Close
Impact

Graduation packages lift farm incomes when inputs and markets exist

(9 min read)

A grant on its own rarely moves a farm household out of poverty. Evidence from 2015 to May 2026 points to what does: a package of an asset, coaching and savings, delivered in a way a government can afford, into a market where inputs, roads and buyers exist. This article follows that evidence from the household to the processor and asks a practical question for African cooperatives, agribusinesses and governments: how does the next dollar of support, and the next dollar of credit, become farm income that lasts?

Read this first. We read in full the May 2026 PEI newsletter on economic inclusion, the World Bank newsletter of 29 May 2026, the WBG Civil Society Update and the Agritech Digest issue. We did not read the PEI meta analysis, which is ongoing, the country evaluations behind the newsletter, or the Science paper itself. Ethiopia figures come from J PAL’s summary of that paper, and the declarations from African Union documents, press reports and policy briefs. Newsletter findings are PEI’s and the World Bank’s own statements. Searches ran on 16 November 2026.

Five findings

  • A package beats a single input. The meta analysis averages gains of US$33.7 a month in consumption and US$36.2 in income.
  • Gains last. A year after the Ethiopia program ended, consumption was 18.2% higher and assets 68% higher.
  • Delivery design is a lever. In Tanzania, livelihood support was nearly six times more cost effective than public works.
  • Credit may scale, but is unproven. Ethiopia’s facilitated credit performed strongly, with open questions on repayment and moral hazard.
  • Inputs and trust decide whether it sticks. Africa’s fertilizer use is about 18 kg a hectare against a target of 50.

What a package adds

PEI reports that economic inclusion programs raise household consumption by US$33.7 a month and income by US$36.2, in 2020 purchasing power terms. For the very poor households they target, that is over 11% and 13%. Households also start more micro businesses, save more and build assets.

What a package adds to a very poor household

Average gain per month from economic inclusion programs. Ongoing PEI meta analysis, as reported in the PEI newsletter, May 2026

0102030400%4%8%12%16%Monthly consumption: US$33.7US$33.7Monthly consumptionMonthly income: US$36.2US$36.2Monthly incomeover 11%over 13%US$ per month, 2020 PPP (left axis)Per cent gain (right axis)

The clearest African test is older. In Tigray, Ethiopia, 925 households took part in one of six trials reported in Science in 2015. The program ran from 2010 to 2012 with the Relief Society of Tigray and the microfinance institution DECSI. It combined five parts: a one time productive asset worth US$1,228, skills training, food support, a compulsory savings account and weekly coaching visits. Most households chose small livestock.

What households chose to rear

Share of participants by productive asset chosen, Ethiopia (Tigray) graduation trial. J-PAL summary of Banerjee et al., Science, 2015; “other” is our arithmetic

Sheep and goats: 62%Sheep and goats62%Oxen: 24%Oxen24%Bees: 10%Bees10%Other: 4%Other4%US$1,228asset per household

One year after the program ended, treatment households consumed 18.2% more, held 68% more assets and had saved 372% more. Livestock revenue doubled, and households spent 43 more minutes a day on productive work.

What changed a year after the program

Gain over comparison households, per cent, Ethiopia graduation trial, one year after the program ended. J-PAL summary

0%100%200%300%400%Monthly consumption: +18.2%+18.2%Monthly consumptionTotal assets: +68%+68%Total assetsTotal savings: +372%+372%Total savings

J-PAL puts total costs at US$4,157 per household and estimated benefits at US$10,805, which it calls a 260% return. In five of the six countries, long run benefits outweighed up front costs.

Benefits against cost, per household

2014 PPP US$ per household, Ethiopia graduation trial. J-PAL summary; the net gain is our arithmetic

04,0008,00012,000Benefits: US$10,805US$10,805Benefits per householdCost: US$4,157–US$4,157Program costNet: US$6,648US$6,648Net gain

How support arrives changes what it buys

The newest results are about delivery. In Cameroon, a government program for urban youth found that paying grants in one tranche or two gave similar results, so one tranche may be cheaper to run. In Malawi, a graduation package raised food security, consumption, assets and business revenues, while group based livelihood support of comparable monetary value had limited impact. In Tanzania, adding public works to cash transfers had limited impact, while enhanced livelihoods proved nearly six times more cost effective.

Ethiopia’s Productive Safety Net Program is testing credit against grants. Early findings say facilitated credit can perform strongly, which could open a route through financial institutions, but selection, repayment, insurance, moral hazard and spillovers are still being studied. The same logic is visible at the other end of the value chain. Sabou Capital has extended a US$2 million impact linked debt facility to Tomato Jos, a Nigerian processor aiming to replace imported tomato paste with local production, as reported by Agritech Digest. Local processing is also a Kampala target, as the next section shows.

The market around the grant

A grant buys farm output only if inputs are within reach. The 2006 Abuja Declaration set a target of 50 kg of fertilizer nutrients per hectare. African Union summit documents say use rose from 8 kg to about 18 kg by 2022, and that crop productivity is about 30% of global averages. African heads of state endorsed the Nairobi Declaration in May 2024, with a ten year action plan and a recommended financing mechanism.

