording or otherwise, without prior permission from ICRISAT andFAO. It is hoped this copyright declaration will not diminish the bona fide use of research findings foragricultural research and development.

The views expressed in this publication are those of the authors and not necessarily those of ICRISAT orFAO. The designations employed and the presentations of material do not imply the expression of anyopinion whatsoever concerning the legal status of any country, territory, city, or area, or of its authorities, orconcerning the delimitation of its frontiers or boundaries. Where trade names are used this does notconstitute endorsement of or discrimination against any product by ICRISAT or FAO.

Guidelines for Agricultural ReliefPrograms in Zimbabwe

David Rohrbach, Rod Charters, and Jacob Nyagweta

ICRISAT

International Crops Research Institute for the Semi-Arid TropicsPO Box 776, Bulawayo, Zimbabwe

2004

Acknowledgments

This publication was funded by the European Commission’s Humanitarian Aid Office (ECHO)and the Department for International Development (DFID), UK.

The Guidelines are the product of collaborative efforts involving many parties. They were initiallyproposed by the Agriculture Coordinating Committee for Zimbabwe, an ad hoc grouping ofNGOs, donors, government staff, and others, organized to share planning and experience relatingto agricultural relief and recovery programs in the country. The Committee meetings have beencoordinated by the FAO Emergency Coordinator for Zimbabwe. A voluntary sub-committee wasestablished to draft the Guidelines. The three authors of this report played the most substantiverole in drafting. However, strong input was also provided by various committee membersincluding: K Furany (CARE), S Huddle (World Vision), M Jenrich (FAO), N Kutukwa (CARE),M Mudiwa (FAO), T Mupetesi (FACHIG Trust), L Mutiso (Oxfam), M Murata (CRS), and HSittig (Help).

Contents

1. Guidelines for emergency relief projects in Zimbabwe........................................................ 12. Rationale for emergency assistance in Zimbabwe ................................................................ 33. Institutional arrangements ................................................................................................... 5

4. Needs assessment and geographical targeting....................................................................... 95. Household targeting........................................................................................................... 11

6. Basic input support basket ................................................................................................. 157. Input procurement ............................................................................................................ 238. Timing of input distribution............................................................................................... 26

9. Distribution strategies ....................................................................................................... 2810. Technical support for agricultural recovery ...................................................................... 32

11. Integrating recovery efforts with development ................................................................ 3512. Relief and recovery in livestock systems .......................................................................... 3813. Monitoring and evaluation ............................................................................................... 41

AppendicesAppendix 1. Key contacts in Zimbabwe ................................................................................ 44

Appendix 2. Sample household registration form .................................................................. 46Appendix 3. Distribution of crops grown in the communal sector ......................................... 47Appendix 4. Simplified method for testing germination rates................................................ 49

Appendix 5. Seed varieties released for sale in Zimbabwe..................................................... 50Appendix 6. Sample contract for procurement of inputs ....................................................... 51

Appendix 7. Warehouse checklist .......................................................................................... 56Appendix 8. Datasheet for monitoring distribution of inputs to ward level ........................... 58Appendix 9. Sample of data reported on district level input distribution .............................. 59

Appendix 10. Program monitoring plan ................................................................................. 60Appendix 11. Sample of post-planting assessment survey questionnaire ............................... 61

1

1. Guidelines for Emergency Relief Projects in Zimbabwe:Seed and Fertilizer Relief

During the 2002/03 and 2003/04 plantingseasons, free seed and fertilizer were widelydistributed in Zimbabwe in response todrought and a general economic decline.This distribution was necessitated by thereduction in household seed stocks causedby poor harvests. Production losses havebeen exacerbated by a sharp decline ineconomic growth, and limited availability andhigh retail price of food grains. As a result,household and community seed stocks aremore likely to be consumed.

Zimbabwe’s recent problems have beensevere but not unique, or even uncommon.Input relief has been distributed in thiscountry during 7 of the past 12 years. Similarprograms have been periodically launched invirtually every neighboring country. Thedistribution of seed through relief andrecovery programs has become so commonthat several smaller seed companies haveemerged to service this market. Larger seedcompanies maintain at least some stocks ofa range of food crops to respond to thisdemand.

The substantial investment in inputdistribution programs naturally leads toquestions about theirefficiency. In late2002, the Food andAgricultureOrganization of theUnited Nations (FAO)began sponsoringmonthly meetings ofgovernmental andnon-governmentalorganizations (NGOs)providing input reliefin Zimbabwe in orderto improvecoordination. Thesemeetings initiallyinvolved NGOs linkedwith FAO programs,but eventually

included most of the major NGOs involved ininput distribution in the country.

The meetings of the informal FAO/NGOAgricultural Recovery CoordinationCommittee, involving major donors,international organizations and otherstakeholders, encouraged the sharing ofexperiences and stimulated the jointassessment of possible solutions to commonproblems.

Distribution of seed, fertilizer, and relatedagricultural inputs has undoubtedly helpedsmallholder agriculture recover after the pasttwo seasons of drought. However, manyquestions have been raised about programstrategies and impacts. It is generallyacknowledged that the effectiveness ofassistance would be improved by betterinformation flow – for example avoidingduplication or overlapping coverage of areas.

Related questions have been raised abouthow best to target farm households. Shouldinputs be given to the poorest and most foodinsecure, or to ‘more serious’ or ‘better’farmers more likely to achieve productiongains and thus improve food security in thecommunity? Can targeting be more efficient?

Proportion of households receiving relief seed and growingalternative crops, 2003/04

Relief programs in Zimbabwe have expanded rapidly in recent years,providing food, seed, fertilizer and other assistance

2

Should farmers in drought prone regions begiven maize seed? Is there a ‘best’ packageof inputs for each agro-ecological zone?What indicators should be included in impactmonitoring?

In late 2003, the FAO/NGO AgriculturalRecovery Coordination Committeeendorsed the drafting of a set of guidelinesto provide advice to government and NGOsabout how best to assist farmers in need ofrelief. Several members of the Committeevolunteered to help draft these guidelinesbased on their diverse experiences.ICRISAT agreed to help coordinate thepreparation of the guidelines. In the initialdraft, six NGOs – CARE, Catholic ReliefServices, GOAL, Oxfam (UK), FACHIG, andWorld Vision – contributed brief descriptionsof what they considered best practices. Thisdraft has been rewritten to provide a moreconsistent description of program options,as well as suggestions for best practice.Examples of the practices of specific NGOsare included.

This publication primarily considers issuesrelating to seed and fertilizer delivery in post-drought recovery programs in Zimbabwe.However, the Committee has agreed thatthis should be extended to a wider range ofbest practices relating to agricultural reliefand recovery. These include options fortillage support, crop management advice,

water management, livestock systemsmanagement, and linking farmers tocommercial markets for agriculturalproducts. Therefore, these guidelines shouldbe viewed as a work in progress.

The guidelines focus on programs with a 6 to12 month lifespan, targeting relief andrecovery after drought, as well as chroniccrises such as HIV/AIDS and the decliningeconomy. However, most of this advice canalso inform program planning in areasaffected by flooding. An effort is made tohighlight problems relating to input deliveryto households severely affected by HIV/AIDS. However, much more analysis isrequired of the needs and capabilities ofsuch households. This document representsour views of current best practice. Thisadvice is expected to evolve with time.

Finally, we note one persistent problemencountered during the drafting. We aretrying to provide simple advice to guide thedevelopment and implementation of relief andrecovery programs. Yet opinions differ aboutwhat constitutes best practice. Programs arechanging as more experience is gained withalternative methodologies. Therefore, theseguidelines provide specific sets of advice aswell as a discussion of program options,which could help users modify the advice tobetter suit a particular situation. More detail isprovided in a series of appendices.

3

2. Rationale for Emergency Assistance in Zimbabwe

The primary aim of agricultural relief andrecovery programs is to improve the foodsecurity of vulnerable1 farm households. Inparticular, these programs aim to strengthenthe capacity of these farmers to producetheir own food supplies; and thus expandand stabilize food supplies at communitylevel.Several reasons have been commonly citedto justify the need for agricultural assistanceduring the past two years.These include:

• Poor rainfall leading towidespread shortfalls infood production relative tohousehold and communityneeds

• Shortages of basicfoodstuffs on the retailmarket, increasing theprobability that farmers willconsume some of their seedsupplies

• The sharp decline ineconomic growth, reducing remittanceincome and off-farm employment

• Shortages and consequent high prices ofagricultural inputs on the retail market

• The high incidence of HIV/AIDS resultingin labor shortages, capital losses and alarger proportion of child-headedhouseholds.

In effect, the impacts of drought have beenconsiderably worsened by problems in thelarger agro-economy. It is likely that reliefprograms will still be needed even after morefavorable rains return. A significantproportion of farm households will remainfood-deficient and in need of assistance,perhaps for several years more.

The severity of these problems has led manyNGOs to include both food aid and inputsupply in their relief programs. These includedistribution of food aid (commonly maize, oil,and a legume), school feeding programs,under-five feeding and food for work. Most

NGOs distribute seed to help farmers re-establish their fields and expand theirplantings. Some NGOs distribute fertilizer,micro-irrigation kits and tillage support, andalso provide crop management. A few areinvolved with the rehabilitation of irrigationschemes, boreholes, dams and rural roads.

In many cases, relief programs are anextension of development activitiespreviously being implemented by the NGO

community. Some of thesedevelopment activities remainin place. For example, CAREcontinues to build thecapacities of rural retailers andrural micro-finance groupsthrough whom they distributeinput aid. ITDG has used foodaid to encourage farmers toestablish contours to improvewater infiltration into croplands. FAO has linked itsemergency program to its

farmer field schools and livestock forageproduction initiatives.

Opinions differ about the degree to whichrelief programs should also havedevelopment objectives. Some argue thatobjectives to improve the food security ofvulnerable households should not beconfused with development initiatives thatmay favor households with better resources.Others argue that relief programs ought tohave longer term development impacts.These distinctions are reinforced by the factthat many donor portfolios tend to clearlydifferentiate between humanitarian reliefprograms and development programs. InZimbabwe, programs funded by ECHO orUSAID/OFDA target relief impacts only;while DFID funding generally targets reliefwith longer term impacts. However, theseagendas may change over time.

Relief programs must be judged in terms oftheir impact indicators. The success of reliefseed distribution is generally judged in termsof the expansion of cropped area and1. See Chapter 4 for a broad definition of vulnerability

Short-term relief andlonger-term developmentinterventions often targetdifferent types ofhouseholds. Many donorportfolios clearlydifferentiate betweenrelief and development;and fund one type ofactivity rather than theother.

4

production. Seed is distributed to farmersbelieved to have little or no seed stocks. Ifthis seed was not distributed, food cropswould be planted over a smaller area, andfood insecurity would persist.

Similarly, the delivery of fertilizer is assessedin terms of the additional grain produced.Food-insecure farmers receiving seed andfertilizer should generally be able to harvesta larger crop than farmers who do notreceive these inputs.

Some programs also assess a range ofsecondary indicators. Most relate to thelogistics of delivery – whether all seed isaccounted for, the timeliness of delivery, thenumber of recipients, satisfaction with qualityand quantity, and the proportion of seedplanted.

Such indicators correspond with the shortterm output of these relief efforts. Thus far,

few efforts are made to measure broaderdevelopment impacts such as contribution tohousehold and community food supply,contribution to household income, successof technology adoption, increased efficiencyof crop production, or improvements ininstitutional arrangements for inputmarketing or agricultural relief.

The objectives of relief and developmentcannot always be strictly distinguished, butrelief programs can impact on development.Where possible, relief objectives should bepursued with positive development impacts.At a minimum, development impacts shouldbe neutral. But relief programs shouldensure they avoid deleterious impacts. Forexample, relief distribution of seed andfertilizer should not undermine thedevelopment of seed and input markets.Where possible, these programs shouldsupport the development of such markets.

Input distribution programs provide short term relief – but do not always contribute to long termdevelopment

5

3. Institutional Arrangements

For a relief program to be successful, theimplementing agency, such as aninternational organization or an NGO, mustestablish linkages with public agencies andprivate sector organizations. This chapterreviews several of the key linkages.

National MinistriesThe Ministry of Labor and Social Welfare isresponsible for declaring the need for relief,and then coordinating assistance to needycommunities. The Ministry of Agriculture andRural Development is responsible for district,provincial and national crop productionestimates. These are collected on a fortnightlybasis. This ministry houses the Department ofAgricultural Research and Extension Services(AREX) which is responsible for research andextension advice.

The Ministry of Transport and Energyhouses the Department of MeteorologicalServices which maintains national rainfallrecords, tracks incidences of flooding anddrought, and issues a weekly rainfall bulletin.All IOs/NGOs involved in relief programs areencouraged to maintain communication withsuch government ministries.

Local GovernmentAuthoritiesLiaison is essential withlocal administrativeofficials, though the exactnature of thesearrangements will differ ineach Province. In mostcases, permissions andcoordination are neededwith provincial and districtadministrative staff.Traditional leadershipstructures may also needto be consulted.

Each Province is dividedinto districts. Districts aredivided into wards. Each

ward falls under the responsibility of anelected councilor. In general, local levelpermissions must be sought from thecouncilor. This may be done first through theDistrict Development Committee, whichcomprises the government-appointed DistrictAdministrator plus all councilors in thedistrict. Zimbabwe has 10 provinces (2urban, 8 rural) and 83 districts (14 urban, 69rural), each of which contains 15-30 wards.

A ward consists of several villages, each ledby a headman; the position is generallyhereditary. The village headman (or kraalhead, in some areas) is responsible forallocating village land and resolving locallevel disputes. He generally maintains villagelists and is responsible for identifyinghouseholds in need. The headman’spermission must be sought before entering avillage community.

Once again, there is no single set ofprocedures, throughout the country, forcommunity entry. The important thing is torespect the existence and concerns of localauthorities; and good relations depend asmuch on good communication as on protocol.

The provinces of Zimbabwe; districts shown but not labeled

6

Agricultural Research andExtension Services (AREX)Agricultural research and extension inZimbabwe is coordinated by a singlegovernment department called theAgricultural Research and ExtensionServices. The national research service isthe primary source of information aboutcropping technologies, and the primarysource of new seed varieties for crops otherthan maize. The extension service developsand disseminates crop managementrecommendations. These are summarized inthe Farm ManagementHandbook1 as generic ‘blanket’recommendations for thecountry as a whole. However,these recommendations tendto be adapted, on a somewhatad hoc basis, by provincial anddistrict level extensionauthorities. Each ward issupposed to have an extensionofficer.

The extension service is alsothe primary source of crop productionestimates. Ward level extension workersprepare fortnightly reports containingestimates of area, production, and cropcondition. These are collated at district,province and national levels. Each year, theestimates are reviewed, and possiblyamended, by a national Crop ForecastingCommittee. This Committee then announcesa First Official Crop Forecast, usually inFebruary, and a Second Official CropForecast, usually in April. The secondforecast is generally the final official estimatefor the season.

In general, NGOs operating in rural areasbenefit from support of local extension staff.At a minimum, it is a courtesy to inform localextension staff while entering a community.District and ward level extension staff canalso provide valuable information about thelocal agro-ecology, crops commonly grown,

and crop management practices. Extensionstaff may also provide assistance withtraining farmers about new varieties, andhelp run demonstration trials to promotegood management practices.

International AgriculturalResearch CentersZimbabwe hosts several programs of theinternational centers of the ConsultativeGroup for International Agricultural Research(CGIAR). These are valuable sources oftechnical advice about crop varieties, crop

management andZimbabwean agriculture.ICRISAT maintains asignificant program at theMatopos Research Stationsituated 26 km south ofBulawayo. It holds a globalmandate for sorghum, pearlmillet, groundnut, chickpeaand pigeonpea research, andthus offers an excellent sourceof information about thesecrops.

CIMMYT has a global mandate for wheatand maize research, and maintains a maizeprogram at the University of Zimbabweresearch farm approximately 15 km north ofHarare. ICRAF maintains a program of agro-forestry research while CIFOR works moredirectly on forestry issues. The relevantcontact addresses for these institutions arelisted in Appendix 1.

Seed CompaniesMore than ten seed companies operate inZimbabwe. Many of them also serviceneighboring countries. The Seed Companyof Zimbabwe (SeedCo) accounts for morethan half of all sales in the country. Otherlarge companies include PANNAR, Pioneer,Monsanto, National Tested Seed, and PrimeSeed (see Appendix 1). These companiesare the main source of certified seed in the

1. AGRITEX 2003. Farm management handbook, 4th ed. Harare, Zimbabwe : Department of Agricultural TechnicalExtension Services (AGRITEX).

Seed companies focus onhybrid maize, cotton andother commercial crops.They produce only smallquantities of traditional foodcrops (eg sorghum,cowpea) because theybelieve commercialdemand is limited andhighly variable, dependingon relief requirements.

7

country. They will also import seed stocks ifthey are not available locally, mostly fromneighboring countries – particularly Zambia,Botswana and South Africa. However,occasionally, seed has been imported fromas far away as India and Egypt.

The main problem with the seed industry isthat seed stocks (except for maize) are low.Seed companies are reluctant to producesignificant quantities of crops such assorghum, pearl millet, groundnut and cowpeabecause they believe commercial demand islimited. They do not want to be left with largestocks in the event that relief anddevelopment programs in Zimbabwe andneighboring countries cut down onpurchases. This problem is magnified by thefact that the ultimate level of seed stocks maynot be known until several months after thepreceeding season’s harvest. The seedcompanies contract farmers to produce theirseed, but they remain uncertain about thelevel of deliveries for several months.

