The Protein Sovereignty Index

Regional assessment

East Africa & Horn (excl. Ethiopia)

We assess East Africa and the Horn as Critically exposed (39.3). The region grows most of the protein it eats, holds 131.8 million cattle and supplies 82 per cent of Saudi Arabia's live sheep and goat imports,[15] yet delivers only 59.0 grams of protein a day, of which 15.1 grams are animal, across the nine of its ten members for which a figure exists, 293.1 million of its 296.7 million people.[3][7][12]

Access and affordability is now the binding constraint, scored at 17: on a population-weighted basis, 76.0 per cent of people across the seven members with published data cannot afford a healthy diet, a figure that almost certainly understates the truth because Sudan and Somalia, the two members in open war, have no published estimate at all.[41]

Shock endurance is the next weakest pillar, at 30: cereal stocks run to about 33 days, four of the ten member states report less than a month of import cover and two more have not published a usable figure for decades, and the humanitarian pipeline that substitutes for national reserves was between 29 and 40 per cent funded in July 2026.[13][12][30]

Covers: Kenya, Tanzania, Uganda, Rwanda, Burundi, Somalia, Djibouti, Eritrea, South Sudan, Sudan

The protein system

Weighted by protein consumption, Tanzania accounts for roughly 24.5 per cent of the region, Sudan 21.1, Kenya 18.8, Uganda 17.2, South Sudan 5.2, Rwanda 4.8, Somalia 4.5, Burundi 3.2 and Djibouti 0.5.[3][12] Eritrea has no protein supply figure in either FAOSTAT release and is excluded from the weighting; its 3.6 million people are 1.2 per cent of the regional population.[12] Four states determine about 82 per cent of the score.[3][12]

This is a plant-and-pasture system. Roughly 44 of the 59.0 grams come from crops, including 4.19 million tonnes of dry beans and 2.84 million tonnes of groundnuts in 2024.[5] The animal fraction comes almost entirely from grazed ruminants rather than from anything fed on purchased grain: 26.0 million tonnes of milk in 2024, 2.18 million of it from camels, plus 1.81 million tonnes of beef, 899,000 tonnes of sheep and goat meat, 541,000 tonnes of chicken and 523,000 tonnes of eggs.[5] Fisheries matter more than the arid-Horn framing suggests, at 1.46 million tonnes landed wild and 290,000 tonnes farmed.[8][9]

The spread of outcomes is as wide as anywhere in the index, from 38.9 grams a day in Burundi and 39.8 in Somalia to 74.0 in South Sudan and 75.5 in Djibouti.[3][1] The high readings show the gap between supply and access: South Sudan's 30.3 grams of animal protein is the region's highest, and 57 per cent of its people were in Crisis or worse during the 2025 lean season.[2][19]

Durable domestic capacity: 43

Quantitative score 50, adjusted down 7 points

Across the five members with current FAO Food Balance Sheets, population-weighted protein self-sufficiency is 86.1 per cent, computed by weighting each item's production-over-domestic-supply ratio, capped at 100 per cent, by its protein in food supply: Tanzania 96.0, Uganda 92.6, Rwanda 83.7, Kenya 70.0 and Djibouti 12.7, with protein import dependency between 6.6 and 25.5 per cent.[3] The ranking, Tanzania > Uganda > Rwanda > Kenya >> Djibouti, holds under every protein-content mapping tested. Measured that way the region looks close to sovereign, and the measure misleads in three ways.

First, the level. A nominally self-sufficient system delivering 38.9 grams a day is not a functioning protein system.[3] Undernourishment stands at 56.5 per cent in Somalia and 34.9 per cent in Kenya, and Kenyan dietary energy supply adequacy is 94 per cent, below requirement.[4]

Second, the base is overwhelmingly rain-fed. Only 4.9 per cent of the region's 56.6 million hectares of arable land is equipped for irrigation, and the three largest rain-fed members sit below 5 per cent: Uganda 0.1 per cent, Tanzania 2.7 and Kenya 4.3.[36] Output therefore tracks two rainy seasons directly, and Somalia's 2025 deyr harvest came in 75 per cent below average at 15,600 tonnes.[18] Sudan is the exception at 8.8 per cent, in the Gezira and other Nile schemes the war has disrupted.[36]

