Africa’s young workers are arriving faster than formal jobs, and most of them will start on or near a farm. The World Bank’s Development Research Group expects 1.2 billion young people to reach working age in low and middle income countries over the next ten years, and a new IFPRI paper, citing the World Bank, puts the addition in sub Saharan Africa at roughly 20 to 25 million workers a year. The Bank’s March 2026 research newsletter sorts what works into three groups: foundations such as roads, power and skills; policies that let firms hire; and finance that reaches them. This article reads those findings through one question for African cooperatives, agribusinesses and governments: if farming will keep absorbing young people, what has to be built around the farm so that more of the jobs are good ones?
Read this first. We read the World Bank newsletter and IFPRI Discussion Paper 2439 of August 2026, including its country tables. We did not read the World Bank papers the newsletter summarises, the ILO Africa Youth Employment Outlook 2026, or the AGRA keynote in full; for those we relied on extracts and press reports, and we say so where it matters. The IFPRI paper is a discussion paper that has not been through formal review, and Kenya is not among its 15 countries. Some figures come from outside Africa and are marked. Research stops at 14 November 2026.
Seven findings
- Farming is still the main job. In 15 sub Saharan countries, agrifood systems account for 68.2% of employment, and agriculture itself for 52.5%.
- The exit from farming is slow. The agrifood share of employment falls by only 0.3 to 0.4 percentage points a year, and young people are not leaving agriculture faster than adults.
- The middle of the chain is thin. Workers who do leave farming mostly land outside agrifood altogether, so processing, trade and logistics absorb few of them.
- Investments work as bundles. Roads with extension in Ethiopia raised productivity by 11% where each alone did nothing; electricity with highways in Brazil lifted local output by at least 29%.
- Training works when buyers ask for it. Job programmes rarely lift overall employment unless paired with demand for labour.
- Women are more tied to agrifood work and face more constraints. Childcare laws raise women’s participation, but take up decides whether offers become jobs.
- Finance lags the sector’s weight. Agriculture receives about 3% to 12% of bank lending in the countries we found, against a 20% to 40% share of GDP.
A workforce that arrives faster than formal jobs
The ILO’s Africa Youth Employment Outlook 2026, as reported by Sunday World, counts more than 10 million young people entering Africa’s labour market every year, against only about 3 million formal jobs created by current growth patterns. Agriculture still accounts for 47% of youth jobs in 2025, about 143 million, yet only 2% of agricultural jobs are classified as formal, against about 20% in services. The outlook expects services to overtake agriculture as the largest youth employer by 2033. FAO adds, in its Status of Youth in Agrifood Systems report, that the global share of working youth in agrifood systems fell from 54% in 2005 to 44% in 2021, and that the youth population of sub Saharan Africa is projected to grow 65% by 2050.
Million young people a year. ILO Africa Youth Employment Outlook 2026 as reported by Sunday World. The ILO figures are “more than 10 million” and “about 3 million”, so the bars are approximate; the right hand bar is our arithmetic
Subtracting the two ILO figures leaves about 7 million young people a year without a formal job. That is why the quality of work on and around the farm matters as much as the number of jobs: for most entrants, an agrifood job will be the first job, and whether it is informal, seasonal and unpaid or a step toward something better depends on what surrounds it.
Where the workers are
The most complete recent count comes from IFPRI Discussion Paper 2439, which uses ILO labour data for 15 sub Saharan countries. Agrifood systems, meaning farming plus the processing, trade, transport and food service jobs that depend on it, employ 68.2% of workers on average. Farming alone employs 52.5%, so about three in four agrifood jobs are still on the farm. The paper’s estimate is a lower bound because trade and transport jobs are only partly captured.
Share of total employment, per cent, average across 15 sub Saharan countries. IFPRI Discussion Paper 2439, August 2026. The 31.7% remainder is our arithmetic; slices follow the paper’s rounded figures
The paper sorts the countries by how far their economies have changed. Agriculture dependent economies (DR Congo, Ethiopia, Mozambique, Tanzania, Uganda, Zambia and Zimbabwe) have two nonfarm agrifood workers for every ten farm workers. Diversified economies (Botswana, Namibia and Senegal) have about seven. Transitioning economies (Angola, Côte d’Ivoire, Ghana, Nigeria and Rwanda) sit between.