NGOs also need to be aware that a commonway to resolve seed supply deficits is topurchase grain from local markets and cleanit to seed specification – essentially forphysical purity and germination. This helpsensure that farmers receive varieties that areadapted to local conditions; but the seed isoften genetically mixed – Common Gradeseed, in industry parlance. Chapter 6provides more information on these issues.

Fertilizer CompaniesZimbabwe currently has five major chemicalfertilizer companies (see Appendix 1).Three are licensed for commercial salethrough the wholesale and retail trade –Zimbabwe Fertilizer Corporation (ZFC),Windmill and Omnia. The other two arelicensed to produce fertilizer – SableChemicals and Zimphos. During the pasttwo years, the operations of thesecompanies have been severely restricted bythe country’s foreign exchange shortages.Zimbabwe produces its own phosphorus,but must import all its ingredients fornitrogen, as well as most minor nutrients.

Sable Chemicals is the only producer ofnitrogen in the country. In 2003, thecompany stated it would give priority tobuyers able to provide foreign exchange forthe import of ammonia from South Africa.

Other Sources of InformationIn 2002, Zimbabwe established a nationalVulnerability Assessment Committee (VAC) –a subcommittee of the Social ServicesCabinet Action Committee, in turn composedof a consortium of government, NGO andUN agencies. The VAC coordinatesassessments of the level and scope of theemergency at the national level. A series ofnational surveys were conducted to assessagricultural production, nutrition, and ruraland urban poverty. In March 2003 the VACissued a major report on food insecurity inrural areas. A similar report was completedfor urban areas.

The Famine Early Warning Network(FEWSNET) is a good source of informationon agricultural conditions and food securityin the country. FEWSNET issues periodicreports estimating production levels, foodsupply levels, nutritional indicators, and thestatus of agricultural production.

The Surveyor General’s office offers a goodsource of maps including detailed maps ofrural districts – though village demarcationson these maps are not always accurate. Thisoffice also sells aerial photos of variouslevels of resolution for much of the country –though, for any given area, these are likelyto be several years old.

The Central Statistical Office sells periodicreports summarizing agriculture in thecountry, national accounts and economicstatistics, and inflation estimates. Thesereports are generally available at theGovernment Publications office.

The Meteorological Service currentlyprovides a weekly report summarizingnational weather conditions including moredetailed data for a small cross-section ofareas. This is available for a smallsubscription fee.

8

FAOThe FAO is taking the lead in coordinatinghumanitarian interventions in the agriculturalsector, through monthly stakeholdermeetings and a series of sub-workinggroups. The meetings involve organizationsdirectly involved in relief work, donors, andother stakeholders participating directly andindirectly. Information about these meetingsmay be obtained through FAO’s EmergencyCoordination Unit for Zimbabwe. FAO also

has a database of agricultural interventionsdisaggregated by ward, NGO and crop.

UNDP/European UnionIn 2003, the UNDP established a Relief andRecovery Unit (RRU) with EU and DFIDfunding. The RRU has a mandate to collectand distribute information about relief andrecovery programs in the country. It alsofacilitates coordination of some of theseactivities.

9

4. Needs Assessment and Geographical Targeting

Two different assessments are requiredbefore program beneficiaries can beidentified. The first, discussed in this chapter,is to identify what parts of the country areaffected by crop losses, ie what parts of thecountry to assist. The second, reviewed inChapter 5, is a more detailed assessment ofneeds within target communities, andselection of household beneficiaries in thesecommunities.

Assessing the Scope andDistribution of the EmergencyDeclaration of an agricultural emergency inZimbabwe is the responsibility of the Ministerof Labor and Social Welfare.The Minister is responsible forreviewing production datacollected by the Ministry ofAgriculture and RuralDevelopment and assessingthe severity of rural and urbanfood insecurity. The finalassessment relies heavily onthe Second Official CropForecast, which summarizes aggregate cropproduction levels for the country as a wholeand for various classes of farmers (large-scale commercial, small-scale commercial,communal and resettlement). Province-wisebreakdowns may be provided for maizeproduction.

Once an emergency is declared, villageheadmen and councilors are asked tocompile lists of needy households to betargeted for food aid and input relief. Theseare aggregated to district and province levelsto estimate the number of households inneed.

Individual NGOs also seek information fromtheir local contacts in rural communitieswhere they are running programs. This mayinvolve consultation with district and wardofficials, AREX officers, chiefs, and villageheadmen.

For the first time, in 2003, VAC survey dataallowed a disaggregated assessment of food

The 1993 VAC surveyestimated that 4.4 millionrural residents (56% of therural population) wouldneed food aid – 754,800tons of grain would needto be imported.

supply shortfalls in each district. Thesefindings were summarized in EmergencyFood Security and Vulnerability Assessment(Report No. 3), commonly known as the VACReport. The report estimated that 4.4 millionrural residents (56% of the rural population)would need food aid, and that thegovernment would need to import 754,800tons of grain to meet these needs. Mostimportantly, this report provided abreakdown, by district, of the level of need –measured in terms of the number andpercentage of households experiencingcereal supply deficits.

Selection of Districts and WardsNGOs select which parts ofthe country to target for reliefassistance based on their pastexperience and the location oftheir current programs.Informal consultations withother NGOs will also helpselect target areas. OlderNGOs tend to follow their

established contacts in particular districtsand wards. Newer NGOs look for districtsand wards that are not covered. Permissionis then sought from Province and Districtauthorities to distribute aid within a particulardistrict or ward.

The FAO Emergency Unit for Zimbabwecollects data on NGO activity by district andward. The 2002/03 database highlighted thelikelihood of over-coverage of certaindistricts and under-coverage of others –based on the perceived distribution of need.The 2003/04 database highlighted similarproblems. Many districts received moreinputs than needed to assist the entirepopulation identified by the VAC report asbeing in need.

FAO also collects data on the distribution ofagricultural inputs by ward. This againshowed over-coverage of many regions, withmultiple NGOs sometimes serving the same

10

ward. These findings were confirmed insurveys conducted by ICRISAT.

Most observers in the NGOcommunity acknowledge theneed for better coordination oftheir relief efforts, down to theward level. But this will remaindifficult, as long as NGOs’plans continue to evolve duringthe early stages of thecropping season. Thishappens if funding grantsarrive late, the identification oflocal partners is prolonged orsimply because NGO staff take an extendedperiod of time to establish contacts with local

Over 30 NGOs (plus government agencies ARDA and GMB)were distributing relief inputs in one area studied

Substantial over-coverageand overlaps are reported.Many districts receive farmore inputs than needed;multiple NGOs sometimesserve the same ward. Onemajor reason is lack ofcoordination betweenNGOs operating in anarea.

% of farmers receiving inputs from multiple NGOs authorities. In the worst cases,decisions about ward and villagecoverage are still being made whileinput delivery is underway.

In order to improve coordination, theFAO Emergency Unit seeks to obtaindata from each NGO on theirintended ward level input distributionearly in the cropping season.

Ultimately, stricter agreements maybe needed, earlier in the planningprocess, about geographicalcoverage. As long as multiple NGOsoperate in a district, strict delineationof zones of coverage may prove

difficult. However, an early agreement onward coverage should help reduce the

current overlaps in communitycoverage. This agreementwould need to be in place atleast four months before thefirst inputs are distributed. AnIO/NGO protocol is currentlybeing developed, highlighting(among other issues) the needfor early identification anddissemination of assistanceplans. This will help addresssome of the problems

associated with over-coverage andoverlapping.

11

5. Household Targeting

The first priority for any relief program is toestablish robust criteria for interventions athousehold level. To identify whichhouseholds should receive assistance, twobasic sets of decisions must be made. First,NGOs must identify the types of householdsto be targeted. Second, they must establisha procedure to identify these households. Ineither case, it is useful to involve the localcommunity in the selection process. Localtransparency increases the chances ofaccurate targeting.

Types of HouseholdsTargetedThere are no universal criteriato identify vulnerablehouseholds. The choice ofindicators or criteria dependson the NGO; its priorities,mandate, specific constraints,and the political environment.Most relief programs target the mostvulnerable households. However, manyNGOs also target assistance to householdsparticipating in their ongoing developmentprograms.

Many indicators have been proposed for theselection of vulnerable households. Theseare, in effect, proxy indicators for food supplydeficits. The most common:• Child headed households• Female headed households• Elderly• Widows and widowers• Households with disabled members,

chronically ill members or orphans• Households with limited cash income, no

pension and no formal employment• Households with high dependency ratios• Households with limited assetsQuestions then arise about how theseindicators are best measured. Many of thecriteria are vague – how should the term‘elderly’ be defined; what is meant by ‘limitedassets’ or ‘high dependency ratio’. Oneoption is to establish explicit criteria for each

Stringent, clearly definedcriteria can minimizetargeting errors – but maybe costly and sometimesinappropriate. Moreflexible criteria lay downthe general principles forselection and allowcommunity participation inselection of beneficiaries.

variable – eg minimum/maximum age ceilingon household assets, target dependencyratio. Such stringent criteria have theobvious advantages of clarity, and canminimize inclusion/exclusion errors.However, they tend to be costly toimplement. Few NGOs have the field staffnecessary to collect and evaluate suchinformation for each target community. As aresult, most NGOs prefer more flexiblecriteria, laying down the general principles

for selection and leaving theresponsibility for selection tothe community.

Questions are also beingasked why only the pooresthouseholds benefit. Somecommunity leaders argue thatvillage food security can bebetter enhanced by providingassistance to better endowedhouseholds, which are more

likely to use the inputs to produce a cropsurplus. These households are said to bemore likely to employ members of morevulnerable households and to provide themost needy with food and seed. Theseclaims have been backed by evidence thatmany poorer households only use part of theinputs they receive.

To cope with these arguments, some NGOscombine their vulnerability indicators with anadditional set identifying households likely to‘seriously’ invest in crop production. Thesecommonly include:• Households with adequate land for

planting ( a minimum level is specified)• Households with adequate draft power for

land preparation• Households with enough labor to grow a

crop• Households not engaged in gold panning

or other off-farm activitiesYet several of these latter indicators contradictthose used to identify the most vulnerablehouseholds. This is particularly true for draftpower. In many communities more than 40%

12

of households – generally the poorest andmost food insecure – do not have enoughcattle or donkeys to field a minimum team of2 animals. They may rent or borrow animalsor draft services from relatives and neighbors.This leads to late planting and low yields.However, most of these households shouldstill be classified as committed farmers.

HIV/AIDS Affected HouseholdsApproximately 26% of Zimbabwe’s adultpopulation (ages 15 to 49) is officially classifiedas HIV-positive. Rates are marginally lower inthe rural areas, but highly variable acrossdistricts. The high rates of prevalence have leddonors to recommend explicit targeting of HIV/AIDS affected households. This advice isreflected in many of the indicators used toidentify vulnerable households – child headedhouseholds, households with chronically illmembers, or households with many orphansor large dependency ratios.

While there has been much talk of childheaded households, the actual proportion ofthese families inrural Zimbabweappears low –less than 2% ofthe population.Children arecommonly takenin by aunts,uncles andgrandparents.However, NGOsneed to watch forcases where anolder relative issaid to beresponsible for agroup of children,but misuses relief assistance. There isanecdotal evidence of cases where an uncleresponsible for a group of orphans simplycollects relief inputs for his own use, leavingthe children to fend for themselves.

The capacity of HIV/AIDS affectedhouseholds to make effective use of

agricultural inputs is also uncertain. A recentassessment by the VAC1 highlights sharpdeclines in both area planted and productionlevels among affected households. Theseare due to labor constraints, as well as theloss of farming assets sold to pay formedicines or funeral expenses. The mostsignificant declines in production are linkedwith the recent death of an adult decisionmaker. However, little is known about thecoping strategies these families pursue tooffset resource constraints over a period ofseveral years.

One option is to provide specialized relief anddevelopment support targeting HIV/AIDSaffected households. In one such program,orphans, or youth from households that havelost adult decision makers, are being giventraining in new income earning opportunitiessuch as small-scale poultry enterprises. Inother cases, HIV/AIDS affected householdsare being encouraged to concentrate onimproving food production from nearby

household plots. Many relief programsincorporate HIV/AIDS awareness training.

Procedures for IdentifyingVulnerable HouseholdsMost NGOs struggle to identify whichspecific households should be classified as

1. Vulenerability Assessment Committe (VAC). 2003. Zimbabwe emergency food security and vulenerabilityassessment, April 2003. Harare, Zimababwe: VAC.

Proportion of households with severe grainshortfall (harvest < 500 kg), 2003/04

13

vulnerable, and thus meriting of assistance.Strategies range from more formalprocedures of data collection, to less formaldiscussions with community leaders or fullcommunity meetings.

CARE follows a more formal practice (seebox). Following community meetings tointroduce the intended program, allcommunity members are asked to registerfor possible assistance. Householdscomplete a simple one page registrationform (see Appendix 2) providing informationon the main proxy indicators of vulnerability.The data from these forms are computerizedand the vulnerable are identified using cut-offs linked with the amount of assistanceavailable. The beneficiary list is then taken tothe community for verification.

This approach requires strong advanceplanning and ground level support, morelikely to be available among large, wellestablished NGOs. Farmers like it becausethe procedure is viewed as impartial.However, this procedure can be prone toerror insofar as potential aid recipients areaware of the criteria being used to judgewhether they qualify. While ground-truthingcan reduce these biases, this process alsocan be manipulated by local elites.

An alternative strategy is to call acommunity meeting and let the communityitself identify the most vulnerable or needyhouseholds. In general, village leaders areasked to gather all households to a meetingwhich is addressed by local and NGOofficials. The general criteria areannounced, and the community thenidentifies which households should receiveassistance. A public discussion is believedto reduce cheating. It also ensures broadawareness of the reasons why somehouseholds receive aid while others do not.This strategy is also quicker, cheaper andmore transparent than formal datacollection.

Yet this procedure is also said to be prone tobias. Some farmers may complain that theywere shut out of the process as a result of

Steps in registration andhousehold targeting –Example 11. Hold province and district level

meetings to introduce theintervention

2. Call for awareness meetings(through local councilors andtraditional leadership) to explainintervention and selection criteria(see application form)

3. Communities line up by village; pre-screening by communities begins

4. Application forms issued,completed and sent to NGOoffice. Vulnerables who do notattend the meetings cannot applyfor benefits

5. Applicants entered into database,selection criteria applied

6. Initial beneficiary register produced7. Verification meetings to clean

registers8. Final beneficiary lists prepared

their political views or distance fromcommunity elites. This is particularlyproblematic in periods prior to elections.

While methods for identifying targethouseholds are becoming more exact, theproportion of households receivingassistance remains high. In manycommunities in the drier, more drought proneparts of the country, 80% or more ofhouseholds are receiving assistance. Ineffect, the definition of vulnerability appearshighly elastic depending on the resourcesavailable.

Debates about what proportion (or number)of households should be assisted can beresolved only through a closer analysis ofthe relationship between the proxy indicatorsof poverty or farming capability, and theimpacts of input distribution. To date, therehave been no significant efforts to test theserelationships.

14

Steps in registration and household targeting – Example 2

Guidelines are first developed for the selection of beneficiaries to ensure uniformity inoperations by district teams. The criteria include:• Households without draft animals and with limited small stock• Female-headed (de jure) households• Households with limited cash income, no pension, no formal employment and little

or no remittances• Households with a high dependency ratio, ie, high numbers of children, orphans,

handicapped, terminally ill and the elderly• Male-headed households with limited assets (satisfying criteria 1 and 3 above)• Households should be acceptable to the community as being able to utilize the seed

offered.

If sufficient quantities of inputs are available, selection may be opened to householdsthat have less than a defined number of draft animals (this ‘limit’ to be determined atdistrict level), and satisfy criteria 3 and 4 above.

Registration is conducted in open community meetings in the respective wards andfood distribution points. The registration team reads out and explains the selectioncriteria to the community. Each community then collectively selects beneficiaryhouseholds who meet the criteria. A cross-section of households is later visited toverify the information supplied.

15

6. Basic Input Support Basket

The types of inputs to be provided depend onthe objectives of the relief program and thecharacteristics of the local farming system.Most relief programs distribute seed in orderto help farmers re-establish themselves aftera flood or drought. Seed is a simpletechnology that can readily be divided toassist large numbers of households. But seedalone may contribute little to household foodsecurity. Other kinds of inputs/interventionsmay offer higher returns on investment. Thefollowing discussion summarizes some ofthese trade-offs.

How Much Seed to Distribute?Drought is the most common factor triggeringseed interventions in Zimbabwe, but not theonly factor. Some parts of thecountry periodically experienceflooding. But access to seedmay also be limited becauseseed markets fail, or becausethe capacity to purchase inputsmay be reduced by high ratesof inflation.

To determine how much seedto provide, NGOs must firstconsider the fact that, contraryto common opinion,smallholder communitiesrarely, if ever, run completely out of seed. Indrought years, some farmers will lose mostof their seed stocks, and the aggregate levelof community stocks may be reduced. Butmany farmers will remain with seed, andlocal seed markets perform reasonably wellat reallocating these stocks. Followingseasons of favorable rainfall, most of thesetransactions involve cash or barter trade.Following drought years, gifts predominate.As farmers themselves describe it,

“I have an obligation to help my neighbor withseed, just as she has an obligation to help mein return when my harvest falls short.”

