Women do more of Africa’s farm work than almost anyone credits, and their plots yield less. The FAO’s 2023 report on women in agrifood systems finds that 66% of women’s employment in Sub Saharan Africa is in agrifood systems, against 60% of men’s, and that female managed farms have 24% lower land productivity than male managed farms of the same size. A World Bank working paper published on 24 June 2026 puts the gender gap in agricultural productivity at between 7% and 77% across twelve African countries. This article asks what drives that gap and which fixes the evidence supports, including a June 2026 Mozambique experiment that found titling women’s land was not enough. It draws on African field studies, FAO and World Bank evidence, and the African Union’s Kampala CAADP Declaration.
Read this first. We read the pasted newsletters and the abstracts, press summaries or extracts of every paper and report cited, not the full documents. Most 2026 evidence here is from World Bank working papers, which have not been through journal peer review. Gap estimates differ by method and country, so we show ranges and do not combine them. The FAO land rights figures are global, not African. We found no evidence in the sources we read on how fertiliser shocks, soil cards or mobile money taxes affect women farmers specifically, and we say where we are inferring. Research stops at 16 November 2026.
Six findings
- The gap is large and uneven. Gaps range from 7% to 77% across twelve countries, and 4% to 62% remain after controlling for farm, household and plot characteristics.
- Women are not less engaged. Two in three working women in Sub Saharan Africa work in agrifood systems, a higher share than men.
- Crops and labour explain most of it. The paper’s largest driver is women’s concentration in lower value crops, and the second is lower use of, and returns from, labour.
- A land title alone did little. In a three year Mozambique experiment, registration raised perceived tenure security only temporarily and did not change inputs, investment or women’s decision making.
- Inputs without markets can backfire. The same experiment’s subsidised input and training package raised input use but lowered harvests.
- The African Union has set a target. The Kampala declaration commits to halving the male female yield gap by 2035.
The size of the gap
Donald and Tafere (World Bank Policy Research Working Paper 11414) synthesise country level decomposition studies. They find that plots managed by men yield more than plots managed by women, with gaps between 7% and 77% across twelve African countries. Controlling for farmer, household and plot characteristics shrinks the gap but leaves a conditional gap of 4% to 62%. The wider literature reports an overall gender gap in agricultural production of 18% to 61%. By our arithmetic the first range is 70 percentage points wide, which tells you how much depends on the country, the crop and the method.
Donald and Tafere, World Bank Policy Research Working Paper 11414 (June 2026)
Earlier single number estimates sit inside those ranges. The 2014 World Bank and ONE report Levelling the Field found that women produce 25% less than men in Malawi, the largest of its six countries, and 13% less in Uganda, the smallest. UN Women’s 2019 brief on Ethiopia, Malawi, Rwanda, Uganda and Tanzania estimates that closing the gap could raise crop production by up to 19% across the five. The 2014 report says that, with equal access to productive resources, women farmers could raise yields by 20 to 30 percent.
FAO (2023, female managed farms of the same size); World Bank and ONE (2014, output)
Women work the land
The gap is not about effort. The FAO notes that women make up well over half of the agricultural labour force in many Sub Saharan African countries. By our arithmetic the share of women’s employment in agrifood systems is six percentage points higher than men’s, 66% against 60%. UN Women says the gap arises from unequal access to inputs, not because women are less efficient farmers.
FAO, The Status of Women in Agrifood Systems (2023)
Pay shows the same pattern in wage work. The FAO finds women in agricultural wage employment earn 18.4% less than men, or 82 cents for every dollar. The wage gap is not a farm productivity measure, but it points the same way: the market pays less for the same work.
FAO (2023): women receive 82 cents for every dollar men earn. Remainder is our arithmetic
Where the gap comes from
Donald and Tafere name women’s concentration in lower value crops as the largest driver, and lower use of, and returns from, labour as the second. Social norms and unequal access to land and finance keep both in place. Levelling the Field adds that equal access to fertiliser, labour and training does not always give equal returns, and lists land rights, hired labour, labour saving tools, child care, better seed and fertiliser, tailored training, market access and education among its ten recommendations. The paper’s promising approaches are narrower: raising women’s adoption of cash crops, improving their use of labour and inputs, and tailored skills and information.
Land: necessary, not sufficient
Land rights matter because investment follows security. Goldstein and Udry (Journal of Political Economy, 2008) found in Akwapim, Ghana that people in powerful positions in the local political hierarchy have more secure tenure, invest more in land fertility and have substantially higher output, and that investment on a given plot tracks tenure security on that plot. The paper is about power, not gender, but its lesson holds: who can be sure of keeping the land is who invests in it. In Ethiopia, Deininger, Ali and Alemu (Land Economics, 2011) found that low cost certification of some 20 million plots raised tenure security, land investment and rental market participation. The World Bank’s June 2026 feature on Ethiopia’s landscape programme says 81% of registered holdings are owned by women, individually or jointly, and that certified landholders accessed a cumulative $65 million in microfinance loans.