Share of total employment, per cent, by stage of transformation. IFPRI Discussion Paper 2439, group averages as reported in the text
Two findings cut against common belief. First, the exit from farming is slow: across the 15 countries the agrifood share of employment fell from an average of 70.6% to 66.2% over roughly two and a half decades. Second, young people are not fleeing the farm. Across the sample, 57% of workers aged 15 to 24 were in agriculture against 51% of adults, and the youth share fell by 0.2 points a year against 0.7 for adults. The paper reads this as older workers leaving, not youth, and as a sign that other jobs have not appeared fast enough.
Each dot is one country. Farming share is the average of the years in the paper’s country tables; youth (15 to 24) labour force participation is the 2024 modelled ILO estimate. Periods differ, so read this as an association, not a cause
Our own arithmetic on the paper’s tables shows that each extra percentage point of farm employment goes with about 0.5 points higher youth participation (r = 0.64, 15 countries). The paper explains why: in countries such as Tanzania, Mozambique, Ethiopia and Uganda, many young people enter work through family farming and informal rural jobs instead of staying longer in school. Participation is a measure of who is working or looking for work, not of whether the work is productive.
The thin middle of the chain
If young people are staying in agriculture and the rest are leaving, where do the leavers go? Mostly not into agrifood. Across the 11 countries where farm employment fell, released labour split roughly 55% to 45% between other sectors and nonfarm agrifood. Excluding Côte d’Ivoire, Nigeria and Namibia, where the pattern reverses, other sectors gained 0.71 points a year against 0.21 for nonfarm agrifood. The authors call this a hidden middle that is too thin to absorb workers at scale, and they list what would thicken it: processing, cold chains, logistics, wholesale systems and food safety infrastructure, plus small firm finance, reliable power and help with standards.
Nonfarm agrifood share of total employment, per cent, average across the years each country is observed. IFPRI Discussion Paper 2439, country tables. Periods differ by country; Kenya is not in the sample
Senegal, Nigeria, Côte d’Ivoire and Ghana have the largest nonfarm shares, between 22% and 23%. Mozambique, Botswana and Ethiopia have the smallest, near 11%. The trend inside a country can be steep or uneven. In Nigeria, nonfarm agrifood grew from 12.1% of employment in 2003 to 32.6% in 2022 while farming fell from 57.9% to 34.7%, and an IFPRI synopsis of the Nigerian data says women’s share in nonfarm agrifood roughly tripled and youth participation roughly quadrupled. In Ghana the share rose from 16.1% in 2000 to 27.1% in 2015, then slipped to 23.1% in 2017, a reminder that gains are not automatic.
Nonfarm agrifood share of total employment, per cent. IFPRI Discussion Paper 2439, country tables. For Nigeria the early years come from household surveys and the later years from ILO data, so the join is not exact
The reason to care is partly jobs and partly losses, because they sit in the same place. CSIR estimates for South Africa, as reported by SA Harvest, put 82% of food loss before the consumption stage, with processing and packaging alone at 49%. These are South African figures and cannot be applied to other countries, but they show how much value can be lost after the farm gate. A keynote at AGRA’s 20th anniversary on 31 August 2026, published by allAfrica, makes the same point from the trade side: Africa exports raw cocoa, coffee, cashew and cotton and imports chocolate, instant coffee, tomato paste and milled rice, and it argues that output cannot be raised by expanding acreage. It urges processing next to production, and measuring the share processed domestically, post harvest loss by commodity and the farmer’s share of the retail price.
Share of food loss by stage, per cent. CSIR estimates for South Africa as reported by SA Harvest. The 6% not itemised before consumption is our arithmetic (82 − 49 − 19 − 8)
Policy papers on the hidden middle point the same way. A policy paper from a research project implemented by Rimisp and partners reports that storage, cold chains, transport and processing can raise smallholder incomes by reducing losses and preserving quality, and cites Ethiopia’s agro industrial parks as an example of shared processing hubs that create jobs and link smallholders to export markets. It also notes that its evidence is concentrated in sub Saharan Africa and that effects of multi part interventions are hard to isolate.
Infrastructure works as a bundle
The newsletter’s clearest infrastructure result comes from Brazil. Selod and co authors followed municipalities for four decades and found that neither highways nor electricity alone consistently raised local output. Small and medium municipalities that received electricity after already having highway access saw output per capita rise by at least 29%, and industry expanded. Brazil is not Africa, but an Ethiopian study finds the same pattern on the farm. Mesay Gebresilasse, in a working paper, found that rural roads and agricultural extension were ineffective alone, while villages with both saw productivity rise 11%, by raising the take up of advice, credit and fertiliser.