Unfortunately, years of seed relief haveconditioned farmers to complain about seedshortages even if the shortage is not severe.Drought affected communities are similarlyconditioned to show visiting NGOs theirempty granaries. The visitors assume that allseed is lost. But the fields actually plantedthe following season almost always contain awide range of traditional and improvedvarieties – far beyond the limited range ofseeds distributed through relief programs.

While recognizing the proclivity of farmers toask for aid, studying the community’s seedneeds will help an NGO assess what seeds

of what crops are most likely tobe in short supply. ICRISAThas worked with farmers andvillage authorities inneighboring Mozambique todevelop a seed needsassessment methodology1 thatprovides a first approximationof community requirements.This considers the availabilityof stocks from better thanaverage farmers, from uplandfields, from grain markets and

from the retail market.

The success of seed fairs in many countries(including Somalia, Sudan, Kenya, Tanzaniaand Mozambique) provides ample testimonyto the capacity of farming communities tomaintain local seed stocks (Chapter 9discusses seed fairs in more detail). Seedfairs based on the supply of seed from localfarmers have been successfullyimplemented in many communities inZimbabwe.2 These commonly supported the

Seed relief helps farmersre-establish themselvesafter a flood or drought. Butseed alone may contributelittle to household foodsecurity. Alternative inputsor interventions – cropproduction and fertilizertraining, marketinformation, improvedvarieties – may offer higherreturns on investment.

1. ICRISAT/INIA. 2003. Organizing seed fairs in emergency situations: improving the efficiency of seeddistribution. Maputo, Mozambique: ICRISAT and INIA.

2. During the 2003/04 season, seed fairs were implemented by CRS/CTDT, FACHIG and World Vision in morethan 10 districts in Zimbabwe. Local seed markets failed to provide a significant share of seed in only one area,Matobo District south of Bulawayo. This is the first known example where seed fairs have not attractedsignificant local seed stocks. This was likely due to two factors – many households in the area rely onremittance income, and the season was preceded by two consecutive years of severe drought.

16

delivery of commercial maize seed tocomplement locally available seed of cropssuch as sorghum, pearl millet, cowpea,pumpkin and various squashes.

Such experiences again highlight theimportance of targeting. Some householdswill have seed of many different crops whileothers maintain only limited stocks.Shortages are more likely for legumes,compared with cereals. In Zimbabwe,farmers are accustomed to purchasinghybrid maize seed on the retail market. Insome communities farmers particularlysought maize seed through relief programsin order to reduce their purchases.

Improved vs Traditional VarietiesSeed distribution programs should not beused to distribute new varieties of uncertainadaptation and acceptance. Seed from localgrain markets may be better than seedimported from unknown sources outside thecountry. However, many Zimbabweanfarmers look to relief programs as a sourceof access to new varieties.

Relief programs are largely responsible forthe widespread adoption of new whitesorghum (SV 2 and Macia) and pearl millet(PMV 2, PMV 3) varieties in the country.These varieties mature early, increasing theprobability of a harvest if rains end early.Relief programs are also the main (thoughlimited) source of the drought tolerantcowpea variety IT 18. During the past year,relief programs have been virtually the onlysource of new, open-pollinated maizevarieties now being grown in Zimbabwe. Allthese new varieties were widely tested in thecountry before being released by thegovernment and distributed by NGOs.

The main problem, revealed in recentsurveys of relief programs, is that farmerscommonly lack information about thevarieties they are receiving. In a recentsurvey of 1500 households, at least 40% offarmers did not know the names of thevarieties provided by various NGOs. Theydid not know the growth characteristics ofthe crop, or even whether this was hybrid oropen pollinated seed. This helps explain whymany farmers did not plant all the seed they

Improved varieties offer substantial yield gains over traditional varieties – as much as 50% in a bad season

17

received – logically, they sought to assessthe variety’s performance before allocatingmore of their land.

The payoffs to relief programs can often beconsiderably enhanced if they are able toprovide farmers access to more productivenew varieties. The international agriculturalresearch institutes of the CGIAR as well as thenational research and extension services canprovide useful information about the existence,performance and availability of new varieties.

Hybrid vs Open PollinatedVarieties of MaizeMore than 90% of Zimbabwean smallholdershave experience growing hybrid maize. Thiswidespread adoption resulted from theremarkable performance – even in droughtprone areas – of several hybrids releasedduring the 1970s. Most of these farmersdiscarded open-pollinated varieties (OPVs)of maize in favor of hybrid seed; and thegovernment subsequently outlawed the saleof OPV maize in the country. Consequently,sales of hybrid maize were entrenched.

In late 2002, Zimbabwe changed its policyand allowed ‘fast track’ registration of fiveopen pollinated maize varieties fordistribution and sale in the country. One ofthese varieties, Kalahari Early Pearl, is anolder variety widely promoted in reliefprograms in neighboring countries. Theother four (ZM 421, ZM 521, Matuba andObatampa) are new varieties from CIMMYT.Small quantities of several of these varietieswere distributed for the first time throughrelief programs in 2003/04. Larger quantitiesof seed are now being produced for deliveryduring the 2004/05 planting season.

Many NGOs have expressed interest inpromoting sales of OPVs so farmers canmore readily save seed from their harvest.Some claim farmers are better off growingOPVs. However, this decision should bemade by farmers, not by NGOs. On-farmtrials have shown that OPVs perform well.

Farmers now need to be educated about thedifferences between OPV and hybrid maize,

and offered the opportunity to try both. Forexample, NGOs can distribute 5 kg of eachtype and encourage farmers to experiment.Such experimentation is likely to havelonger-term impacts than simply distributingwhatever seed is most readily available.

Maize in Drought Prone RegionsSome NGOs similarly argue that maize seedshould not be distributed in drought proneregions because it simply perpetuates theneed for food aid. But farmers in these areasare well aware of the risks of maize production;and adopt strategies such as timely planting onbetter soils, or multiple plantings, in order toimprove the probability of a harvest.

Hybrids versus OPVs

Given adequate rainfall and inputs,hybrids will generally give higher yieldsthan open-pollinated varieties (OPVs).However, the actual difference in yieldsdepends on the choice of variety.

A main advantage of OPVs is thatfarmers can re-use a variety for severalseasons before buying fresh stocks ofcertified seed. This saves money.Correspondingly, OPVs may be suitablefor smallholder farmers who cannotafford to purchase inputs, and thoseliving in outlying areas with weak inputmarkets.

Zimbabwe has a long history of hybridcrop breeding. It was the second country(after the USA) to release maize hybrids.By the mid 1980s, virtually every small-scale farmer in the country hadexperience growing hybrid maize. Infact, sales of OPV maize seed werebanned. In recent years, however, highseed prices have encouraged a shiftback to OPVs. In 2003 the sale of OPVmaize seed was legalized.

NGOs should be aware of this history ofhybrid seed development, and be surethat farmers understand whether they arereceiving hybrid or open pollinated seed.

18

Undoubtedly, maize is less drought tolerantthan sorghum or pearl millet. Farmers inmost drier regions of the country consistentlyplant these small grain crops as a form ofinsurance against rainfall risk. Yet maize alsohas well known advantages. If rainfall is fairlyconsistent through the season, maize willgenerally give higher yields than sorghum orpearl millet. It is also less prone to birddamage, and easier to consume as a greenvegetable during the period before the mainharvest, when household food supplies arelowest. Maize is commonlyviewed to be easier to processinto meal or flour.

Once again, farmers should begiven the choice. Most willchoose to grow maize pluseither sorghum or pearl millet,regardless of the views ofNGOs.

Seed PacksThere is no single best seedpack to distribute to all farmersin Zimbabwe, or even to allfarmers within a particularagro-ecological zone. Reliefprograms first need to consider what cropsare commonly grown in the target district orcommunity, and then consider what seed ismost likely to be needed.

The distribution of crops grown in Zimbabweis summarized in Figure 1 (and in greaterdetail in Appendix 3). Maize is widely grown.Sorghum and pearl millet are most importantin the drier parts of the country – NaturalRegions IV and V. Groundnut is locallyimportant in many areas while cowpea tendsto be more common in the drier areas. Sugarbeans tend to be grown in higher rainfallzones and in areas close to urban markets.

Survey evidence suggests that ruralcommunities are most likely to retainsignificant seed stocks of drought-tolerantcrops with high multiplication ratios, such assorghum and pearl millet. Farmerscommonly look to relief programs for newvarieties of these crops.

Maize seed is more difficult to retain, if onlybecause much of the seed being planted isof hybrids. Farmers hope relief programs willdistribute fresh maize seed so they need notbuy it. However, there is also growinginterest in gaining access to the newlyavailable maize OPVs.

Legume seeds are most likely to be in shortsupply, because of the low multiplicationratios and the high demand for legumegrain, which is considered a premium food.

Plot sizes are generally smalland available stocks arequickly consumed as a snackor relish. Larger supplies maybe sold in the high pricedconfectionery market,particularly by householdsneeding a bit of cash.

The next question is howmuch seed to distribute.Again, much depends on therelative availability of seedfrom the formal and informalmarkets. However, a startingpoint for poorer householdswho have lost most of theirseed is to provide enough

seed to plant just over 1 ha of a mixed cerealand legume crop. Table 1 shows twopossible crop packs – one suitable for areaswhere maize dominates (eg the higherrainfall areas of Natural Regions I, II and III),and the second for areas where sorghum orpearl millet are more important (low rainfallareas, Natural Regions IV and V).

The harvest will depend more on the cropmanagement practices employed than thequantity of seed provided. Table 1 shows theapproximate harvest expected if rains arefair to good. Yields will be much higher, ineither environment, if improved managementpractices are used. Yields will be lower ifcrops are planted late (relative to the timingof individual rainfall events) or on poorquality soil.

Both these alternative seed packs containmore legume seed than is normally providedin Zimbabwe’s relief programs. Groundnut orcowpea allotments are usually small

Farmers are usually selfsufficient in seed ofsorghum and pearl millet;but relief programs canplay a crucial role inintroducing new varieties.For maize farmers wanthybrid seed as well asnew OPVs.Legume seeds is often inshort supply, because ofthe low multiplicationratios, small quantitiesharvested, and highdemand for legume grain.

19

because the seed is expensive, and onlylimited stocks are available on local markets.Investments in legume crop seedmultiplication may be needed in order tomake this package more practical.

Also, NGOs frequently complain that farmersconsume their groundnut or cowpea seedallocations. The reasons include: limitedquantities of legumes are available on localgrain markets, or from food aid. Groundnut,cowpea and beans are highly valued as acomplement to the grains provided in aidallocations. Correspondingly, there may begood justification for distributing morelegumes in food aid allotments during thelegume planting season.

Seed Quality IssuesThere are frequent reports of poor qualityseed being distributed through reliefprograms – low germination, mixed geneticpurity, or poorly adapted varieties. The bestway to avoid these problems is to purchasewell-labeled seed of known varieties ofreliable origin.

Seed companies normally test thegermination rates of seed they intend to sell.Purchase contracts may specify a minimumlevel of germination. In general, seed ofcereal grains should have a germination rateabove 85%. Seed of legumes, such asgroundnut or cowpea, should have agermination rate above 70%. Problems oflow germination are more common forlegume seed than for cereal grains. NGOscan easily test germination rates of the seed

they receive before distribution. Simpleselection and testing procedures aredescribed in Appendix 4.

The most common problem with relief seedfor secondary crops such as sorghum, pearlmillet and cowpea is mixed seed lots. Somecases have been noted, where seedcompanies bought and sold common gradeseed when their own stocks ran short. Mostof this seed is purchased from the local grainmarket, cleaned for physical purity andchecked for germination. But it may notalways be clearly labeled as common gradeseed. Instead, it may be labeled genericallyas mixed cowpea, or simply as whitesorghum or pearl millet. NGOs must thendecide whether to offer farmers seed of anuncertain and possibly mixed variety, or noseed at all. At least, this seed is likely to besuited to local environments, although it maybe no better than the seed available in local,informal seed markets.

During the past 12 years, any poorly adaptedseed sold in Zimbabwe has invariably beenimported. Pearl millet seed purchased fromIndia following the 1991/92 drought, wasgenerally late maturing and low yielding. In2002 and 2003, several seed companiessold large quantities of a white sorghumvariety labeled as Macia. However, thisproved to be a late maturing varietypurchased from the Mozambique grainmarket. Most of this crop yielded no grainbecause it matured long after the rains wereover. Most unfortunately, this seed was mis-labeled. Macia is a popular early maturingvariety tested and released in Zimbabwe,

Table 1. Two alternative crop packs to plant approximately 1 ha

Higher rainfall zone Lower rainfall zone

Seed quantity Area Expected harvest* Seed quantity Area Expected harvestplanted planted**

10 kg maize 0.5 ha*** 600 kg 8 kg maize 0.4 ha 300 kg4 kg sorghum 0.5 ha 500 kg 5 kg sorghum or10 kg groundnut 0.15 ha 100 kg 3 kg pearl millet 0.75 ha 550 kg

10 kg groundnut or8 kg cowpea 0.15 ha 50 kg

* An average 6-member household needs approximately 1 ton of grain per year** Cereals are planted at lower density in drier regions*** Officially recommended planting rate for maize is 25 kg/ha, but farmers commonly plant at 20 kg/ha

20

and several other countries in southernAfrica. The seed in question was probably alandrace variety found in the longer seasonhighlands of Manica Province. Poor seedquality, poor markings and bad labeling arebeing addressed through discussions withthe major seed houses. Protocols are alsobeing drafted, that stakeholders areexpected to adhere to.

Timing of Seed DistributionZimbabwe’s planting season somewhatuniquely extends over a three month period.Planting commonly begins just before rainsstart in late October. Pearl millet is oftenplanted into dry ground, while maize andlegumes are planted in the days immediatelyfollowing a major rainfall event. Novemberrainfall is usually highly variable in bothtiming and distribution. Rains are morecommon in December, so most planting isdone between 15 Dec and 15 Jan. Usually,planting is completed by midJanuary.However, rainfall in the 2003/04 season washighly variable, so plantings continued intoearly February.

Farmers generally plant legume crops early(before midDecember) and cereals over anextended period on multiple plots. The latterstrategy helps offset the risks of floweringduring a mid-season dry spell. Thisparticularly affects the productivity of maize.Sugar beans may be planted late in theseason, sometimes on the residual moistureavailable after the harvest of an early crop.

Since planting may begin as early as lateOctober, farmers want their relief seed bylate September. Strictly speaking, early seeddistribution contributes more to householdcertainty about seed availability, than to earlyplanting. If relief seed is delivered early, thefarmer does not have to purchase or borrowseed; this in turn provides a measure oflivelihood security.

Community Seed ProductionCommunity seed production programs havebeen widely promoted as a means tomultiply and distribute new (and traditional)varieties of limited interest to the commercialsector. Some of these programs encourage

All plants in this photograph were grown from seed labeled Macia – the tall plants in the backgroundproduce leaves but little grain, under Zimbabwean conditions

21

the community to produce OPV seed oncontract for a commercial seed company.But most involve production for localizedsale within the community – seed producersare expected to sell to neighborsexperiencing production deficits.

These programs have generally failed tomeet expectations. None has proved self-sustaining, and in most cases the willingnessof farmers to pay a premium price for seedfrom their neighbors has been over-estimated. Most programs findit hard to consistently obtainpure foundation seed after thefirst year or two of specialproject attention. Farmers areencouraged to adopt stricterquality control measures thanthey would normally use fortheir own seed crops, but suchmeasures are costly to policeand of uncertain value to localcommunities. Virtually all theseprojects have difficulty withseed sales. If production levelsare high, participating farmerscommonly ask the NGO topurchase their seed for sale to othercommunities.

Nonetheless, community seed productioncan provide a practical way to accelerate thedistribution of new varieties of limited interestto the private sector. In this case publicsubsidies to support the distribution offoundation seed, farmer training andtechnical supervision, may be justified asinvestments to speed the adoption of newhigh yielding varieties. NGOs should consultwith AREX or the international centers toidentify new varieties suitable for theseprograms. Potential varieties currentlyinclude SV 3 (white sorghum), PMV 3 (pearlmillet), Nyanda (groundnut) and IT 18(cowpea).

Draft PowerSeed distribution is often necessary to helpfarmers expand planted area. But the morebinding constraint to expansion is lack ofdraft power, particularly among the poorer

farmers. In fact, lack of draft resources is aprimary indicator of poverty. ICRISATsurveys in southern Zimbabwe show a closecorrelation between area of land planted andthe availability of cattle for plowing.Shortages of draft power also delay planting,which in turn reduces yields.

Data on the impacts of drought on livestocknumbers are limited. However, in more open-ended questions about input requirements,many small-scale farmers express a greater

need for tillage support than forseed. But this constraint isdifficult to resolve – restockingof cattle is expensive and timeconsuming, and contributes toovergrazing and landdegradation. Efforts to promotetractor hire schemes havebeen fraught with problems.Plowing services areexpensive, and tend first to beallocated to better connectedfarmers. Tractor drivers arepoorly trained in themaintenance and use of their

equipment. Again, farmers waiting for theseservices often fail to plant on a timely basis.

One solution is to encourage farmers toprepare their land earlier, soon after theharvest. Agronomic trials from bothBotswana and Zimbabwe consistently showhigh returns to winter plowing. During thepost harvest period, draft animals arestronger and can plow larger areas. Demandfor tractors is also low at this time, allowingmore area coverage.