Third, the largest fishery has been fished past its ceiling. Our computation from FAO's species-level landings series shows Nile perch capture across Kenya, Tanzania and Uganda falling 51 per cent, from 371,526 tonnes in 1990 to 183,069 in 2023, while landings of silver cyprinid, the small dagaa lower down the food web, rose from 46,738 to 321,178 tonnes.[10] Kenyan inland capture fell 42 per cent in a decade.[10] Tonnage has been held up by catching smaller and cheaper fish, the signature of a stock past its limit, and 37.5 per cent of assessed Western Indian Ocean stocks were overfished at the last assessment.[11]

We applied a downward adjustment of seven points for active conflict degrading nominal capacity. Sudan, South Sudan and Somalia hold 83.5 million people, 28 per cent of the region and 31 per cent of the consumption weight, and all three are at war: 22.0 to 22.99 million Sudanese needed food assistance through September 2026 with credible famine risk in nine named areas.[28] FAOSTAT continues to report notional output these populations cannot reach.

Resource headroom: 51

Quantitative score 58, adjusted down 7 points

In feed-conversion terms this is one of the least resource-hungry protein systems in the index. Ruminants graze pasture and crop residues and almost no cereal passes through livestock. Kenya, the largest formal feed market with 57.5 million people, imported 111,845 tonnes of soybean meal and 421 tonnes of fishmeal in 2024.[14] Aquaculture farms tilapia and catfish on largely plant-based rations rather than carnivores on wild fishmeal, so the 290,000 tonnes produced in 2024 draws almost nothing from a peaked wild stock.[9][14]

Nutrient headroom is very large on paper: total application in 2023 ranged from 0.18 kilograms per hectare of cropland in South Sudan and 1.99 in Uganda up to 17.6 in Tanzania and 43.5 in Kenya, so yields could rise substantially without a hectare of new land.[6]

Land and water headroom is nonetheless exhausted at one end of the region. Rwanda has 0.083 hectares of arable land per person and Burundi 0.097, and Kenya at 382 cubic metres of renewable internal freshwater per person, Somalia at 337, Djibouti at 264 and Sudan at 81 sit far inside the 1,700 cubic metre stress threshold.[12] Demography consumes headroom faster than yields have risen: the region grew from 228.4 million people in 2015 to 296.7 million in 2025, a compound 2.65 per cent, with South Sudan at 3.93 per cent a year and Somalia at 3.48.[12] Set against a longer window, arable land grew from 38.80 to 56.61 million hectares between 2000 and 2024, up 45.9 per cent, but population grew 93.9 per cent over the same span, so arable land per person fell from 0.260 to 0.196 hectares, down 24.7 per cent.[36][46] The expansion is real; it has simply not kept pace with people.

The irrigation-equipped share of that expanding land base tells the same story from a different angle: it fell from 6.1 per cent in 2000 to 4.9 per cent in 2024, even as the arable area beneath it grew by nearly half.[36] Kenya, 1.7 to 4.4 per cent, and Tanzania, 1.9 to 2.7, are the only two members where the irrigated share genuinely rose; Burundi, Rwanda, Eritrea and Uganda all fell, Somalia's 19.2 to 18.2 per cent rests on an irrigated area unchanged in absolute hectares since 2000, and Sudan and South Sudan combined fell from 11.5 per cent (undivided Sudan, 2000) to 8.0 (2024).[36] Two decades of cropland growth have been overwhelmingly rain-fed rather than banked. Kenya's own flagship counter-example makes the scale of the gap concrete: the Galana-Kulalu Food Security Project was launched in 2015 on a planned 1.0 million acres, reached only 10,000 of 20,000 leased acres in active use by 2020, and by November 2025 had 2,500 acres running under 18 centre pivots, roughly a quarter of one per cent of the original target a decade in.[47][48]

We applied a downward adjustment of four points because the largest land reserve is unreachable. Sudan holds 0.420 hectares of arable land per person and South Sudan 0.217, the two highest figures in the region and the two least usable. Sudan's Gezira scheme, 924,000 hectares and 42 per cent of the country's irrigated area, lost 9 per cent of its cultivated land between 2019/20 and 2023/24 and recorded 733 security incidents between April 2023 and August 2024; wheat production fell 58 per cent in the 2023-24 season after the RSF captured Wad Madani, Gezira state's capital, in December 2023, with only a fragile recovery through 2025.[49][50] South Sudan applies 0.18 kilograms of nutrient per hectare and has 0.8 per cent of its arable land equipped for irrigation.[6][36][28] Land that cannot be farmed, fertilised or connected to a market is not headroom, and a hectares-per-person indicator cannot see the difference.