Fertilizer use against the Abuja target

Kilograms of nutrients per hectare, Africa average. African Union summit documents and press reports; the gap is our arithmetic

0204060Abuja target: 50 kg/ha8 kg/haabout 18 kg/ha20062022Gap in 2022: about 32 kg/ha
From almost none to more than ten times the target

Fertilizer use per hectare, selected African countries and benchmarks. African Union and press reports; years vary

0.010.11101001,000SudanSudan: 0.03 kg/ha0.03SomaliaSomalia: 1.04 kg/ha1.04Africa average, 2022Africa average, 2022: about 18 kg/haabout 18Abuja targetAbuja target: 50 kg/ha50EgyptEgypt: 542.57 kg/ha542.57Kilograms of nutrients per hectare of arable land, log scale

Prices pull the other way. The World Bank’s 29 May 2026 newsletter reports energy, food, fertilizer and metals prices climbing, and seven multilateral development banks have pledged support over Middle East conflict disruptions to energy and fertilizer markets. A fixed grant buys fewer bags when prices rise, which is one more reason to test designs, as the LEADS approach (Learn, Adapt, Scale) proposes.

Roads and buyers matter as much. In Madagascar, a US$1.8 million rehabilitation of the 140 metre Manambery Bridge keeps harvests of about 70,000 vanilla farmers moving to market before beans spoil. The Kampala CAADP Declaration, adopted in January 2025 for 2026 to 2035, sets 22 targets, including 45% more agrifood output, 50% less post harvest loss, triple intra African agrifood trade and 35% of agrifood GDP from locally processed food.

Trust decides adoption

Farmers adopt what they trust. In the Agritech Digest podcast, the farmer and advocate Aziz says trust is built through simple communication, field demonstrations, local communities and farmer to farmer recommendations, and that founders go wrong when they design without farmers. A PEI partner blog by Wendy Chamberlin reaches a similar view for Kenya’s pastoralist communities, citing limited phone access, weak app uptake and the lasting role of trusted local networks. Graduation programs build this in through peer coaching and couples training, as in Malawi’s Umodzi model. Scale is arriving: PEI’s Impact Collaborative supports six World Bank operations covering nearly 1.2 million households, and a Zambian women’s program reached over 145,000 women.

Who the programs reach

Bubble area in proportion to the number. Units differ, so this is not a like for like comparison. PEI and World Bank newsletters, May 2026

Impact Collaborative: 1.2 million households, six operations1.2 millionImpact Collaborativehouseholds, six operationsZambia: 145,000 women reached145,000Zambiawomen reachedMadagascar: 70,000 vanilla farmers70,000Madagascarvanilla farmersDominican Republic: 50,000 coffee families50,000Dominican Republiccoffee families

Kenya shows the stakes beyond the farm. The WBG Civil Society Update says poverty and joblessness are entrenched among host communities and refugees as humanitarian aid declines, and that the Shirika Plan’s success hinges on mobility, private investment and shock responsive support.

Six actions

ActionEvidence
Bundle an asset with coaching and savingsEthiopia trial: 18.2% higher consumption, benefits 2.6 times costs
Test tranches and credit before fixing a designCameroon tranches; Ethiopia credit against grants
Put inputs beside the grantFertilizer at about 18 kg a hectare against 50; rising prices
Finance processors as well as farmersTomato Jos facility; Kampala target of 35% local processing
Build adoption through farmer to farmer demonstrationAziz podcast; Chamberlin on Kenya
Measure results on actual plots and householdsLEADS approach; open questions on credit

Poverty exit programs, input markets and processing credit are usually discussed separately, yet on the farm they are one chain. The Ethiopia trial shows that a well designed package pays for itself, the newer country studies show that design details can halve or double cost effectiveness, and the fertilizer figures show how much room remains on the land itself. Cooperatives and agribusinesses that keep records, test delivery choices on their own members and connect grants, inputs and buyers will be best placed to turn the next round of public and private money into incomes that last.

Sources & notes

SOURCES: Partnership for Economic Inclusion, Special Edition newsletter (May 2026); World Bank Group newsletter (29 May 2026); WBG Civil Society Update (May 2026); Agritech Digest (this week’s edition); J-PAL, Graduating the Ultra Poor in Ethiopia; Banerjee et al., Science, 2015; African Union, Africa Fertilizer and Soil Health Action Plan 2024 to 2034; African Union, Unlocking the potential of Africa’s soils; African Union, fertilizer consumption by country; Down To Earth, African ministers call for affordable fertiliser; The Star, African nations vow to triple fertiliser production; Devex, African leaders pledge to triple fertilizer use; CGTN, Summit closes with a 10 year action plan; African Union, ten year strategy and action plan; AKADEMIYA2063, Kampala CAADP Declaration at a glance; IITA, Kampala CAADP Declaration; AgriKonnekt.

Differences between sources: the global fertilizer average is about 120 kg a hectare in Down To Earth and 135 in CGTN; Africa’s average is 18 kg in 2022 in most reports and 18 to 25 kg in others; Egypt is about 400 kg in Down To Earth and 542.57 in the African Union data, which the chart uses; J-PAL calls benefits of US$10,805 on costs of US$4,157 a 260% return.

Our own arithmetic: the 4% “other” asset share; the net gain of US$6,648; the 2022 gap of about 32 kg a hectare; the benefit to cost ratio of 2.6, which makes the net gain about 160% of cost.