Alternatively, Zimbabwe’s small-scalefarmers can be encouraged to prepare partsof their fields by hand. Here the experiencesof the Zambian Conservation Faming Unitare instructive. Planting basins can beprepared over an extended period followingthe previous harvest, allowing timely plantingwhen the rains eventually begin.

Fertilizer DistributionMost of the soils cultivated by small-scalefarmers, particularly in the drought prone

Community seed productionprograms have suffered arange of problems – non-viability, dependence oncontinuous external support;difficulties in marketing theseed produced, and inobtaining foundation seed.Nonetheless, communityseed production canaccelerate the distribution ofnew varieties of limitedinterest to the private sector.

22

regions, are grossly deficient in nutrients. Inmost years, lack of nitrogen is a more severeconstraint than lack of water. Consequently,even ‘improved’ varieties give only limitedyield gains. Average grain yields are lessthan 1.2 tons per hectare for maize, even infavorable seasons; and less than 800 kg perhectare for sorghum and pearl millet.

Farmers commonly complain that chemicalfertilizer is expensive. In drier areas theyclaim it is either not available or too risky.Some argue it burns the crop, or‘impoverishes’ the soil. In this context,national extension recommendations toapply three 50 kg bags of Compound D

Illustration from a Shona/Ndebele pamphlet on fertilizer use, distributedto 160,000 farmers across Zimbabwe

(commonly known asmaize fertilizer), andtwo 50 kg bags ofammonium nitrate(AN) per hectare areessentially irrelevant.

ICRISAT has collectedconsiderable data fromfarm managementsurveys in the drierparts of Zimbabwe.One surprising findingwas that many farmersfail to use the manureavailable to them –60% of cattle ordonkey owners did notuse any manure ontheir field crops.Instead, it was left torot in piles by the kraal.

Several years of on-farm experimentationhave shown that mostfarmers are simply

unaware of the benefits of fertilitymanagement. Training and on-farmexperimentation involving the use of smallquantities of manure and chemical fertilizercan significantly improve yields. However,this requires more than simply distributingfertilizer – farmers must be taught to apply itto an early-planted, well-weeded crop. In2003/04, approximately 160,000 farmersacross the country received 25 kg packs offertilizer with a pamphlet in the locallanguage (Shona or Ndebele depending onthe area) explaining how to use the fertilizermost efficiently. In the future, these effortsneed to be extended with on-farmdemonstrations and discussions.

23

7. Input Procurement

Zimbabwe has one of the more sophisticatedagricultural input markets in Africa. Thecountry has more than ten seed companiesand three retailers of chemical fertilizer.Wholesale and retail trade channels are wellestablished for inputs relating to commercialcrops. Despite this, during the past twoyears, large-scale relief programs havedifficulty sourcing inputs for distribution.Seed (except for maize) and fertilizer are inshort supply. Power lies withthe input seller rather than thebuyer. Correspondingly, qualityand timeliness of supply arecompromised; late buyers areforced to ‘take what they canget’.

Availability of InputStocksNGOs should first inventorythe market to see what inputsare available; and consult like AREX or theinternational research centers about therelative benefits of alternative products.Appendix 5 lists the varieties released onthe local market. However, many are notconsistently available. There are over 75white maize hybrids on the nationalregistration list, but only about ten of theseare readily available and suitable for reliefprograms. Only one of the four sorghumOPVs, and one of the three pearl milletvarieties is readily available, and only insmall quantities (relative to relief demand).No significant stocks of groundnut areavailable; Zimbabwe consistently imports anolder groundnut variety (Natal Common)from South Africa.

The country produces two main types offertilizer targeted for maize production.Compound D, now being advertised asmaize fertilizer, offers a mix of nitrogen,phosphorus and potassium (8-14-10). It issuitable as a basal dressing for most cerealcrops. Other compound or basal fertilizers

were available earlier, but production hasbeen discontinued.

Zimbabwe has long promoted the applicationof ammonium nitrate (AN) as a top dressingfertilizer for cereal grains. It is sold by twocompanies – Zimbabwe Fertilizer Companyand Windmill. A third company (Omnia) hassought to sell urea to relief programs, butfew Zimbabwean smallholders are familiarwith this input.

Input TendersMost NGOs either purchasetheir inputs through tender orobtain them through bulktenders issued bycollaborating partners. Severalcommon problems arise. First,most tenders specify the cropbut not the variety. This allowsseed companies to providewhatever variety is available.

This presents little problem in the case ofmaize because certified seed is readilyavailable. However, the buyer will still needto specify a preference for OPVs vs hybrids,or for varieties suited to drier, drought proneregions.

Purchasing seed of most other grain andlegume crops is more difficult. If varieties arenot specified, the buyer is likely to be offeredcommon grade seed of uncertain origin. Thisdoes not necessarily present a problem if itis purchased from local farmers. However,an opportunity may be lost to distributeimproved varieties of greater interest tomany farmers. In order to obtain thesevarieties, orders must be placed early andproperly specified.

Contracts should specify quality standardsfor the seed – germination rates, maximumpermissible content of foreign matter, and (ifa particular variety is specified) geneticpurity, ie maximum level of contaminationwith other varieties.

Zimbabwe has a strongprivate sector foragricultural inputs – butlargescale relief programshave difficulty sourcingseed (except for maize)and fertilizer. In a seller’smarket, quality andtimeliness of supply areoften compromised.

24

All seed production and trade in Zimbabwe isregulated by the Seed Services Department.All seed is supposed to be inspected in thefield and again in the laboratory to ensurephysical and genetic purity, germinationrates and seed health. This system worksreasonably well for maize, and for the smallquantities of other crops sold on thecommercial market. But the system cannotcope with the large quantities of seed beingsold and distributed under relief programs.As a result, seed stocks are of variablequality, and genetic (varietal) purity is often aproblem.

In this situation, buyers mustoften choose betweenpurchasing standard gradeseed and having little or noseed to distribute. This trade-off is difficult to judge. Ingeneral, common grade seedhas not harmed smallholderfarmers, and generallyproduces a fairly uniform crop.However, in areas where localseed stocks are more plentiful,NGOs are better off limitingtheir purchases to high quality seed ofimproved varieties and investing in fertilizeror tillage support.

Seed companies commonly need four to sixweeks lead time to make deliveries.However, NGOs should be aware that reliefseed and fertilizer are often delivered late,particularly if tenders are awarded late in thebuying season (due to delayed funding andother factors). This is due to limitations inseed cleaning and packing, and constraintsin the manufacture of chemical fertilizer.These difficulties have been worsened bythe decline in the overall economy. Buyersmay need to have clauses in their contractspenalizing suppliers for late delivery.Appendix 6 shows a sample contract.

StorageIt is important to arrange warehouse spacewhen tendering for large seed and fertilizerpurchases. Such stores may be hired in

larger business centers through rentalagreements with local wholesalers or bigretail shops. They should have waterproofconstruction, good ventilation, and amplespace for orderly stacking of commodities. Achecklist for assessing storage space isprovided in Appendix 7.

LabelingOne very common problem in reliefprograms is the lack of clear labeling ofagricultural inputs – and of seed in particular.This results partly from limitations in the

availability of printedpackaging. But in addition,most seed companies do nothave the capacity to labelthousands of smaller bags. Atbest, labels identify the varietyand lot number required totrace the origins of the seed.In many cases, even thisinformation is not available.Instead, farmers receive clearpackages with no informationabout the seed type or varietycharacteristics.

Lack of labeling is of particular concern inthe context of increasing distribution of OPVmaize seed. Zimbabwean farmers are usedto growing hybrids and will not necessarilyexpect to receive an OPV. If seed packs arenot labeled, most will simply assume thatseed must be repurchased the followingseason. This benefits the seed companies,but a development opportunity has beenlost.

Finally, good labeling allows NGOs to trackand document seed problems. Thecharacteristics of the crop can be checked inrelation to a variety name. The origins ofmixed or mis-labeled seed can be trackedwith a lot number. When problems areidentified, seed companies often blame thefarmer for contaminating their seed. But ifunopened, but labeled, seed packets arefound to be similarly contaminated, orotherwise deficient, it is easier to hold theseed seller accountable.

Zimbabwean regulationsrequire rigorous fieldand lab inspections ofseed. This system worksreasonably well formaize, and for the smallnon-maize commercialmarket – but not for thehuge relief market.Relief seed often hasproblems of quality orvarietal purity.

25

PackagingMost NGOs request that seed be packed in aform suitable for delivery to individualhouseholds. A set of different seed crops maybe packed into a larger bag for distribution toeach farmer. Or smaller packets of severaldifferent crops may be distributed individually.The former strategy allows faster distributionin the village, and is moreconvenient for farmers (easierto carry a single 15 kg bag thanseveral smaller bags).

The latter strategy is moreflexible, allowing the NGO toadjust delivery strategies asneeded. For example, onecommunity may grow moresorghum, while a neighboringvillage may grow more pearlmillet. Smaller package sizesincrease flexibility, but may also make thelogistics more complex.

In areas where localseed stocks are plentiful– even if the quality isvariable – NGOs couldconcentrate theirresources on distributinghigh quality seed ofimproved varieties andinvesting in fertilizer ortillage support.

In a few cases, seed companies deliver inlarger 50 kg bags, which the NGO repacksinto smaller units. This may reduce the priceof seed and improve delivery schedules,because the major bottleneck is in packingseed in smaller lots. However, the time andeffort required for repacking seed should notbe underestimated; and the NGO itself mustprovide some sort of labeling.

During the 2003/04 croppingseason, 160,000 farmerswere each provided with 25kg of ammonium nitratefertilizer. The standardfertilizer bag is 50 kg. Therewas no way to repack it insmaller bags withoutsignificant additional expenseand long delays. Ultimately,two farmers were asked toshare each 50 kg bag. The

success of this program is currently beingevaluated.

26

8. Timing of Input Distribution

There never seems to be enough time tocomplete every objective of a wellfunctioning and well monitored agriculturalinput relief program. The funding arrives late,new staff members are difficult to find, andrequire more training and supervision thananticipated; and inevitably one problem oranother delays the implementation of theprogram. Program planning and tracking canbenefit, at least, from the creation of acalendar of activities. These need to belinked, ultimately, back to the farmingcalendar commonly employed byZimbabwe’s smallholders.

Farming CalendarMost smallholder farmers in Zimbabwe planttheir main field crops over a single summerplanting season. This usually starts just beforethe onset of rains in late October and endswith the completion of harvest in June or earlyJuly of the following year. Table 2 provides arough calendar for the country as a whole.

Maize is commonly planted more than oncein order to offset the risks of mid-seasondrought. Therefore one plot of maize may be

planted in November and a second plot inlate December. Groundnut and cowpea tendto be planted earlier. Cereal grains are morelikely to be planted later. Poorer householdswith limited access to draft resourcescommonly extend their planting into January,and occasionally to early February. This isbecause they may wait to borrow draftresources from neighbors who first want tocomplete their own planting. Most farmerswould prefer to plant late than to till their landby hand.

According to this calendar, relief seed oughtto be delivered to farmers by late October.However, if supplies are late, seed can stillbe delivered and productively used by poorerhouseholds in late December or earlyJanuary. Basal fertilizer similarly needs to bein the hands of farmers by October, thoughtop dress fertilizer may be distributed in lateNovember or December.

Program CalendarTables 2 and 3 show a possible calendar forrelief operations, in relation to the farmingcalendar.

27

Tabl

e 2.

Sum

mer

sea

son

farm

ing

cale

ndar

Tabl

e 3.

Cal

enda

r for

relie

f inp

ut p

rogr

am

June

July

Aug

Sep

Oct

Nov

Dec

Jan

Feb

Mar

Apr

May

June

Identi

fy ar

eas n

eedin

g as

sistan

ce

Quan

tify a

ssist

ance

nee

ded

Prop

osal

for fu

nding

Tend

er fo

r inp

uts

Orga

nize

deliv

ery s

ystem

Regis

ter re

cipien

ts

Distr

ibute

seed

and

bas

al fer

tilize

r

Distr

ibute

top d

ress

fertil

izer

Asse

ss d

istrib

ution

Asse

ss u

se o

f inpu

ts

Evalu

ate p

rogr

am im

pacts

* F

arme

rs in

Zim

babw

e co

mmon

ly ap

ply b

asal

comp

ound

s only

afte

r ger

mina

tion,

to re

duce

inve

stmen

t risk

s

Oct

Nov

Dec

Jan

Feb

Mar

Apr

May

June

July

Aug

Sep

Summ

er p

lowing

Plan

ting

Wee

ding

Basa

l fertil

izer

Top

dres

s fer

tilize

r

Harve

st

Thre

shing

Wint

er p

lowing

Legu

mes f

irst,

cere

als la

ter, e

xcep

t for

gree

nma

ize h

arve

st

Usua

lly 1

mon

th o

r mor

e af

ter g

ermi

natio

n,up

to flo

werin

g

At pl

antin

g or 2

-3 w

eeks

after

germ

inatio

n *

28

9. Distribution Strategies

Relief programs in Zimbabwe haveexperimented with several strategies forseed distribution in recent years. In manycases, direct distribution of free handouts isbeing replaced with more market orientedstrategies. These include:• Credit programs requiring farmers to

deliver a portion of their harvest inexchange for inputs

• Vouchers redeemable for inputs ‘sold’through rural retail shops

• Seed fairs where vouchers are exchangedfor seed supplied by commercial andinformal traders.

The advantages and disadvantages of theseapproaches are still being assessed. Ingeneral, efficient input distribution isexpected to:• Minimize errors of inclusion/exclusion (ie,

assisting non-deserving households orleaving out deserving households) at thebeneficiary identification and selectionstage

• Provide farmers with inputs for which theyhave the agronomic knowledge and skills,and which relate to their crop productionpreferences. In addition, new products canbe introduced provided farmers are givenrelevant training

• Enable farmers to receive inputs in atransparent and corruption-free fashion

• Minimize administrative costs of delivery• Minimize donor dependency• Minimize the disruption of input markets

and where possible, facilitate marketdevelopment.

Ultimately, multiple strategies may beneeded in order to reach different segmentsof a population. The choice of strategy maydepend on the capacity of NGOs, thestrength of local markets, and the politicaland economic environment.

Direct Free DistributionThis is the traditional method – small packsof seed and related inputs are distributeddirectly to individual farm households.Recipients are first selected and registered

(see Chapter 4). The registered householdsmay receive an identification card or simplyhave their national identification cards listed.When seed or fertilizer is available, they areinformed and mobilized.

Distribution is done in coordination with localdistrict and ward authorities. NGO staff mayaddress a meeting explaining the distributionprocedure. Beneficiaries are grouped intovillage units, the names checked against theregistration lists, and recipients’ identitiesconfirmed by village officials. Thebeneficiaries are then asked to sign for theirinput package before collecting it. SomeNGOs ask for additional confirmation: thebeneficiary signs a statement acknowledgingreceipt, and/or local officials or AREX staffcounter-sign to confirm that the seed hasbeen collected.

The key advantage of this strategy is itssimplicity. Communities are mobilized in thesame way as for food aid distribution, socommunity leaders and participatinghouseholds are familiar with the process.Some argue this method is cheaper toimplement than some of the alternativemethods described below. However, a strictanalysis of the costs of alternativedistribution strategies would need to be doneto confirm this.

Input Delivery on CreditIn the main government program of inputdelivery, seed and fertilizer are distributed oncredit through the Grain Marketing Board(GMB). In effect, farmers receive seed andfertilizer in return for a promise to sell theirgrain to GMB at the end of the season. In2003/04, farmers had to show they had soldgrain to GMB in previous years. Farmerswho had repaid their earlier credit for inputscould essentially obtain what inputs theydesired.

Little information is available about thedetails and success of this program. Anunknown proportion of farmers receiving

29

credit in 2002/03 did not repay, andtherefore did not qualify for loans thefollowing season. The requirement ofsales to the GMB effectively restrictsparticipation to better-than-averagefarmers who had surplus graindespite the drought last season.Correspondingly, very few farmers indrier parts of the country havequalified for these loans.

The advantage of this program is itreduces dependency on free inputs.The terms of the loan are moreliberal than terms offered through thebanking system; so farmers can builda credit rating while bearing onlylimited risks. The imposition ofmarket controls appointing GMB asthe sole buyer in the country, hasincreased the likelihood ofrepayment. But this may be difficultto enforce.

A variant of this strategy is for anNGO to take responsibility forcollecting grain after harvest. This isnot currently legal in Zimbabwe, buthas often been used in neighboringcountries. The main difficulty is thelogistics of collecting small amounts of grainfrom large numbers of farmers – collectioncosts may be higher than the value of thegrain. And farmers have an incentive toprovide poor quality grain to meet numericalrepayment quotas.

Some NGOs have also asked farmersreceiving inputs (particularly seed) to returna small quantity of seed from their grainharvests. The problem is that the returnedseed is of uncertain quality. It is difficult toassess seed quality in the field and the‘seed’ being returned may be actually grain;of mixed varieties or with low germinationrates.

Seed FairsCatholic Relief Services (CRS), incollaboration with ICRISAT, have developed

Seed fairs have proved effective and popular in severalcountries, including Zimbabwe, Kenya, Uganda and Sudan

1. CRS, ICRISAT and ODI. 2002. Seed vouchers and fairs: a manual for seed-based agricultural recovery in Africa.Nairobi, Kenya: Catholic Relief Services, ICRISAT, and Overseas Development Institute, UK.

a strategy of using community seed fairs todeliver relief seed to small-scale farmers.The main objective of these fairs is to givefarmers a choice of seed types and varieties.In particular, farmers can obtain traditionalvarieties from other farmers or local traders,instead of unfamiliar modern varieties. CRSand ICRISAT have written several manuals1

describing how to implement seed fairs.