We applied a further three points on the same realisability ground, this time to the nutrient headroom. The paper fertiliser gap described above assumes continued import of nitrogen, and that is the same channel scored as failing in upstream dependence: tanker traffic through the Strait of Hormuz has fallen by more than 90 per cent since March 2026, and roughly half of Kenyan and a third of Tanzanian nitrogen by value originates in Gulf suppliers now largely cut off.[33][14] Nutrient headroom the region cannot currently import is no more realisable than land it cannot farm.

Agricultural credit points the other way, and is why the deduction stops at seven points rather than going further. Across the six members that report a credit-to-agriculture series, covering 87.7 per cent of the region's population, a population-weighted 10.7 per cent of total bank credit goes to agriculture, forestry and fishing, above the FAOSTAT Africa aggregate of 4.0 per cent: Sudan 18.6 per cent in 2022, Burundi 15.8, Uganda 11.9, Tanzania 10.6, Kenya 3.9 and Rwanda 0.8.[45] The agriculture orientation index, which weighs that credit share against agriculture's share of GDP, is a population-weighted 0.42 against parity of 1.00, so credit still trails the sector's economic weight everywhere it is measured, but the region's largest economies are not starved of agricultural capital in gross terms. Somalia, Djibouti, Eritrea and South Sudan, 35.7 million people, publish no agriculture-credit breakdown at all; Somalia carries no FAOSTAT credit record of any kind.[45] The constraint this evidence points to is not an absent capital market, it is capital that has financed rain-fed area rather than irrigation.

Import exposure: 51

Quantitative score 54, adjusted down 3 points

As a ratio this is the better pillar, and the aggregate conceals two different systems. Across the five members with USDA balance sheets, covering 72 per cent of the population, cereal imports equal 33.9 per cent of consumption in MY2025/26.[13] On FAO's measure, cereal import dependency is 1.8 per cent in Tanzania, 15.3 in Uganda, 35.3 in Rwanda, 49.5 in Kenya and 100 in Djibouti.[4]

The region is also a net exporter of animal protein. Sudan supplied 45.3 per cent, Somalia 27.7 and Djibouti 9.1 of Saudi Arabia's live sheep and goat imports in 2024, 82.1 per cent between them, worth USD 727 million, USD 265 million and USD 82 million, and Somalia sold Oman a further USD 161 million.[15] Uganda and Tanzania exported 27,551 tonnes of fish fillets, predominantly Nile perch, worth USD 179 million, mainly to the European Union and Israel.[16] The region ships out its highest-value protein and eats dagaa.

Where exposure exists it is severe and concentrated. Somalia grows no wheat, imported 1.15 million tonnes of it plus 650,000 tonnes of rice in MY2025/26, and runs a cereal import dependency of 92.2 per cent.[13] Kenya meets 2.91 million tonnes of wheat consumption with 2.90 million tonnes of imports, 99.7 per cent of the total.[13] Russia supplied 81.5 per cent of Kenya's wheat, 84.3 per cent of Uganda's and 66.1 per cent of Tanzania's in 2024.[14] India and Pakistan together supplied 81.7 per cent of Kenyan rice, and India alone shipped 500,321 tonnes to Somalia and 215,093 to Djibouti.[14][15] Affordability compounds geography: food imports equal 266 per cent of Somalia's total merchandise exports, 81 per cent of Burundi's and 57 per cent of Sudan's.[4]

We applied a net downward adjustment of three points from two opposing grounds. Against the region, Black Sea wheat reaches East African ports through Suez and Bab el-Mandeb, a corridor under attack since November 2023 that also serves Port Sudan, Djibouti and the Eritrean ports; UNCTAD recorded ship tonnage entering the Gulf of Aden down by more than 70 per cent between the first half of December 2023 and the first half of February 2024.[37] One origin and one route therefore carry most of the wheat three large importers eat. In its favour, much measured exposure is intra-regional and carries no chokepoint risk: 93.6 per cent of Kenya's 305,686 tonnes of maize imports arrived overland from Tanzania in 2024, and Kenyan soybean meal came from India, Zambia, Malawi and Uganda rather than the Atlantic soy complex.[14]