The Mozambique experiment tempers that story. Brudevold Newman and colleagues (Working Paper 11421, June 2026) ran a three year field experiment with women farmers that randomly assigned statutory land registration and, separately, a subsidised input and climate smart agriculture training package. Demand for registration was high but effects were limited: it produced only a temporary gain in perceived tenure security, did not raise willingness to pay for inputs or downstream investment, and, despite gender targeted titling, did not improve women’s intra household security or decision making power. The input package raised input use but reduced harvests, with exploratory evidence of a shift toward a more shock prone crop, and the losses persisted after households returned to earlier practices. Non recipients with more treated network connections reduced their adoption of climate smart practices over time. Doing both together produced nothing beyond either alone.
The two results are not in conflict. A title is a document, and a gain in security needs the household, the community and the market to treat it as one. The Ethiopian count measures who is registered, not who decides, and the Mozambique result says registration is not decision power. Worldwide, the FAO finds that men hold land ownership or secure tenure at a higher rate than women in 40 of 46 reporting countries, and half of the countries reporting on the SDG indicator for women’s land rights have weak legal protections. It also finds the share of women among landowners rose in 10 of 18 countries over the last decade, with marked improvements in several Sub Saharan African countries.
FAO (2023). 46 countries reporting; global, not African. Shares are our arithmetic
Roads, markets and the informal firm
A crop only pays if it can leave the farm. A World Bank working paper on Rwanda by Musonera, Nsabimana and Overbeck combined all highway upgrades with 12 years of tax and census records. A summary of the paper reports that municipalities within 2 kilometres of newly paved highways saw local tax revenue rise 136% within five years, rental tax revenue 269% and trade licence revenue 55%. Small firms of one to nine workers, mostly informal, grew most, and upgrades that were not fully paved had little effect. Central government tax collections were largely unchanged, so the paper suggests revenue sharing and simpler registration. We found no gender breakdown in the sources we read, so we do not claim that the gains reached women traders. What the paper does show is that the firms that grew were small and mostly informal, which is the scale at which most farm trade happens.
Within 2 km of newly paved highways, five years after. Devdiscourse summary of the working paper
Money: credit, phones and taxes
Finance and records decide whether land and roads turn into income. The Global Findex 2025 shows Sub Saharan account ownership at 58% in 2024 and mobile money at 40%, but women in low and middle income economies are still nine points less likely than men to own a phone. An IMF working paper by Barczay, Hebous, Sawadogo and Wen (December 2025) studied mobile money taxes using transaction data from Cameroon, the Central African Republic and Mali. It finds sharp declines in usage after taxation, stronger among banked users, a disproportionate burden on unbanked and rural users, substitution into cash, and an excess burden equal to 35% of the revenue raised. The abstract does not break results out by sex, so the link to women farmers is our inference from the phone gap, and a question worth testing.
Nigeria’s new soil scheme shows the same design question. BusinessDay reported on 5 May 2026 that the federal government and GIZ unveiled a nationwide readiness assessment for the Nigerian Farmers’ Soil Health Card Scheme, with 774 soil testing laboratories, a national soil information system and capacity for up to two million farms a year. It aims to replace generic fertiliser advice with site specific recommendations and to link soil data to credit and insurance. The article mentions no gender component, while the Africa Gender Innovation Lab’s June 2026 newsletter says its Nigeria lab is working with the federal agriculture ministry on the card across 19 states. Whether women’s plots are tested and women’s names are on the cards is not yet measured.
The pressure on inputs
Women’s farms already use fewer inputs, so a price shock reaches them first. Fregene and Jenane (AfDB and World Bank, April 2026) write that about 80% of fertiliser used in Sub Saharan Africa is imported. The World Bank’s April 2026 Africa Economic Update cut the region’s 2026 growth forecast from 4.4% to 4.1% because of the Middle East conflict, according to the Think Africa Partnership newsletter. We have no source that links these shocks to women’s plots, so the effect on them is our inference.
What African policy already promises
The Kampala CAADP Declaration, adopted in January 2025, has six commitments, 35 intervention areas and 22 targets for 2035. Its targets include reducing the yield gap between male and female farmers by 50%, empowering 30% of women and 30% of youth, and at least 30% of agricultural land under sustainable land management. It also sets tripled intra African agrifood trade and US$100 billion of investment. By our arithmetic, halving a gap like the FAO’s 24% would leave about 12%. The Declaration’s biennial review begins in 2027 and reports to the AU Assembly start in 2028, so there is no verified progress yet.