Farmer beneficiaries are first identified andregistered. Arrangements are then made fora market day (fair) where anyone wanting tosell seed can participate: formal seedcompanies, local small-scale retailers, andlocal farmers with surplus seed. Eachbeneficiary is given a set of vouchers(different denominations for convenience)using which they can ‘buy’ seed from anyseller at the fair.

30

On the day of the fair, traders are registeredand allowed to display their seed stocks.There is no restriction on formal sectorparticipation. Some fairs are open to thoseselling not only seed but other inputs such astools or fertilizer. Some fairs may restrict thenumber of informal traders (ie local farmersselling seed), to keep the fair of manageablesize. In most fairs in Zimbabwe, seed pricesare set in advance, eg standard prices for abag or cup of cereal/legume seed. In othercountries, prices are left to market forces. Theseed prices being set at fairs in Zimbabweare commonly higher than seed prices onlocal markets. This is viewed as necessary toencourage more traders to participate, but italso contributes to seed price inflation.

Informal traders are expected to have theirseed inspected for quality. However, this issometimes difficult for field staff trying tomanage large numbers of trader and buyers.Good advance planning is essential.

The total quantity of seed offered by eachtrader is weighed and recorded both at thebeginning of the fair, and at the end. Thishelps the organizers monitor prices and limitcheating.

Farmers must spend all their vouchers onthe day of the fair. After this the vouchers areworthless.

Seed fairs essentially build on the recognitionthat local seed stocks are usually available,often in substantial quantities, despite floodsor drought. Village seed markets continue tooperate. The fairs are predicated on theassumption that the main constraint facingvulnerable households is their inability topurchase the available seed. This assumptionmay merit further investigation.

One of the main gains from the fair is asubstantial infusion of money into the ruralcommunity. This may have multiplier effectsas seed sellers use their earnings topurchase labor or other village-producedcommodities.

The main disadvantage of the fair is thecomplex logistics. The organizers neverquite know how much seed will be available

until the fair opens and may consequentlyset seed prices higher than needed. Little isknown about the impacts of seed fairs onlocal seed markets. Surveys in Zimbabwesuggest that fairs are monetizing atransaction that would normally take the formof a gift. Poorer households normally obtainfree seed from their neighbors followingdrought years; and are expected toreciprocate when fortunes are reversed. Theintroduction of external institutions and largeamounts of money to finance seedtransactions may undermine suchcommunity obligations. However, thesignificance of this problem is unknown.

Vouchers Redeemable at RetailShopsCARE is implementing a program similar toseed fairs, where vouchers are distributed totarget beneficiaries, and are redeemable forinputs at local retail shops. This takesadvantage of the substantial marketinfrastructure in Zimbabwe’s rural areas.Shops that regularly sell hybrid maize seedcan be found in most rural communities. Butthis retail trade is severely compromised by thefree distribution of relief seed. Even theprospect of such programs discouragesretailers from stocking seed. Similarly, seedcompanies are discouraged from distributingthrough wholesale and retail channels if theycan more easily, and perhaps more profitably,sell in bulk to a few relief programs.

The voucher program is a natural extensionof CARE’s recent efforts to provide businesstraining to rural retailers, in order to improvetheir investment practices andentrepreneurial skills. In non-drought years,commercial suppliers were encouraged toprovide inputs on credit to retailers thustrained. CARE agreed to provide monitoringsupport. A similar program of businesstraining was implemented by the Citizen’sNetwork for Foreign Affairs, in a programultimately managed by AGMARK.

Though this is a more market friendlyapproach to the delivery of relief assistance,CARE still takes responsibility for buying allinputs. This has remained necessary, partly

31

because suppliers no longer provide inputson credit to retail traders, in turn because ofhigh rates of inflation in Zimbabwe (over500%).

When inputs are available, they aredistributed to the designated retailers, andfarmers are notified to trade in their vouchersfor a mandated package. The farmers mustpay a small fee to cover the costs of storageand handling by the local retailer. This is saidto reduce dependency on free inputs andensure that farmers recognize the value ofwhat they are receiving. Even the poorestfarmers seem to have little difficulty payingthese service fees. The income earned bythe retailer is expected to be reinvested intothe business.

In the future, farmers may be given morechoice of what inputs to purchase. This iscurrently not possible because of the

logistics. The seed and fertilizer packagesbeing distributed must match the value of thevouchers being redeemed. Also, seed mustbe reserved in bulk to ensure it is available.But subsidies empowering poorer farmers tomake choices are probably more efficientmeans to provide agricultural assistancethan mandated purchases.

The main disadvantage of this program isthe complex logistics. Retailers must beidentified and trained, vouchers distributed,and farmers organized on the day of inputdelivery. The main advantage is this createsor strengthens market linkages that willcontinue to function after the relief programhas ended. As such, this may be the mostsustainable of the delivery strategiescurrently being implemented in Zimbabwe.

Table 4 summarizes the advantages anddisadvantages of these programs.

Table 4. Advantages and disadvantages of alternative distribution strategies

Advantages Disadvantages

Free, direct distributionSimilar in logistical requirements to food aid delivery Provides no choice of inputsMost NGOs are familiar with the procedures for free input delivery Undermines retail trade of inputsLow establishment costs Creates dependency on free handoutsCredit programsReduces dependency on free handouts insofar as farmers have May undermine formal credit systems if credit isto repay loans subsidizedProvides some choice in the type and quantity of inputs to be Risky to administer in drought prone regionsobtainedSeed fairsProvides farmers with a choice of inputs to be purchased High start-up costs in staff training and communityEncourages local seed producers to expand their production organizationEncourages development of informal, community seed market May undermine local seed marketsBrings cash into the rural community Inflates local seed prices

Input availability is not guaranteed; need to check ifseed is really available on local markets; can bedifficult to determine if farmers are hiding stocks inorder to qualify for handoutsElderly, disabled may have difficulty obtaining seed ina crowded fairMay increase dependency on external interventions

Vouchers redeemable at retail shopsEncourages development of wholesale and retail input markets High start-up costs in organizing and training retailMay provide choice depending on how the program is run tradersReduces risks of stocking agricultural inputs Still unclear how much collateral investment will be

made by input supply companies in developing suchretail tradePossibly prone to corruption – eg trader providesonly partial allotments or asks for bribes

32

10.Technical Support for Agricultural Recovery

Many NGOs involved in agricultural reliefand recovery programs provide sometechnical support to complement the deliveryof inputs. The support may include farmertraining on varieties and crop managementpractices, implementation of demonstrationtrials, or distribution of informationpamphlets. The key issue is: how muchshould we invest in theseprograms? And as a corollary,how to build linkages withpartners who have thenecessary technicalcompetence?

Information aboutSeed VarietiesRecent surveys have shownthat too many recipients ofrelief seed do not know whatvarieties they are receivingfrom NGOs. The bag is sometimes labeledwith the variety name; but this alone ismeaningless without additional information.Few farmers will know the differencebetween the maize hybrids SC 403 and DK8031; or between the OPV ZM 521 and thehybrid SC 501. Often, the seed packet is noteven labeled. In such circumstances,farmers may be well advised to only plantpart of the seed to check its performance.

A huge opportunity is lost if farmers lack theinformation needed to use their inputseffectively. If the unknown variety performspoorly, relief seed becomes suspect. If thevariety performs well, farmers may beinterested in purchasing it next year – butcannot, because the identity is unknown; theopportunity to develop a market is lost.Clearly, relief programs in Zimbabwe mustallocate more investment to seed labelingand farmer education.

Information about FertilizerAt least 80% of farmers in drought proneregions (Natural Regions IV and V) have

little experience with chemical fertilizer. Theymay have received free fertilizer from a reliefprogram, but will rarely purchase it. As aresult, when fertilizer is distributed, it is usedinefficiently. Some farmers will reject italtogether, worried that it ‘worsens theirsoils’. Others may sell it to traders or toneighbors with irrigated plots. There are

stories of farmers who use thefertilizer to paint their house orkeep snakes out of the yard!

In most cases the problem islack of technical information onhow to apply the limitedquantities provided throughrelief programs. Manyagronomists, even, havedifficulty answering the questionof how best to allocate 25 kg ofammonium nitrate. Should it bespread broadly, or concentratedon a small portion of the crop?

Should it be used on maize or sorghum?Should it be buried or broadcast?

Most NGOs rely on AREX extension staff toprovide advice on fertilizer use. But theycannot simply assume that this advice will beprovided in the normal course of extensionprograms. Rural extension programs do nothave enough resources to make regularvisits to small-scale farmers. The majority ofthese farmers never see an extension agent.And many extension staff have little trainingrelating to severely resource constrainedsystems. Few can readily answer thequestion of how best to apply smallquantities of fertilizer, other than to suggestthat it ought to be applied with the rains.Training of extension staff may be needed.

During the past few years, ICRISAT hasbeen working with provincial and districtAREX staff to test alternative strategies forallocating small amounts of chemicalfertilizer. This work is founded on theassumption that most farmers adopt costlytechnologies in a stepwise fashion. They willstart with a small investment in fertilizer

Most relief recipients lackthe information needed touse the inputs effectively.This is a huge opportunitylost. Relief programsmust expand investmentsin farmer education –seed labels and whatthey mean, cropmanagement, fertilizerapplication rates andmethods.

33

Early gains in drought-prone environments

Investmentgrowth

Researchrecommendations

Typical N response curve – Zimbabwe semi-arid areas

Official fertilizer recommendations are placed at the high end of the nitrogenresponse curve – recommendations at a lower level would be more practical,and give higher returns on investment

before considering themuch higherrecommended rates.The willingness ofthese farmers toexpand usagedepends on thepayoffs to the initialsmall investment.

The key to thisprocess is goodtargeting. In 2003/04,ICRISAT worked withthe NGO community todistribute an illustratedone-page pamphletdescribing how best toallocate exactly thequantities of fertilizerbeing provided through one major donor. Itexplained the need to target fertilizer to theplant, apply on a timely basis relative to bothrainfall and crop growth, and to control weedgrowth. Every farmer received a pamphletalong with his/her allocation of fertilizer.

While the impacts of this effort are still beingmeasured, at least some farmers havelearned something more about a key cropmanagement practice.

Demonstration TrialsSeveral NGOs support demonstration trialsto promote improved crop management

On-farm trials in the semi-arid Matopos district in Zimbabweshow significant gains from a single bag of fertilizer

technologies. World Vision, for example,manages an extensive training program for‘lead’ farmers who are chosen by thecommunity to work with and be trained byAREX officials. These lead farmers thenestablish a demonstration plot in the village,and in turn train five to 10 farmers on cropmanagement methods. These Field SchoolGroups (FSG) are centrally located toencourage more farmers to participate. Thepayoffs to such demonstration programsmerit further assessment.

Sources of Technical SupportAREX is generally the primary source of

technical support, capable ofbackstopping relief and recoveryprograms. Indeed AREX staff arelocated in most of the country’swards and have a mandate forfarmer training. Once again,however, NGOs should note thatmuch of the training received bythese staff applies to relatively high-input and more commercializedsystems. AREX staff themselvesmay benefit from additional trainingrelating to low-input systems.

ICRISAT has been working with theNGO COSV to train extensionworkers on low-input fertility

34

management strategies in Hwange district.Similar programs of training anddemonstration trials have been launchedwith World Vision and CARE. There may bescope for obtaining support from otherinternational agricultural research centersbased in Zimbabwe, including CIMMYT andICRAF.

Seed companies should be encouraged towork more actively with NGOs involved inseed distribution. The companies should beencouraged to view the distribution of newvarieties as an investment in building amarket; provide farmers with moreinformation about their varieties; and

The majority of relief beneficiaries lack information on how best to use the inputs received. Farmertraining must be better integrated into relief budgets

perhaps sponsor demonstration trials andfield days. Fertilizer companies may besimilarly encouraged to invest instrengthening extension programs relating totheir products.

Finally, NGOs need to take advantage of thewealth of experience with Zimbabwe farmingsystems and input delivery available withintheir own community. During the past year,CRS has provided training to staff of otherNGOs about how to implement seed fairs.But such initiatives are rare. An inventory ofspecialized programs within the NGOcommunity could reveal an unexpectedwealth of expertise.

35

11. Integrating Recovery Efforts with Development

Most donors have separate budgets for reliefand for development. Funding for reliefprograms is for short term investments thathelp people keep themselves alive (food aid)and re-establish their livelihoods (agriculturalinputs). This is distinct from developmentprograms with medium to longer term impacts.

This distinction derives from two factors.Some argue that the pursuit of developmentgoals undermines efforts to provide shortterm welfare gains to the most vulnerablehouseholds. Such questions have arisen inrecent discussions about seed and fertilizerallocations. Should they be allocated tohouseholds capable of producing more foodper unit of input? Or to the poorest and mostvulnerable households – who are likely toachieve lower yields (perhaps because ofdraft power constraints), but who wouldotherwise find it difficult to re-establish theirfarming activities?

The second factor is largely political. Somedonors are willing to provide Zimbabwe withhumanitarian assistance, but not developmentassistance. Relief programs may be funded,but not development programs.

In most cases, however, it is impossible tomake a strict distinction between relief anddevelopment activities. Relief interventionsinevitably contribute to communitydevelopment – by introducing new varieties,improving rural incomes etc. Two questionsthen arise:• How can we deliver agricultural aid in ways

that are least disruptive of longer termdevelopment efforts, but still contributemost to household food security?

• How can longer term development effortsbe linked with the provision of reliefassistance?

In Zimbabwe, this debate has arisen in thecontext of several initiatives:

1. Seed market developmentFree seed distribution, whether throughdirect distribution, through seed fairs or

through vouchers redeemable at retailshops, inevitably distorts seed markets.Large scale purchases by NGOs encourageseed companies to hold seed off the marketin the hope of bulk sales to relief programs.Seed packing equipment is tied up servingthe subsidized trade. Free, direct distributiondiscourages wholesalers and retailers fromstocking seed, even if it is available throughcommercial seed companies.

The need for emergency assistance mayjustify short term interventions that disruptlocal markets. However, NGOs may alsoconsider a range of strategies for makingrelief distribution more market friendly. Forexample, tighter targeting of assistance topoorer households with limited capacity tobuy from the commercial market. In general,voucher type programs, though harder toorganize, can provide an element of choicewithin the context of existing marketinstitutions. One key question is how muchchoice is logistically feasible.

The seed trade may also benefit frominvestments in the establishment of seedsecurity stocks. There have been noticeableproblems of seed quality, for crops otherthan maize. One reason is that following afavorable season, commercial markets forthese seed crops are perceived to be limited.Seed stocks remain low, and when demandsharply increases, little is available. Adevelopment investment in building seedsecurity stocks may have a high payoff whenthe next emergency occurs.

2. Community seed productionAfter every drought, there is renewedenthusiasm for community or local seedproduction schemes. These are promoted asmeans to improve seed security within ruralcommunities. In most cases, communityseed production is linked with themultiplication and distribution of newvarieties. But farmers are also beingencouraged to multiply and maintain stocksof favored traditional varieties.

36

Some relief programs also distribute farming implements, to encourage adoption of improved cropmanagement practices

Community seed schemes potentially offer agood means to support the distribution ofvarieties of limited interest to commercialcompanies. However, NGOs need toconsider the fact that these schemes dependon continuing financial and technical supportfrom the public sector. The assumption thatfarmers will maintain community seedproduction programs on their own hasproved unfounded. And there are problemsof foundation seed supply, quality controland marketing that require careful planning.

3. Conservation agricultureThe objective of Conservation Agriculture orFarming (CF) in Zimbabwe is to enablesmallholders to adopt more productive andenvironmentally sustainable farming systemsthat improve yields, food security and soilfertility.

A CF Task Force has been set up inZimbabwe, representing the majorstakeholders, and has designed a CFpackage allowing flexibility according to

agro-ecological zone. This starts from thepremise that soil should be managed in amore sustainable manner. The Task Forcerecommends no plowing (reduced tillage),maintenance and incorporation of cropresidues, early planting at appropriate plantpopulation, timely weed control, androtations, preferably with legumes.

The package initially requires a largeinvestment of family labor in land preparationand weed control. Farmers without draftresources are encouraged to dig plantingbasins during winter, before the rainyseason. Those with draft power can simplycreate planting lines or furrows. Plantingbasins or furrows concentrate early rainfallaround crop seeds, accelerating emergenceand improving plant stands. Fertilizer andmanure are more carefully placed near theplant, making more efficient use of availablewater and nutrients. Retaining crop residuesreduces surface temperatures and over time,improves soil fertility. Deep rooting crops canbe used in a rotation to break pans andaerate the soil. Rotations with legumes

37

reduce the quantities of mineral fertilizerneeded.

Weed burdens tend to be high in the first yearor two. However, weed populations quicklydrop as the limited tillage reduces thegermination of weed seeds. Over time, lesslabor is required than in conventionalproduction systems, soil quality improves, andyield levels increase, even in drought years.