Brazil: local output per capita in small and medium municipalities (Selod and others, World Bank, 2024, via the newsletter). Ethiopia: agricultural productivity (Gebresilasse, 2018, author’s abstract). The outcomes, places and periods differ, so compare the pattern, not the numbers
Electricity shows the same logic in Kenya and Rwanda. A J-PAL policy analysis reports that subsidised grid connections in Kenya raised household electrification but did not increase electricity use or produce detectable gains in employment, education or health, and that similar results appeared a decade on in Rwanda. J PAL’s answer is to plan around productive use, such as agro processing, irrigation and cold storage, and to use anchor customers like food processors to give utilities predictable demand. It sits alongside Mission 300, the World Bank and African Development Bank target of connecting 300 million people in sub Saharan Africa to electricity. Our reading for a cooperative is that a cold room needs a road, a reliable supply and a buyer at the same time, and that a pump with no credit to buy inputs gives less than it should.
Skills and job search: aligned to buyers, or not at all
Government job training and search programmes have mostly failed to lift overall employment, according to Eliana Carranza and David McKenzie in the Journal of Economic Perspectives. The programmes that show promise link training to high demand skills, use results based funding, help jobseekers look in new places, correct their beliefs about the market and certify skills credibly, and are paired with efforts to raise firm productivity and labour demand. A Latin American study of short cycle higher education found that curriculum design, employer links and teaching practices explain up to 40% of the variation in student outcomes within countries, and that teaching numerical competencies was associated with higher formal employment and pay. That is outside Africa, but it speaks to agricultural colleges and technical training in the cold chain, processing and equipment trades. David McKenzie also reports in a blog that he has updated his earlier pessimism on wage subsidies “a little” after newer evidence, which suggests the case is not closed. Beegle and Fox raise the sharper question of whether such programmes create jobs or only shift them, and a forthcoming study in Côte d’Ivoire randomises both youths and firms to find out; we did not find its results in the sources we read.
IFPRI’s paper reaches a related conclusion for the young: demand linked technical training, apprenticeships and first job support suit a group that is the main margin of adjustment when other sectors expand. A study by Clement Joubert and Kathleen Beegle across seven sub Saharan countries looks at business associations as a way to reach informal firms that policy rarely touches. For agrifood, cooperatives play that role, and that is our inference, not a finding of the study.
Women’s work in the same chain
FAO’s Status of Women in Agrifood Systems finds that agrifood systems provide 66% of women’s employment in sub Saharan Africa against 60% of men’s. Globally, farms managed by women are 24% less productive per hectare than men’s farms of the same size, women earn 18.4% less in agricultural wage work, and men have greater ownership or secure tenure rights over farmland in 40 of 46 countries reporting.
Share of each sex’s employment that is in agrifood systems, sub Saharan Africa. FAO, The Status of Women in Agrifood Systems, 2023
IFPRI’s 15 country data add a pattern inside the chain. Women are more concentrated than men in nonfarm agrifood work, 19.2% of employed women against 13% of men on average, and in 12 of the 15 countries women’s share is higher (our arithmetic on the country tables). Mozambique, Rwanda and Uganda are the exceptions. At the same time, women’s overall lead in agrifood employment shrinks as economies diversify, from 15% above men’s in agriculture dependent economies to 12% and then 9%, and the paper warns that gender segmentation deepens, with men concentrated in commercial farming and women in downstream food service roles.
Share of employment in nonfarm agrifood work by sex, per cent, average across observed years; bubble area is the country’s total nonfarm agrifood share. IFPRI Discussion Paper 2439, country tables. Counting countries above the parity line is our arithmetic
Share of each sex’s employment in agrifood systems, per cent (left axis), and the percentage by which women’s share exceeds men’s (right axis). IFPRI Discussion Paper 2439, group averages as reported; the paper notes the diversified group is dominated by Botswana
What helps? A global study of childcare laws in 190 countries, in the World Bank newsletter, finds that enacting them raises female labour force participation by about 1.4 percentage points, a 2.2% gain, with affordability mattering most, ahead of availability and quality. FAO also finds that access to childcare has a large positive effect on mothers’ employment. Two African cases show the difficulty of turning an offer into a job. In an Egyptian experiment reported by Kathleen Beegle, neither childcare vouchers nor job matching, nor both, changed mothers’ employment, because take up was low: only 11% of households used the free childcare and about one in five women in the job services group applied for a job. In the Republic of Congo, only about a third of children aged 3 to 5 have access to early childhood education, and private daycare can cost over $100 a month against a minimum wage near $150. A Ghanaian study of grants to women owned businesses asks why such grants often fail, and looks at pressure to share resources within households. For farm cooperatives, the practical questions are whether care services fit planting and harvest calendars, whether land titles are secure for women, and whether support reaches women directly.