Dates from the sources cited
A chain, not a title
A woman farmer’s income depends on a land right she can use, inputs she can afford, a crop with a market, a road to reach it and a way to be paid. The Mozambique result shows what happens when one link is fixed in isolation.
Our diagram of the links this article follows
What to do next
- Pair land registration with joint decision rules, legal protection and community recognition, since a title alone did not change women’s decision power in Mozambique.
- Test input packages against local markets before scale, and use the evidence that harvests fell when a subsidised package shifted farmers to a shock prone crop.
- Move women into higher value crops with tailored training and market links, since the paper names crop choice as the largest driver.
- Include women’s plots by design in soil health cards and extension, and report coverage by sex.
- Pave the roads that reach markets, not partial upgrades, and design local revenue sharing.
- Measure mobile money taxes by sex and rurality before rates are set.
- Report progress on the Kampala yield gap target with plot level data by sex, from 2027.
What we can and cannot claim
| Statement | Evidence status |
|---|---|
| Gender productivity gaps range 7% to 77% across twelve African countries | World Bank working paper 11414 abstract; not journal peer reviewed |
| Female managed farms have 24% lower land productivity | FAO 2023; different method from the working paper |
| 66% of women’s and 60% of men’s employment is in agrifood systems | FAO 2023, Sub Saharan Africa |
| Land registration did not raise women’s decision power in Mozambique | Experimental, working paper abstract; effect sizes not reported there |
| 81% of Ethiopian registered holdings are owned by women, individually or jointly | World Bank programme report; counts registration, not decisions |
| Paved highways raised Rwandan local revenue 136% within five years | Press summary of a working paper; not independently checked |
| Mobile money taxes cut use and burden rural and unbanked users | IMF working paper, three countries; no sex breakdown seen |
| Soil cards and fertiliser shocks affect women differently | Not shown in the sources we read; our inference |
| Kampala halves the yield gap by 2035 | A target, not a result |
Africa’s farm gender gap is not a puzzle about who works harder, because women already work the land at least as much as men; it is a puzzle about what the work is paid in, which crops it is allowed to grow, and whether a paper title turns into a decision. The June 2026 Mozambique experiment is the sharpest evidence yet that fixing one link, a title or a subsidised input package, can leave the rest of the chain untouched or even make it worse. The Kampala target of halving the gap by 2035 is reachable only if governments and partners measure plots, crops and decisions by sex, and fund the roads, markets and money that sit around the land.
Sources & notes
SOURCES: Donald and Tafere, The Agricultural Gender Gap in Sub Saharan Africa, Policy Research Working Paper 11414 (2026); Brudevold Newman and others, Property Rights, Agricultural Technology Adoption, and Investment, Working Paper 11421 (2026); Musonera, Nsabimana and Overbeck, The (Fiscal) Dividend of Infrastructure, Working Paper 11405 (2026); Devdiscourse, How Rwanda’s paved highways are boosting local revenues; World Bank and ONE, Levelling the Field press release (2014); UN Women, The gender gap in agricultural productivity in Sub Saharan Africa (2019); FAO, The Status of Women in Agrifood Systems (2023); Goldstein and Udry, The Profits of Power, Journal of Political Economy (2008); Deininger, Ali and Alemu, Land Economics (2011); World Bank, Ethiopia’s landscape management program; Barczay, Hebous, Sawadogo and Wen, Taxing Mobile Money, IMF Working Paper (2025); World Bank, Global Findex 2025; BusinessDay, FG launches nationwide soil assessment; Fregene and Jenane, Al Jazeera opinion (April 2026); AU-SAFGRAD, Kampala CAADP Declaration at a glance; AKADEMIYA2063, Kampala CAADP Declaration power brief; World Bank Policy Research Working Papers newsletter, June 2026; Think Africa Partnership newsletter, first half of 2026; World Bank Group Civil Society Update, June 2026.
Differences between sources: the FAO puts female managed farms 24% lower in land productivity, the working paper puts gaps from 7% to 77% across countries, and the 2014 report puts Malawi at 25% and Uganda at 13%; UN Women says closing the gap raises crop production by up to 19%, while the 2014 report says yields could rise 20 to 30 percent; the Kampala target list on the AU-SAFGRAD page was partly garbled, so we used the AKADEMIYA2063 brief to confirm the yield gap and women’s empowerment targets; the Think Africa newsletter describes a mobile money tax study across 15 Sub Saharan countries, while the IMF page we read lists three countries for its transaction data.
Our own arithmetic: the 70 point width of the gap range; the six point difference between 66% and 60%; the 18 cent wage gap and 82 cents; 40 of 46 countries as 87%; the 12% left after halving a 24% gap; and Malawi’s gap at about 1.9 times Uganda’s.