4. Commercial markets foragricultural productsSome relief programs explicitly promote theproduction of more drought tolerant crops. InZimbabwe, for example, the production ofsorghum and pearl millet is encouraged asan alternative to maize in drier parts of thecountry. To reinforce these efforts,commercial linkages may be pursued withgrain traders and processors. But NGOs

need to be aware of the difficulties ofcommercial market development in droughtprone regions. The very likelihood ofdrought, and the associated variability inproduction levels, raises marketing costs. Ifa significant surplus is only available everyother year, trade infrastructure isunderutilized. In addition, low populationdensities in many drier regions lead to highersearch and assembly costs.

The best option may be to establish linkageswith private traders willing to take on most ofthe investment risks involved in commercialmarket development for drought tolerantcrops. Relief programs can play a significantrole by distributing improved seed ormanagement advice for targeted crops. Theymay facilitate crop assembly or informationsharing about market opportunities. Theseefforts can reduce the uncertainty and costsof the private trader’s investment.

Developing links between smallholder farmers and commercial markets is difficult – but can provide highpayoffs to relief investments

38

12.Relief and Recovery in Livestock Systems

Livestock are an important and integral partof Zimbabwe’s agriculture economy,contributing 15 to 25% of the value ofagricultural output in all farming sectors.Cattle production accounts for more thanhalf this output. Goats, sheep, pigs andpoultry are also widely important. Donkeysare important sources of draft power in thedrier parts of the country.

Most livestock production in the communalsector remains semi-subsistence.Reproduction rates and offtake are generallylow. Cattle are valued as much for theiroutputs – for tillage, transport, milk, manureand meat – as for their cash sale value.Goats and sheep are commonly maintainedfor smaller cash needs – to be sold orslaughtered when school fees need to bepaid, or for household consumptionexpenditure. Almost all households maintaina few chickens as a source of protein and forcash purposes; poultry is of considerableimportance to poorer households.

Despite the importance of livestock in thesmallholder economy, investments in reliefand recovery programs targeting this sectorhave been limited. Past governmentprograms have sought to encourage de-stocking, and the use of state landsprovidingsupplementary accessto water and feed.During the past year,the governmentdistributed feed rationsto a small cross-section of districtsworst affected bydrought. SeveralNGOs supplementedthese efforts with re-stocking programs forcattle and goats.

A study conducted inearly 2004 reviewedoptions for expandingthese interventions.

Several of the recommendations from thisstudy are summarized here.

Livestock TargetsNational development efforts tend toemphasize production and veterinarysupport for cattle. But if a relief program is toassist poorer households, more attentionmust be directed to other species. Donkeysare relatively tolerant of drought and are akey source of transport and draft power.Small stock are important sources of foodprotein and petty cash income for essentialneeds such as medicines and school fees –especially in times of drought. Programstargeting improved poultry husbandry maybenefit the poorest small-scale farmers.

Animal FeedFeedlots have been used in the past toprovide drought relief to livestock producers.However, these favor cattle, which usuallybelong to the wealthier households. As such,they should be managed on a cost recoverybasis; farmers could pay for feed either incash or in the form of livestock.

Management of communal feed resourcesis a key part of drought mitigation

Livestock are a key component of farming systems, especially in drier areas.Interventions must take account of crop-livestock synergies

39

programs. Once a drought occurs, plansshould be in place allowing access to statelands for grazing of selected groups ofanimals. Relief programs could assist withthe collection of hay from roadsides orfodder from various browse species. Reliefand recovery programs could alsoconcentrate on delivering feedsupplements to support the maintenanceof a few key animals in the homesteadkraal. These include items such as oilseedcake, sugarcane tops, liquid molasses,salt, urea, and high-energy protein cubes.Bran from the major mills could betargeted to poultry.

Farmers report that they selectively allocatefeed to animals of particular value.However, many also claim to simplyabandon animals that appear weak in orderto focus their limited resources on strongeranimals more likely to survive drought.Further training can improve this decision-making process.

Veterinary ServicesImprovement of animal health should also berecognized as a mitigation strategy. Healthieranimals can better withstand shortages offood and water. At the onset of a droughtpoorer households with fewer animals couldbe targeted to reduce the vulnerability ofremaining stock. Vouchers can be providedto poorer households to help them purchasefeed and veterinary pharmaceuticals.

Water AccessDroughts increase competition for dwindlingcommunity water resources. As surface waterruns dry, animals and people turn to boreholewater. In this context, underground watersupplies need to be evaluated and monitoredto promote sustainable use. NGOs can providelogistical support for village level watermanagement committees, including assistancein repairing and maintaining boreholes.

Production SupportMost communal farmers have limited accessto animal husbandry advice. Technical

support for the production of donkeys andsmall stock is particularly limited. NGOs canassist AREX authorities with thedevelopment and dissemination of advice onhousing, feed and nutrition, wateringregimes, animal health, and reproduction. Inall cases, advice should take account of thevalue of indigenous breeds.

De-stocking and Re-stockingFarmers are commonly advised to de-stock animals at the onset of drought inorder to reduce pressure on limited feedand water resources, and thereby reduceanimal deaths. However, few householdsheed this advice because of the low pricesat such times, and the difficulty of re-establishing herds. These problems areaggravated by underdevelopment oflivestock markets and lack of marketinformation in many communal areas. Oneoption is to more closely link de-stockingefforts with re-stocking support after thedrought. Poorer households sellinganimals could be given vouchers topurchase new stock after the drought.Efforts to promote re-stocking should onlybe initiated when water and feed resourcesare replenished. These should not rely onexotic breeds. Poorer households withlimited capacity to invest in feed andmedicines may be better off withindigenous animals.

Limiting the Need to Sell SmallStockCommunal farmers rely heavily on the saleof small stock for the purchase of basicnecessities. Consequently, households withfew animals risk falling into a poverty trap.When these animals are sold, it becomesdifficult to re-establish herds or flocks. Thesehouseholds can benefit from targetedassistance protecting these minimumholdings from sale. Simple interventionstargeted at poorer households can reducetheir obligations to sell small stock. Theseinclude food aid, external support forpayment of school fees (or a moratorium on

40

school fees), subsidies for medicines forboth animal and human use, andsupplementary protein.

Monitoring SystemsA recent assessment of Zimbabwe’s livestocksector revealed large gaps in the informationavailable about livestock population levelsand dynamics. This makes it particularly

difficult to plan livestock interventions. Prioritymust be placed on improved data collectionand reporting systems for livestock and inparticular for all non-cattle species; improvedinformation on water resources available forlivestock; improved information on livestockfeed availability; and stronger monitoring andevaluation systems for relief and recoveryinterventions.

41

13.Monitoring and Evaluation

Monitoring and evaluation programs will bedesigned in different ways depending onwhether the objective is to:• Document the implementation of the relief

program• Document the allocation of relief

resources, eg how much seed waspurchased and distributed to how manybeneficiaries

• Measure the primary impacts, eg whetherthe seed was planted, how muchadditional land was planted and how muchgrain was harvested; or

• Document secondary impacts, eg didfarmers understand what varieties theywere receiving, how were these judged,was seed saved for the next plantingseason.

Some monitoring and evaluation plans alsoseek to collect a wider range of informationabout such issues as alternative livelihoodstrategies, household asset levels, or theimpacts of HIV/AIDS.

There is no single correct method forconducting a program assessment. Rather,there are many different strategies forconducting different types of evaluations.Some may involve simple checklists to trackthe flow of resources. One example is thewarehouse checklist found in Appendix 7.

Alternatively, we may use a brief one- tothree-page questionnaire focusing on a fewkey indicators. Open-ended discussions maybe used to identify problems with fieldpractices. Longer formal questionnaires maybe used to assess program impact. Each ofthese methods may be used with differentstakeholders at different times.

The FAO Emergency Unit has created aspreadsheet to monitor the distribution ofagricultural inputs by district and ward. Thisaims to assess input distribution coverage inthe country and highlight areas of overlap.The success of this effort depends on thetimeliness and accuracy of the data providedby NGOs. The data collection form for thisspreadsheet is shown in Appendix 8. Some

of the data collected in 2003/04 are shown inAppendix 9.

The value of a program evaluation dependson the quality of the data collection andanalysis. Consideration of a few basicprinciples can considerably improve thequality of these assessments.

1. Clearly define the specificobjectives of the evaluationIf the objectives are unclear, the chances ofcollecting data sufficient to assess the programare sharply reduced. Is the objective to trackthe allocation of relief funds, to assess howfarmers used the assistance, or to evaluate theimpacts of this assistance? One way to clarifythese objectives is to frame hypothesesidentifying the impacts to be measured.

2. Define the indicators by whichthese objectives will be measuredIf performance indicators are not explicitlyspecified, the probability of measuring anygiven impact is sharply reduced. Whatindicators should be used, for example, tomeasure whether relief seed contributed to anexpansion of cropped area? If seed isobtained from various sources, these must bedistinguished in the data collection. Ideally,data may need to be collected from similarsets of households who did or did not receiverelief seed. What indicators should be used toestimate whether relief inputs contributed tofood security? Again, estimates of totalproduction will be inadequate if seed isderived from different sources. Complicationsarise in distinguishing the contributions ofrelief inputs (eg seed and fertilizer) from thecontributions of management factors such astimely access to draft power.

3. Select the data collectionmethod to be employedPerformance data can be collected in manydifferent ways. Two common formats are (a)

42

formal, survey questionnaires and (b)informal, open-ended interviews. The latterinclude procedures like focus groupdiscussions, participatory surveys andtransect walks. Formal, quantitative surveysare most suited for the collection of data thatare quantifiable – for example, quantity ofinputs received, quantity of inputs used andthe amount harvested.

Informal surveys are more suited to thecollection of data that are difficult to quantify,such as opinions about the value of aprogram. They are also better for identifyingproblems. Often the twosurvey methods are used incombination. The informalsurvey may be used to helpfocus a formal survey. Andinformal interviews may thenbe used to learn more aboutproblems identified byquestionnaire responses.

Formal quantitative surveysare considered to be moreexpensive and time consuming thaninformal surveys. But this is not alwaystrue. In either procedure, much dependson the characteristics of the sample frame,and the sorts of information beingcollected.

4. Frame questions so that thechosen indicators can beefficiently and unambiguouslycollectedMany monitoring and evaluation programsstart by drafting a generic questionnaire oninput use and production levels. These sortsof surveys are highly prone to attributionerrors (eg was the seed from a reliefprogram or from the farmer’s own stock) orfailure to collect data critical to the evaluation(eg what is the value of the additional grainat the farmgate). A better practice is to crafta set of questions around two specific lists: alist of hypothesized impacts, and a list of theindicators by which these impacts will bemeasured.

Be sure the questions are properly wordedso that responses will be as consistent aspossible. Many monitoring questionnairesask ambiguous questions. This makes theanalysis difficult if not impossible. Forexample, questions such as “Did you like theseed” are much less useful than a series ofquestions about germination rates, seedlingvigor, plant health, and relative grain yields.

5. Identification of enumeratorsNGOs commonly use their own staff toconduct impact surveys. This creates two

possible sources of bias.First, farmers may alreadyperceive these staff as havinga vested interest in certainresponses. Most recipientsalready know they need tomeet certain criteria in orderto qualify for assistance. Theymay similarly believe thatwhether they receiveassistance next year willdepend on their answers to

questions about what they did with thisyear’s assistance. In addition, there isalways a risk that stakeholders may have avested interest in the result of theevaluation.

6. Train enumerators to ask thequestions with careEffective training of enumerators commonlytakes many days of discussion andpractice. It is important that enumeratorsfully understand the purpose of the surveyand problems of interpretation likely to beencountered in the field. They must beprepared to ask follow-up questions toresolve ambiguities in initial answers.Possible differences in interpretation ofeach question must be analyzed.Enumerators must be trained to beskeptical of first responses and to watch forinconsistencies in responses. And theymust often be trained to learn with theireyes as well as their ears. This takes timeand considerable practice.

There is no single correctmethod for conducting aprogram assessment.The method – a simplechecklist, a briefquestionnaire, or detailedinterviews – will dependon what precisely is beingassessed.

43

7. Establish a sample capable oftesting each impact hypothesisMost monitoring and evaluation programsselect sample frames of programparticipants with the objective of measuringchanges in practice or result after inputshave been provided. This is difficult becausethe likely position of these households if theyhad not received assistance, is unknown. If ahousehold had not received seed from arelief program, for example, it may havereceived it from a neighboring household.There is no way to test this proposition with asample confined to reliefrecipients. An alternative is todeliberately select a samplethat includes households thatreceived assistance and thosethat did not.

The problem here is thatthese two sub-populationsmay have started from suchdifferent levels of resourceendowment (vulnerable vs better offhouseholds) that dif ferences at the end ofthe season remain difficult to attribute.Ultimately, the key issue is to be clear howthe sample is chosen. If generalizationsare to be drawn about a larger populationthe sample must genuinely reflect thecharacteristics of this population. This is

generally achieved through randomization.A random sample of households isselected in a random set of villages in arandom set of wards in the districts wherethe NGO has been active.

8. Organize data processing andanalysis earlyThe longer the time lag between datacollection and data processing, the morelikely difficulties will arise in interpreting theresults. Problems of coding can be quicklyresolved if data processors are in contact

with enumerators or surveysupervisors. Questions aboutinconsistencies in the datacollected from individuals orgroups of respondents cansimilarly be resolved if analysisfollows soon after datacollection. As time passes,memories of field problems, orunexpected factors likely toinfluence interpretation of the

data, are lost. And interest wanes in thecompletion of a complete and accurateanalysis.

A sample of a generic monitoring plan isprovided in Appendix 10. A sample of aformal questionnaire for post-plantingassessment is provided in Appendix 11.

Data processing andanalysis must followimmediately after datacollection. Delays willlead to difficulties – dueto coding errors, datainconsistencies, andinterpretation of results.

44

Appendix 1. Key Contacts in Zimbabwe

International Agricultural Research Centers

Center Address Telephone Fax Crops

International Maize and PO Box MP 163, Mt 04-301807 04-301327 Maize, wheatWheat Improvement Pleasant, HarareCenter (CIMMYT)International Crops PO Box 776, Bulawayo 083-8311 to 8318 083-8253 or Sorghum, pearl millet,Research Institute for 8307 groundnut, pigeonpea,the Semi-Arid Tropics chickpea(ICRISAT)World Agroforestry P O Box MP128, Mt 04-369122 to 124 04-728340 Agroforestry cropsCenter (ICRAF) Pleasant, HarareCenter for International 74 Harare Drive, Mt 04-369655 04-369657 ForestryForestry Research Pleasant, Harare(CIFOR)

Fertilizer Companies

Company Address Telephone Fax Business

Omnia PO Box BW736, 04-369390 04-369393 Wholesale and retailBorrowdale, Harare

Windmill PO Box WGT560, 04-334911 to 19 04-334910 Wholesale and retailWestgate, Harare

Zimbabwe Fertilizer Co PO Box 385, Harare 04-753882 to 89 04-753881 Wholesale and retailSable Chemicals PO Box 561, Kwekwe 055-23601 to 609 055-23611 ManufacturerZimphos PO Box AY120, Amby, 04-498837 to 47 04-487934 Manufacturer

Harare

Seed Companies

Company Address Telephone Fax Crops

Agricultural Seeds and PO Box 6766, Harare 04-700655 or 728021 04-700655 or Non-maizeServices 728021Monsanto PO Box EH 47, 04-336626 to 28 04-336636 Hybrid maize

Emerald Hill, HarareNational Tested Seeds PO Box 3018, Harare 04-310284 to 87 04-331050 or Numerous crops

310288PANNAR PO Box 99, Ruwa 073-2631 to 34 073-2652 Maize, sorghum, othersPioneer Mutual Gdns 100, 04-339301 to 303 04-339386 Hybrid maize

The Chase West, HararePrime Seeds PO Box BW1798, 04-480501 or 02 04-480501 or 02 All + vegetables

Borrowdale, HarareSeed Co Ltd PO Box WGT 64, 04-308881 to 88 04-304841 Maize, sorghum, pearl

Westgate, Harare millet, sunflower,groundnut, cowpea

Zimbabwe Seed Trade PO Box A1906, 04-332017 04-332017Association Avondale, Harare

45

Key Government Departments

Department Address Telephone Fax Focus

Department of PO Box BE150, 04-778160 04-778161 Rainfall dataMeteorological Services Belvedere, HarareFamine Early Warning PO Box 7810, Harare 04-744434 Climate analysis andNetwork (FEWSNET) predictionSurveyor General’s PO Box CY54, 04-775550 to 52 04-780808 Cartography,Office Causeway, Harare geographical surveysCentral Statistical Office P Bag 7705 Causeway, 04-794571 Data collection and

Harare analysis

46

Appendix 2. Sample Household Registration Form

Emergency Agriculture Recovery Application Form1. Application form no2. Name of interviewer3. Date of interview

Geographical Information4. Region Gweru Masvingo Zvishavane

5. District Gokwe L/Gweru Kwekwe BikitaChivi South Chivi North Masvingo ZakaMberengwa Mwenezi Zvishavane

6. Ward7. Village8. Kraal9. Nearest Business Centre/School

Household Information10. Name of household head11. National Registration Number12. Gender of household head Male Female13. Age of household head14. Marital status of household head Married Divorced

Single Widow15. Level of education of household head Primary Secondary Tertiary16. Is the household head employed Yes No17. Name of respondent (if not the household head)18. Relationship to household head Child Spouse Worker19. Next of kin to household head20. National Registration Number of the next of kin21. Household Size22. How many members of your family fall under each of the following categories?