Who finances the middle
The African Development Bank’s president said in March 2025 that agriculture contributes 30% of Africa’s GDP but receives only 6% of commercial bank lending, as reported by allAfrica. Making Finance Work for Africa lists country shares of 3% in Sierra Leone, 4% in Ghana and Kenya, 6% in Uganda, 8% in Mozambique and 12% in Tanzania, against a 20% to 40% share of GDP across the continent. At a Nairobi forum reported by allAfrica, a Kenyan banker put the Kenyan figure at about 4% against agriculture’s 22% of GDP. The forum described a “missing middle” of loans too large for microfinance and too risky for banks, with collateral demands of up to 120% of the loan.
Share of commercial bank lending that goes to agriculture, per cent. Six countries: Making Finance Work for Africa (undated). Africa overall: African Development Bank president, March 2025. Dashed lines mark agriculture’s 20% to 40% share of GDP across the continent
The gap has a price. The same AfDB event heard that Kenyan government bonds yield about 14% while agricultural SME lending returns an average of about 3%, a spread that, in our reading, discourages lending to the middle of the chain. Estimates of the gap vary. AGRA is quoted at $65 billion, and industry estimates reported by The Star reach $100 billion a year, which is a different basis.
US$ billion. AGRA as quoted by allAfrica, December 2025; industry estimates reported by The Star, April 2026. The two are not on the same basis and should not be read as a trend
Two remedies appear in the evidence. One is risk sharing: Making Finance Work for Africa says AGRA’s guarantee facilities with Equity Bank and Standard Bank leveraged about ten times their public risk capital into private lending to farmers. The other is wider capital markets. A joint IFC and World Bank study of firms in 106 low and middle income countries finds that bond and equity markets let smaller, younger and more productive firms invest and hire, generating about 5% more jobs over three decades; it is not about agrifood in particular. A Bank paper by Cull and co authors cautions that multilateral efforts to mobilise private capital still fall well short of need, held back by thin pipelines of investable projects, limited influence and difficulty proving additionality. A pipeline of verified, bankable agrifood projects is exactly what is missing in the middle of the chain.
Where verified field data fits
Banks, buyers and funders need records they can trust before they lend to a cooperative, a mill or a cold room. According to the AgriKonnekt page, AgriKonnekt is built to check farm plots with sensors and satellites so cooperatives can show results to funders, and a project summary at the Paris Peace Forum describes work with 1,500 farmers across six counties in Kenya. Those are the organisation’s statements. We have not seen it used to verify jobs, post harvest losses or processed volumes, so applying it to the middle of the chain is our proposal, and it would need agreed definitions, baselines and independent checks.
What to do next
- Governments should plan roads, power, cold storage and extension as district packages and report on farms and firms served, not on connections or kilometres alone.
- Ministries should treat processing, packaging and storage as both job and loss targets, and track the share of output processed locally, post harvest loss by crop and the farmer’s share of the retail price.
- Training funders should tie support to buyers’ demand, use results based payment and test whether jobs are created or only shifted.
- Programmes for women should check take up before scaling, price childcare to be affordable and fit farm calendars, and secure women’s land rights and direct access to grants.
- Banks and funders should use guarantees and risk sharing, lend against transaction records where collateral is lacking, and publish agriculture’s share of lending next to its share of GDP.
- Cooperatives should keep member records of output, losses and jobs, because verified data is what turns a farm group into a bankable partner and a channel for services.
- Statistics offices and researchers should add Kenya and other missing countries to harmonised, age and sex disaggregated agrifood employment data, as IFPRI’s authors urge.