<15 15-35 36-55 >55MaleFemaleDisabled personsChronically ill persons

23. In what numbers do you own the following livestock and assets?Number

CattleDonkeysGoatsPlowsCultivatorsPrevious harvest (how many bags of maize)

24. Any other assistance received from other development agentsCARE Govt Other NGO Other

FoodInputs

25. Type of householdChild headed Elderly Widowed Male/Female defacto

47

Appendix 3. Agro – Ecological Zones and Crops Grown inthe Communal Sector of Zimbabwe

Natural Region 1 – High rainfall, at least 900 mmNatural Region IIa – Moderate and fairly consistent rainfall, 750-1000 mmNatural Region IIb – Moderate but less consistent rainfall, 750-1000 mmNatural Region III – Lower and less consistent rainfall, 650-800 mmNatural Region IV – Low and inconsistent rainfall, 450-650 mmNatural Region V – Low and highly inconsistent rainfall, less than 650 mm

Percent of area planted to alternative crops in each Natural Region

Natural Region Maize Sorghum Pearl millet Finger millet Groundnut Cotton

1 69 8 3 3 6 32 76 1 - 2 9 93 58 3 2 3 9 204 55 11 10 2 10 85 35 36 14 1 8 3Source: Ministry of Agriculture and Rural Development, AREX, various years

48

Percent of area planted to alternative crops in each province

Province Maize Sorghum Pearl millet Finger millet Groundnut Cotton

Manicaland 62 9 7 2 9 4Mashonaland East 72 3 3 3 12 3Mashonaland Central 54 4 1 1 7 29Mashonaland West 63 3 1 0 6 21Midlands 60 5 6 4 9 12Masvingo 50 14 6 7 16 3Matabeleland North 52 17 20 0 2 4Matabeleland South 36 35 17 1 8 0

Source: Ministry of Agriculture and Rural Development, AREX, various years

49

Appendix 4. Simplified Method for Testing Germination Rates

Sampling Proceduresa) Pre-packed seed: select one bag at random from each lot, taking account of the appearanceof the seed. This should appear representative of the larger sample.

b) Bulk seed or seed in larger grain bags or seed from seed fairs : select samples from seedlots greater than 100 kg only. Take 5 samples from a single seed lot, mix these in a bucket, andselect 1 kg from this lot. These samples should be selected from different parts of the seed lot,taking account of the following:• Heavy and light seeds segregate in storage with heavier seeds found usually at the bottom• During harvest, seed is combined from dif ferent locations resulting in variations of seed

maturity, disease or occurrence of weeds• Seed is not always properly blended• Harvesting, conditioning, and storage conditions can vary within a bulk and thus change its

uniformity. Seed stored on the top of a bin or wagon may be subjected to different conditionsthan the seed stored in the middle or at the bottom of the bulk.

Testing Procedures1

Place two pre-cut circles of newspaper (plain newspaper, not colored) into the bottom of aplastic lid. Moisten the paper with 5 ml of tap water. This may be measured out using a 5 ml

1. Adapted from Taylor JRN. 2001. Methods to identify and specify characteristics desired by industrialsorghum processors. Technical Report no.2. Prepared for USAID. Pretoria, South Africa: University ofPretoria.

Simple germination tests – requiring only limited skills and equipment – can help ensure that relief seed isof adequate quality

50

Germinative Energy (%) 24 hours 48 hours 72 hoursSample X 84 92 95

medicine measure. The newspaper circles must be saturated with water, but there must be nofree water on the surface. Use a little less water if necessary.

Count out 50 to 100 intact seeds and spread evenly over the surface of the moistened paperso that none of the seeds touch each other. Cover the lid with aluminum foil to close it. Placethe lid in a polystyrene box with a close-fitting lid.

Incubation may be carried out at ambient temperature (20-30°C). Maintain high relativehumidity in the box by placing two layers of thin cotton dishcloth saturated with water at thebottom of the box, covering the entire surface area; and placing two layers of thin cotton dishcloth saturated with water covering the plastic lids. The cloths must be re-saturated with watereach day of the test.

Examine the seeds after 24, 48 and 72 hours. At each time interval, count the germinatedseeds and remove them from the plastic lid. A seed is said to have germinated if the root haspenetrated the pericarp. At each time interval calculate the percentage germinated grains.Duplicate determinations should not differ by more than ±5 grains, eg first determination 95%,second determination 90% or 100%.

Germinative Energy is the mean of the duplicate determinations, expressed as a wholenumber. Results should be expressed as Germinative Energy (%) 24 hours, 48 hours, 72hours:

Appendix 5. Seed Varieties Released for Sale in Zimbabwe

Hybrid maize – there are about 75 white maize hybrids on the list from SeedCo, Pioneer,Pannar, Monsanto, the Crop Breeding Institute and the African Centre for FertilizerDevelopment

Open pollinated maize: ZM 421, ZM 521, Matuba, Kalahari Early Pearl (KEP), Obatanpa

Open pollinated sorghum: SV2, SV3, SV4, Macia

Hybrid sorghum: DV 75, PAN 888, NS 5511, Scud

Open pollinated pearl millet: PMV 1, PMV 2, PMV 3

Groundnut: Falcon, Flamingo, Jesa, Teal, Nyanda

Bambaranut: Kazuma, Mana

Cowpea: CBC1, CBC2, CBC3, PAN 238, PAN 311, IT 18

Beans: Iris, PAN 127, PAN 138, Nandi, PAN 148, Bounty

51

Appendix 6. Sample Contract for Procurement of Inputs

AGREEMENT

Made and entered into by and between

XXX and XXXX, Hereinafter referred to as Contractor(name of NGO) (name of supplier)

on 26 Feb 04

in relation to: Procurement of agriculture inputs for Emergency Agriculture Project, 2003/04agriculture season

1. PURPOSE OF AGREEMENT

This agreement between XXX (name of NGO) and the Contractor concerns their relationshipto supply XXX (name of NGO) with locally produced agricultural-inputs necessary for theirprogram, which involves the supply and delivery of inputs to farming families.

2. CONTRACT OBJECTIVE

The contractor will supply 150 tons of Sugarbeans seed by 15 March 04.

3. GENERAL TERMS OF THE AGREEMENT

Either party must first approve, in writing, any and all deviations from the parameters of thisAgreement and Annexes.

All communication relating to this Agreement shall be directed only to the Country Director ofXXX (name of NGO) (or nominated representative) if for XXX, and the Contractor (ornominated representative) for the services required. The parties will notify each other of therelevant contact persons.

Neither party will undertake any work for other agencies using resources from this projectunless contractual obligations have been met. Nothing in this contract will prevent the otherparty from entering into separate Agreements with other agencies for other projects oractivities, provided that such agreements do not detract in any way from the efficacy of thisAgreement.

Neither party shall sub-contract any of the work covered by this Agreement without the priorwritten permission of the other party.

4. AGREEMENT PERIOD

The effective commencement date of this Agreement is the date this agreement is signed byboth parties. This Agreement shall terminate automatically without notice on 31 Mar 04, unlessterminated earlier pursuant to Articles 9 and 10 of this Agreement.

5. RESPONSIBILITIES OF THE CONTRACTOR

THE contractor Agrees to supply and deliver to XXX, 150 tons of Sugarbeans seed by 15March 04.

This amount will be broken down into seed packs containing 2 kg of Sugarbeans seedcomposited into 50kg packs. Based on the above tonnage the Contractor will provide 75,000 x2 kg seed. The entire consignment of seed packs should be delivered to XXX’s (name of NGO)warehouses by 15 March 04 as outlined in paragraph 5.3.

52

5.1 PRODUCT QUALITY AND SPECIFICATIONS

The Contractor will supply XXX with Sugarbeans seed (Red Speckled (Bonus Type)) withminimum germination 80% and minimum purity 98%. Seeds should be accompanied by a seedanalysis certificate and the following documents where applicable: commercial invoice,certificate of origin, phytosanitary certificate, waybills, GMO-free certificate and a fumigationcertificate.

5.2 PACKAGING

The 150 tons of Sugarbeans seed will be packed into strong plastic packs of 2 kg andcomposited into a total of 3000 x 50kg. Product and weight is to be clearly marked on eachbag.

5.3 DELIVERY

Location Masvingo Gweru

# of 50kg Composite packs 2400 600Tonnage 120 30Address Add necessary detailsPerson in charge Add necessary detailsDate 15 March 04 15 March 04

The contractor will supply and deliver composite seed packs according to the followingschedule:

All supplies delivered must be accompanied by the Contractor’s dispatch note. The receivingofficial at XXX (name of NGO) shall sign at least two copies, keeping one copy for XXX andreturning the other to the Contractor. The signed delivery note bearing the signature of the XXXreceiving officer shall constitute proof of delivery by the Contractor to XXX. The XXX receivingofficer will raise a goods received note which should tally with the Contractor’s delivery note. Thebasis of payment will be this duly signed GRN and an accompanying invoice.

5.4 PRICE

The Contractor will supply and deliver seeds to XXX at the following rates and totalPrice per 2 kg packet US$ 1.79Price per ton US$ 895Total cost US$ 134,250Transport: Will be paid upon submission of transport invoices for a maximum of Z$ 113 804 perton to Masvingo and Z$ 107,281 per ton to Gweru.

5.5 PAYMENT TERMS

The agreed total value of the contract of US$ 134,250 shall be paid 15 days after the deliveryof all the 150 tons as per 5.3 to the following bank (details below) submission of a valid invoiceaccompanied by goods received notes, seed certification documents and where applicable,other certificates as per section 5.1.

Bank:Sort code:SWIFT code:Account name:Account number:

5.6 PENALTIES

Any seed packs that are damaged or found not to meet the specifications detailed in sections5.1 and 5.2 whilst under the Contractor’s responsibility will be replaced by the Contractor at nocost to XXX.

Due to the nature of the project and the importance of avoiding any delays in supply of thecommodity to XXX it is essential that the delivery schedule is strictly adhered to, or bettered.Late deliveries, other than due to Force Majeure (section 9), may invoke penalties at a rate of5% of the unit price per ton per seven days.

XXX may satisfy penalties owing to it by way of deductions from payments to be made to theContractor under this Agreement.

6. DISCUSSIONS WITH THE PRESS

Neither party will enter into any discussions relating to this Agreement with members of thepress, on or off the record, without the permission of the other party.

7. INSURANCE

The Contractor shall, at its own expense, establish and maintain appropriate workerscompensation, medical care and disability insurance cover for its personnel in respect of claimsfor illness, bodily injury or death, or damages to or loss of, their property, that may arise fromthe execution of this Agreement.

8. CONFIDENTIAL NATURE OF THE AGREEMENT

The Contractor shall never reveal to any person outside XXX, any information acquired as aresult of its association with XXX, without authorisation. Nor shall the Contractor at any timeuse such information for its own advantage. These obligations survive the expiration ortermination of this Agreement.

Neither the Contractor nor its staff shall disclose to any person or organization, in any manneror form, during the period of this Agreement or after its expiration, any privileged or confidentialinformation of XXX. Likewise, neither XXX nor its staff shall disclose to any person ororganization, in any manner or form during the period of this Agreement or after its expiration,any privileged or confidential information of the Contractor. The specific exception to thisrequirement shall be XXX’s donors to whom it may disclose such financial and operationalinformation of the Contractor as is necessary for the execution of this Agreement.

9. FORCE MAJEURE

“Event of Force Majeure” means, in relation to either party, an event or circumstance beyondthe reasonable control of that party (the Claiming Party) including, without limitation (whether ornot by the Claiming Party), acts of God, acts of War, invasion, revolution, insurrection or otheracts of a similar nature or force, or sabotage or damage by an enemy of either party.

The Claiming Party shall not be deemed to be in breach of this Agreement or otherwise liable to theother party (the Non-Claiming Party) for any delay in performance or any non performance of anyobligations under this Agreement (and the time for performance shall be extended accordingly) ifand to the extent that the delay or non-performance is due to an event of Force Majeure

PROVIDED THAT:-a) The Claiming party could not have avoided the effect of the Event of Force Majeure by

taking precautions which, having regard to all matters known to it before the occurrence ofthe Event of Force Majeure and all relevant factors, it ought reasonably to have taken but didnot take,

54

b) The Claiming Party has used reasonable endeavours to mitigate the effects of the Event ofForce Majeure and to carry out its obligations under this Agreement in any other way that isreasonably practicable and,

c) Failure to adhere to the contract by reason of non availability/shortage of components andother materials required to fulfil the contract shall not be regarded as Force Majeure, but abreach of contract.

The claiming Party shall promptly notify the Non-Claiming Party in writing of the nature andextent of the circumstances giving rise to the Event of Force Majeure.

If the Event of Force Majeure in question prevails for a continuous period in excess of 10 daysafter the date on which it began, the Non-Claiming Party may give notice to the Claiming Partyterminating this Agreement. The notice to terminate must be given in writing and specify thetermination date, which must not be less than 5 clear days after the date on which the notice toterminate is given. Once a notice to terminate has been validly given, this Agreement willterminate on the termination date set out in the notice. Neither party shall have any liability tothe other in respect of termination of this Agreement due to an Event of Force Majeure, butrights and liabilities which have accrued prior to termination, shall subsist including, withoutlimitation, those under Article 18.

10. TERMINATION AND SUSPENSION

This Agreement may be terminated at any time;• For cause, if it is determined by either party that the other party has failed to comply with the

conditions of this Agreement;• For convenience , if all parties agree that continuation of this Agreement would not produce

beneficial results commensurate with further expenditure of funds. All parties shall agreeupon termination conditions, including the effective date and in case of partial termination,the portion to be terminated.

Certain provisions herein, including without limitation, the provisions of this Article and Articles6, 8 and 12 inclusive, shall survive any termination or expiration of this Agreement.

11. DISPUTES

In the event of a dispute an independent arbitrator will be appointed by the agreement of bothparties.

12. NON-LIABILITY

Neither party shall indemnify and hold harmless the other party, and its officers, successorsand assigns, from and against any and all claims, demands, liabilities, expenses (includingreasonable lawyer’s fees and disbursements, court costs, judgements, settlements andfines), whether of omission or commission, that may be committed or suffered inconnection with the performance of this Agreement by either party or a partner or agent ofeither party. This paragraph shall survive termination or expiration of this Agreement.Neither party assumes liability for any third party claims for damages arising out of thisAgreement.

13. LIMITATION

Neither party by making this contract has any obligation to provide other additional supportservices or contracts to the other party for the purposes of this Agreement or otherwise.

14. AMENDMENT

This Agreement may be amended, in writing, subject to written approval by XXX and theContractor.

55

15. NO PARTNERSHIP

Nothing in this Agreement and no action taken by the parties pursuant to this Agreement shallconstitute, or be deemed to constitute, a partnership, association, joint venture or other co-operative entity between the parties.

16. WAIVER

A waiver of any term or condition of this Agreement shall be effected only if given in writing andsigned by the waiving or consenting party. No failure or delay on the part of either party inexercising any right, power or privilege under this Agreement shall operate as a waiver thereof.No breach of any provision of this Agreement shall be waived or discharged, except with theexpress written consent of the parties. The right and remedies herein provided are cumulativewith and not exclusive of any rights or remedies provided by law.

17. ASSIGNMENT

This Agreement is personal to the parties to it. Accordingly, neither party may, without the priorwritten consent of the other, assign the benefit of or the obligations of either party under thisAgreement.

18. INVALIDITY

If any provision of this Agreement is or becomes invalid, illegal or unenforceable in any respectunder the law of any jurisdiction, the validity, legality and enforceability under the law of that orany other jurisdiction of any provision of this Agreement shall not be affected or impaired in anyway thereby.

19. GOVERNING LAW AND JURISDICTION

This Agreement (and any dispute, controversy and proceedings or claim of whatever naturearising out of or in any way relating to this Agreement or its formation) shall be governed byand construed in accordance with Zimbabwe Law. Each of the parties to this Agreementirrevocably agrees that the Magistrates Court at Harare shall have exclusive jurisdiction to hearand decide any suit, action or proceedings in the first instance and/or to settle any disputes,which may arise out of or in connection with this Agreement and, for these purposes, eachparty irrevocably submits to the jurisdiction of the Magistrates Court of Harare. Either party maywithhold fees and compensation due to the other party until a settlement has been reached.

20. ENTIRE AGREEMENT

This Agreement contains the entire understanding of the parties hereto with respect to theEmergency Agricultural Recovery Project. This Agreement supersedes all prior agreementsand understandings between the parties with respect to such subject matter. Any amendmentto the Agreement shall only be valid if evidenced in writing and signed by the duly authorizedrepresentatives of the parties.

21. AUTHORITY

By his or her signature below, each signatory represents and warrants that he or she is dulyauthorized to enter this Agreement on behalf of the party he or she purports to represent suchthat, upon execution and delivery, this Agreement shall be binding obligation of such party.

XXX (name of NGO) Contractor

NamePositionSignature

56

Appendix 7. Warehouse Checklist

Name of implementing partner:Location of warehouse:Contact details:Responsible official:Date of visit:

Checklist 11. Is the warehouse tightly constructed, waterproof and adequately ventilated?Poor Fair Good

2. Stacking of commodities: are they stacked in an orderly, countable manner, away from thewalls and not too close to the ceiling? Are the stacks marked and is there space in betweenthem for easy mobility and access?Poor Fair Good No stocks

3. Is there evidence of infestation, slack or damaged bags, damaged or rusted containers?Yes No

4. Is there more than one type of commodity on one stack and in the warehouse?Yes No

5. What is the condition of the packaging of the commodities?Poor Fair Good No stocks

6. Are pallets used on stacks?Yes No

7. Is there a procedure for removal/ disposal of damaged commodities from the warehouse?Yes No

Checklist 21. Are rodents present and are there any signs of feeding?Yes No

2. Is there refuse or other conditions that serve as an attraction or breeding area for insects?Yes No

3. Is the warehouse provided with fire extinguishers/ buckets for safety precautions?Yes No

4. Is the warehouse road accessible in all weathers?Yes No

5. Evaluate the communication system ( telephone, telex or other)Poor Fair Good

6. Are there any openings that permit entry of insects or birds through windows, doors orventilators?Yes No

57

Checklist 31. What is the state of the roof? Does it leak or retain water?Poor Fair Good

2. Is there a smooth passage of air through the warehouse so that no moisture collects in thewarehouseYes No

3. Monitor the lifting of bags by loaders (to prevent tearing or weakening of the bag)Poor Fair Good No activity

4. Assess the reporting format used in the warehouse to check if it captures all important stockmovement information, ie receipts, dispatches, stock losses and balances.

There is a ledger in use at the warehouse that details all the receipts and dispatches. Thereshould however be a Goods Received Note that XXX (name of NGO) complete and give to thetransporter instead of relying solely on the transporter’s delivery note.

COMMENTS:Generally the warehouse is good and well kept. The main worry is whether it canaccommodate 300 tons of fertilizer at one time.

58

Appendix 8. Datasheet for Monitoring Distribution ofAgricultural Inputs to Ward Level

The followings are examples: the varieties of inputs included are not meant to be exhaustive(especially for seed fairs and voucher programs). The Organization should add columns toinclude additional inputs. Please see earlier sections of this publication for further information

1. Assistance: seed distributionLocation: Province: Matebeleland South District: Insiza Ward No: 17 (Tombo/Papama)Date: 25 May 03Implementing partner: New FrontierFunding source: own sourceStatus: DistributedComments:No. of households: 750No. of seed packs distributed: 750Pack contains 10 kg OPV maize, 2 kg millet, 2 kg sorghum, 10 grams pumpkin, 5 gramstomato, 25 kg ammonium nitrate fertilizer

Quantity distributed

OPV maize Sorghum Pearl millet Pumpkins Tomato Fertilizer, Limed AN(kg) (kg) (kg) (kg) (kg) (kg)

7500 1500 1500 7.5 3.75 18,750

2. Assistance: seed distributionLocation: Province: Matebeleland South District: Insiza Ward No: 4 (Mleja)Date: 28 May 03Implementing partner: Africa DevelopmentFunding source: own sourceStatus: PlannedComments:No. of households: 250No. of seed packs distributed: 500 (vegetable seeds distributed in a second pack)Pack contains 5 kg OPV maize, 2 kg millet, 10 kg sorghum, 10 grams pumpkin, 5 gramstomato, 25 kg ammonium nitrate fertilizer

Quantity distributed

OPV maize Sorghum Pearl millet Pumpkins Tomato Fertilizer, Limed AN(kg) (kg) (kg) (kg) (kg) (kg)

1250 2500 1250 2.5 1.25 6250

Note: It is compulsory to include Seed Pack Composition. Ward number: compulsory. Ward name:strongly requested if available

59

HH* a

ssist

ed, b

y dist

rict

HH as

sisted

, by N

GOVA

C re

port

data

(Apr

il 200

3)Pr

ogra

m ev

aluati

on

No. o

f HH

No. o

f HH

% of

HH

with

HH a

ssist

edHH

ass

isted

Distr

ictas

sisted

NGO

assis

tedTo

tal p

opTo

tal n

o.of

HH

cere

al de

ficit

vs to

tal H

Hvs

HH

in ne

ed

MANI

CALA

ND P

ROVI

NCE

Buhe

ra11

,648

CRS/

FACH

IG11

,373

220,

161

50,03

760

%23

%39

%Bu

hera

IFRC

275

Chim

anim

ani

12,81

7HE

LP/A

DF5,

000

111,

755

25,39

954

%51

%94

%Ch

iman

iman

iGe

rman

Agr

i-Acti

on57

1Ch

iman

iman

iCR

S/FA

CHIG

4,24

6Ch

iman

iman

iFA

O/SC

N3,

000

Chipi

nge

74,30

0FO

SENE

T/ZW

B6,

485

261,

395

59,40

850

%12

5%24

9%Ch

iping

ePL

AN39

,325

Chipi

nge

PLAN

28,49

0Ma

koni

33,66

7FA

O/GO

AL12

,676

244,8

2355

,642

49%

61%

124%

Mako

niCR

S/CT

DT3,

843

Mako

niAF

RICA

RE5,

000

Mako

niOx

fam

Cana

da/Z

PT2,

000

Mako

niFO

SENE

T/ZP

T9,

623

Mako

niHE

LPAG

E52

5Mu

tare

63,17

2HE

LP/P

LAN

19,99

721

7,843

49,51

058

%12

8%22

0%Mu

tare

PLAN

38,19

0Mu

tare

AFRI

CARE

4,98

5Mu

tasa

57,85

0HE

LP/P

LAN

41,80

016

0,036

36,37

254

%15

9%29

6%Mu

tasa

PLAN

16,05

0Mu

tasa

Oxfam

Can

ada/F

CTZ

Nyan

ga9,

972

CAFO

D/Ca

dec

8,73

911

3,478

25,79

053

%39

%73

%Ny

anga

Germ

an A

gri-A

ction

1,23

3

TOTA

LZI

MBAB

WE

980,8

5898

0,858

10,43

2,131

2,38

2,50

747

%41

%88

%*

HH =

hou

seho

lds

App

endi

x 9.

Sam

ple

of D

ata

Rep

orte

d on

Dis

tric

t Lev

el In

put D

istr

ibut

ion

60

Appendix 10. Program Monitoring Plan

Objective Indicators Monitoring Inputs Final Verification

By April 2004,149,000vulnerablefarmhouseholds in10 districts ofMasvingo andMidlandsprovinces,have utilizedan increasedquantity anddiversity ofcereal grainand legumeplantingmaterial,throughprovision of2533 t ofseeds andpossibly7450 t oftop dressingfertilizer

% of input packs arriving at the primarywarehouses intact

Quantities of inputs dispatched matchthe # of beneficiaries at distributionpoint. Proper loading and offloading

% transporters getting inputs to rightdestination, in the right quantities and atagreed times; % distributors who havesigned contracts, arranged offloadingand storage and received beneficiarylists and vouchers

Ascertain income generated and howused by distributor

% of targeted beneficiaries receivinginput packs in time for 2003/04 plantingseason

% of beneficiaries reporting satisfactionwith quantity, quality and variety of seedreceived, and suitability andeffectiveness of distribution method

% of beneficiaries cultivating seeds asintended

Compare “better” vs “poor” farmers inorder to share success stories amongstcommunities and promote the project

Assess production levels and use ofadditional production derived frominputs package

Review logistics records andperiodically check inputs on receipt

Matching quantities dispatched to the #of beneficiaries per distribution point

Visit destinations and hold one-on-onediscussions with distributors, note theirconcerns

Check records of cash receipts andany invoices of business transactions

Monitoring Officers (MOs) randomlysurvey beneficiaries to ascertaincompliance with targeting criteria

MOs monitor voucher distribution tomeasure timeliness, ensurecompliance with distribution lists andprocedures established by the projectwith beneficiaries, AGENT rural tradersand RMFP groups

Field staff and MOs collect informationon input utilization, land preparation,planting dates, agronomic practices,estimated yields, crop condition andconstraints faced

MOs undertake post distributionmonitoring to assess utilization ofseeds and fertilizer, monitor cropestablishment and growth

MO and field staff assess productionlevels, storage, what quantities aremarketed

External consultant undertakes finalproject evaluation

Actual # ofbeneficiaries whoreceived seed andfertilizer on time

On a sample basisconfirm actualgerminationpercentage of seeddistributed

Sample field cropstands to monitor useof seed and fertilizer asintended

Sample beneficiaryfields to assessestimated and actualyields

All distributors toaccount for utilizationof the money injectedinto their businesses

61

App

endi

x 11

. Sam

ple

of P

ost-P

lant

ing

Ass

essm

ent S

urve

y Q

uest

ionn

aire

This

sur

vey

can

prov

ide

info

rmat

ion

on w

heth

er th

e ho

useh

old

rece

ived

relie

f inp

uts,

how

thes

e w

ere

used

, and

the

alte

rnat

ive

sour

ces

of s

uppl

y of

thes

e in

puts

. Thi

s in

form

atio

n he

lps

asse

ss th

e re

lativ

e co

ntrib

utio

ns o

f rel

ief i

nput

s to

cro

p pr

oduc

tion.

Dat

e: _

____

____

____

____

____

_E

num

erat

or: _

____

____

____

____

____

_P

rovi

nce:

___

____

____

____

____

Dis

trict

: ___

____

____

____

____

____

___

War

d: _

____

____

____

____

____

_Vi

llage

: ___

____

____

____

____

____

___

Res

pond

ent s

houl

d be

an

adul

t who

is a

mai

n de

cisi

on m

aker

for t

his

hous

ehol

d. If

hus

band

and

wife

join

tly m

anag

e fa

rmin

g ac

tiviti

es, b

oth

shou

ldbe

inte

rvie

wed

toge

ther

. Par

ticip

atio

n of

the

wife

sho

uld

be e

ncou

rage

d.

Res

pond

ent a

nd h

ouse

hold

det

ails

* Hea

lth st

atus

: goo

d =

little

or n

o sic

knes

s fair

= si

ck a

ppro

ximat

ely 1

5-40

day

s per

year

poo

r = si

ck m

ore

than

40

days

per

year

Name

of re

spon

dent

Statu

sGe

nder

Age

Year

s of e

duca

tion

Resid

ency

Healt

h sta

tus*

Male

HH he

ad=1

Male=

1Fu

ll tim

e=1

Good

=1Fe

male

HH he

ad=2

Fema

le=2

Part

Time=

2Fa

ir=2

Othe

r (sp

ecify

)=3

Poor

=3

Full t

ime

(9 m

onth

s or m

ore)

Part

time

(2-8

mon

ths)

Adult

s (15

+ ye

ars)

Child

ren (

0-14

year

s)

1. H

ow m

any

resi

dent

s liv

e in

this

hou

seho

ld (i

nclu

ding

resp

onde

nts

liste

d ab

ove)

?

62

Inpu

ts r

ecei

ved

2. D

id y

ou re

ceiv

e an

y se

ed fr

om o

utsi

de a

genc

ies

this

cro

ppin

g se

ason

? __

____

____

____

_Ye

s=1,

No=

2[if

NO

, go

to Q

4]

3. W

hat s

eed

did

you

rece

ive,

and

from

who

m?

4. D

id y

ou re

ceiv

e an

y ch

emic

al fe

rtiliz

er fr

om o

utsi

de a

genc

ies

this

cro

ppin

g se

ason

? __

____

____

____

_Ye

s=1,

No=

2 [I

f NO

, go

to Q

6]

5. W

hat t

ypes

of f

ertil

izer

did

you

rece

ive?

Wha

t age

ncy p

rovid

edW

hen w

as it

Whe

re w

as it

How

far is

this

Any p

roble

ms th

e see

dpr

ovide

dpr

ovide

dfro

m he

reCr

opVa

riety

Quan

tityUn

itswi

th thi

s see

d?

(wee

k/mon

th)(m

inute

s walk

)Ma

ize=1

Prov

ide na

mekg

=1No

ne=1

W s

orgh

um=2

orOt

her

Poor

germ

inatio

n=2

R

sorg

hum=

3Un

know

n=9

(spec

ify)=

2Va

riety

unkn

own=

3P

mi

llet=

4Cr

op un

know

n=4

Grou

ndnu

t=5

Mixe

d var

iety=

5Co

wpea

=6Ot

her (

spec

ify)=

6Ot

her (

spec

ify)=

7

Wha

t age

ncy p

rovid

edW

hen w

as it

Whe

re w

as it

How

far is

this

the f

ertili

zer

prov

ided

prov

ided

from

here

Type

of fe

rtiliz

erQu

antity

Units

(wee

k/mon

th)(m

inute

s walk

)kg

=150

kg ba

g=2

Othe

r (sp

ecify

)=3

63

Cro

ps a

nd c

rop

man

agem

ent

6. W

hat h

ave

you

plan

ted

this

sea

son

and

how

hav

e yo

u m

anag

ed th

ese

crop

s?

Kg of

basa

lKg

of to

p-dr

ess

Plan

ting

Kg of

Var

# 1

Kg of

Var #

2Kg

of m

anur

efer

tilize

rfer

tilize

rPl

ot #

Crop

date

Varie

ty #1

Sour

cepla

nted

Varie

ty #2

Sour

cepla

nted

appli

edap

plied

appli

ed

Maize

=1(w

eek/m

onth)

(inclu

deOw

n stoc

k=1

(inclu

deOw

n stoc

k=1

W so

rghu

m=2

loca

lNe

ighbo

r=2

local

Neigh

bor=

2R

sorg

hum=

3na

mes)

Dista

nt re

lative

=3na

mes)

Dista

nt re

lative

=3Pe

arl m

illet=

4Re

tail s

hop=

4Re

tail s

hop=

4Gr

ound

nut=

5Gr

ain m

arke

t=5

Grain

mar

ket=

5Co

wpea

=6NG

O (g

ive na

me)=

6NG

O (g

ive na

me)=

6Ot

her (

spec

ify)=

7Ot

her (

spec

ify)=

7Ot

her (

spec

ify)=

7

Seed

sal

es, g

ifts,

bar

ter,

stoc

ks, c

onsu

mpt

ion

7. D

o yo

u st

ill ha

ve a

ny s

eed

in s

tock

? _

____

____

___

Yes=

1,N

o=2

8. D

id y

ou s

ell,

trade

, bar

ter o

r giv

e aw

ay a

ny s

eed

this

sea

son?

___

____

__Ye

s=1,

No=

29.

Did

you

con

sum

e an

y of

you

r see

d, o

r the

see

d yo

u re

ceiv

ed fr

om N

GO

s, th

is s

easo

n? _

____

____

____

___

Yes=

1,N

o=2

10. [

if th

e an

swer

to Q

7, 8

or 9

is y

es]

Wha

t did

you

do

with

the

seed

you

did

not

pla

nt?

Cro

pSe

ed so

urce

Tran

sacti

onQu

antity

Price

Reas

on

Maiz

e=1

Own s

tock=

1Sa

le=1

state

units

If sa

le/ba

rter –

for e

ntire

tran

sacti

onEx

cess

to ne

ed=1

W so

rghu

m=2

Neigh

bor=

2Gi

ft =2

To h

elp a

nothe

r far

mer=

2 R

sorg

hum=

3Di

stant

relat

ive=3

Barte

r=3

To ea

rn ca

sh=3

Pea

rl mille

t=4

Retai

l sho

p=4

Cons

ume=

4Di

d not

like v

ariet

y=4

Gro

undn

ut=5

Grain

mar

ket=

5St

ill in

stock

=5Un

able

to pla

nt=5

Cow

pea=

6NG

O (g

ive na

me)=

6Ot

her (

spec

ify)=

6Ot

her (

spec

ify)=

6 O

ther (

spec

ify)=

7Ot

her (

spec

ify)=

7

Than

k th

e re

spon

dent

s an

d as

k if

they

hav

e an

y qu

estio

ns. R

ecor

d th

ese.

566–2004

ICRISAT-Patancheru(Headquarters)Patancheru 502 324Andhra Pradesh, IndiaTel +91 40 23296161Fax +91 40 23241239icrisat@cgiar.org

ICRISAT-Niamey(Regional hub WCA)BP 12404Niamey, Niger (Via Paris)Tel +227 722529, 722725Fax +227 734329icrisatsc@cgiar.org

ICRISAT-Nairobi(Regional hub ESA)PO Box 39063, Nairobi, KenyaTel +254 20 524555Fax +254 20 524001icrisat-nairobi@cgiar.org

ICRISAT-BulawayoMatopos Research StationPO Box 776,Bulawayo, ZimbabweTel +263 83 8311-15Fax +263 83 8253/8307icrisatzw@cgiar.org

ICRISAT-BamakoBP 320Bamako, MaliTel +223 2223375Fax +223 2228683icrisat-w-mali@cgiar.org

ICRISAT-Maputoc/o INIA, Av. das FPLM No 2698Caixa Postal 1906Maputo, MozambiqueTel +258-1-461657Fax +258-1-461581icrisatmoz@panintra.com

ICRISAT-LilongweChitedze Agricultural Research StationPO Box 1096Lilongwe, MalawiTel +265-1-707297/071/067/057Fax +265-1-707298icrisat-malawi@cgiar.org

The International Crops Research Institute for the Semi-Arid Tropics (ICRISAT) is a non-profit, non-political, international organization for science-based agricultural development. ICRISAT conductsresearch on sorghum, pearl millet, chickpea, pigeonpea and groundnut – crops that support the livelihoodsof the poorest of the poor in the semi-arid tropics encompassing 48 countries. ICRISAT also sharesinformation and knowledge through capacity building, publications and ICTs. Established in 1972, it is oneof 15 Centers supported by the Consultative Group on International Agricultural Research (CGIAR).

Contact information:

About ICRISAT

Visit us at www.icrisat.org