What we can and cannot claim
| Statement | Evidence status |
|---|---|
| Agrifood systems employ 68.2% of workers in 15 countries | IFPRI discussion paper, read in full text; not formally reviewed; Kenya not included; a lower bound |
| 57% of youth and 51% of adults work in agriculture | IFPRI; ages 15 to 24; cross country average, not a trend |
| Other sectors absorb 0.71 points a year against 0.21 for nonfarm agrifood | IFPRI; excludes three countries; descriptive, not causal |
| More than 10 million entrants, about 3 million formal jobs, 47% of youth jobs in agriculture | ILO outlook as reported by Sunday World; outlook not read |
| Electricity with highways lifted output by at least 29% in Brazil | World Bank newsletter summary; paper not read; Brazil, not Africa |
| Roads with extension raised productivity by 11% in Ethiopia | Author’s abstract of a 2018 working paper; older; paper not read |
| Childcare laws raise female participation by 1.4 points | Newsletter summary of a 190 country study; not specific to agriculture or Africa |
| Agriculture gets 3% to 12% of bank lending | Making Finance Work for Africa (undated) and AfDB 2025; definitions and years differ |
| A financing gap of $65 billion or $100 billion a year | Press reports on different bases; not comparable |
| 49% of South African food loss is in processing and packaging | CSIR figures as reported by SA Harvest; South Africa only; CSIR not read |
| AgriKonnekt checks farm plots with sensors and satellites | Organisation statements; not independently verified |
Africa’s young workforce will not wait for a perfect plan, and the evidence suggests that most of it will begin in agrifood work whether or not policy is ready. The question is what that work looks like. Where roads, power, advice and buyers arrive together, farms become more productive and the processing, trading and food service jobs around them have something to grow from; where they arrive one at a time, the gains tend to disappear. Where finance follows the sector’s true weight in the economy, the missing middle can be built, and where women’s care burdens, land rights and access to credit are treated as part of the design, more of the new jobs will reach them. None of this requires a new idea, only the discipline to bundle investments, to measure what happens on actual farms and in actual processing plants, and to share what is learned. The countries and cooperatives that do so will turn an arriving generation into the foundation of a stronger agrifood economy rather than a queue for work that does not exist.
Sources & notes
SOURCES: World Bank Development Research Group, What Works for Jobs?: Foundations, Policies, and Finance for a Dynamic Private Sector, Research Newsletter (March 2026); Bachewe, Abay, Chamberlin and Korir, Who Is Leaving African Agriculture? Nine Stylized Facts on Youth, Women, and Agrifood Employment in 15 Countries, IFPRI Discussion Paper 2439 (August 2026); IFPRI, Synopsis: The changing demographics in Nigeria’s food systems; Sunday World, Africa’s youth workforce shifts from agriculture to services; FAO, The Status of Youth in Agrifood Systems; FAO, The Status of Women in Agrifood Systems (2023); FAO newsroom on the Status of Women report; allAfrica, Africa: A Harvest That Employs the Century (AGRA keynote, 31 August 2026); SA Harvest, Food loss in South Africa starts before the cold room; Rimisp and partners, Unlocking the Hidden Middle: Reclaiming the Role of Intermediaries in Agrifood Systems; Gebresilasse, Rural Roads, Agricultural Extension, and Productivity; J-PAL, Cracking Africa’s chicken and egg problem with energy and economic growth; Making Finance Work for Africa, Agricultural Finance; allAfrica, Global and African Financial Experts Urge Action to Enhance Smallholder Farmer Financing (March 2025); allAfrica, Kenya: Financing Gap Risks Undermining Kenya’s Agriculture Growth (December 2025); The Star, Africa faces US$100 billion annual agriculture financing gap (April 2026); Paris Peace Forum, Democratising Climate Finance for Smallholder Farmers in Sub Saharan Africa; AgriKonnekt.
Differences between sources: the share of youth in agriculture is 47% of youth jobs across Africa in 2025 (ILO) and 57% of workers aged 15 to 24 across 15 countries (IFPRI), because coverage and years differ; agrifood supplies 66% of women’s and 60% of men’s jobs in sub Saharan Africa (FAO, 2023) and 72.6% and 64% in IFPRI’s 15 countries; agriculture’s share of GDP is 30% (AfDB president) and 20% to 40% (Making Finance Work for Africa); the agricultural financing gap is $65 billion (AGRA) and $100 billion a year (industry estimates).
Our own arithmetic: the 31.7% of employment outside agrifood; about 7 million young people a year without a formal job (more than 10 million minus about 3 million); the 6% of South African food loss not itemised before consumption (82 − 49 − 19 − 8); the trendline slope and correlation between farming share and youth participation; the count of 12 of 15 countries where women are more concentrated in nonfarm agrifood work.