Upstream dependence: 41

Quantitative score 46, adjusted down 5 points

Upstream dependence is total in principle and shallow in practice. Not one of the nine member states covered by FAOSTAT's fertiliser series produced any nitrogen fertiliser in 2023, Somalia being unreported, and the only nutrient produced anywhere in the region was 18,028 tonnes of P2O5 in Tanzania, so every kilogram of nitrogen applied across 296.7 million people is imported.[6] Because application rates are among the lowest on earth, the crop system is only lightly hostage to that, and the ruminant system supplying most animal protein uses indigenous genetics, pasture and crop residues and buys almost nothing. Even dairy trade is regional, with 82.8 per cent of the value of Kenyan milk powder imports coming from Uganda.[14]

The genuine dependence is energy, and its concentration is the weakness. Kenya imported 4.91 million tonnes of refined petroleum worth USD 4.13 billion in 2024, 52.4 per cent of it by value from the United Arab Emirates, 12.2 from Oman and 11.6 from Kuwait.[14] That fuel runs the mills, pumps, cold chain and the haulage that carries Tanzanian maize to Nairobi. Fertiliser follows the same map: Saudi Arabia supplied 27.2 per cent and Qatar 22.5 per cent of the value of Kenya's 325,590 tonnes of nitrogenous fertiliser, and Qatar 21.3, Russia 19.1 and Oman 14.5 per cent of Tanzania's 416,960 tonnes.[14]

We applied a downward adjustment of five points because that chokepoint is closed. FAO reports roughly 30 per cent of world fertiliser trade normally transits the Strait of Hormuz and that tanker traffic there has fallen by more than 90 per cent since March 2026.[33] About half of Kenyan nitrogen by value and about 64 per cent of the value of its refined product imports originate inside the strait.[14] A static import-share score reads these as ordinary trade; in July 2026 they are a live supply failure. No project comparable to Ethiopia's Gode urea complex, a 3 million tonne plant agreed in August 2025 whose construction broke ground in October 2025, is under way anywhere in this region.[38][39]

Access and affordability: 17

Quantitative score 21, adjusted down 4 points

This is the region's weakest pillar, and it is weak because the food system described above is one that most of the people in it cannot afford. Among the seven members with a published Cost and Affordability of a Healthy Diet estimate, covering 74.7 per cent of the region's population, a population-weighted 76.0 per cent of people could not afford a healthy diet in 2025, from Djibouti's 49.8 per cent, the region's best reading, to South Sudan's 93.9 per cent and Burundi's 90.9 per cent, a combined 168.5 million people.[41] Population-weighted undernourishment across the members with a figure is 26.9 per cent, led by Somalia at 56.5 per cent,[4] and child stunting, measured for all ten members, is 28.4 per cent, led by Burundi at 55.3 per cent and Eritrea at 48.0 per cent.[4]

Price volatility is split down the middle of the region. Maize, sorghum and wheat flour prices were stable or marginally lower in May 2026 in Burundi, Kenya, Uganda, South Sudan and Tanzania on the first bimodal harvests, while rising in Djibouti, Somalia and Sudan,[43] and Kenyan food inflation ran at a manageable 8.6 per cent in the year to June 2026, against 7.7 per cent in the year to November 2025.[44] Sudan is the sharpest case: the WFP Local Food Basket rose 21 per cent year on year to March 2026, wheat flour 33 per cent, goat and sheep 57 per cent and groundnuts 100 per cent, against a currency that fell 43 per cent on the parallel market and 55 per cent at commercial banks over the same period, and WFP records besieged areas facing outright restricted access to food items.[42]

We applied a downward adjustment of four points because Sudan and Somalia, 71.3 million people, have no published unaffordability estimate at all, and every available proxy for them, Somalia's undernourishment and Sudan's currency collapse, points towards conditions at or beyond the region's worst measured case rather than better than it.

Shock endurance: 30

Quantitative score 27, adjusted up 3 points

This is the second weakest pillar and it is failing in fact rather than in scenario.

Physical cover is thin: cereal stocks across the five members with USDA balance sheets equal 9.2 per cent of consumption, about 33 days, from 15.0 per cent in Uganda down to 5.0 in Sudan and nothing in Somalia.[13] Financial cover is thinner. Kenya held 4.0 months of import cover in 2024 and Rwanda 4.2, Uganda 2.4, Djibouti 0.85 and Burundi 0.73; Sudan last reported 0.19 months in 2017, South Sudan 0.17 in 2023, Eritrea 0.85 as long ago as 2000, and Somalia has had no reserves observation since 1989.[12] The capacity to outbid on world markets that carries Japan and the Gulf through their own import dependence is simply absent here.

That leaves the humanitarian system as absorber of last resort, and it is not absorbing. At 31 July 2026, Sudan's 2026 response plan had received USD 1.156 billion of a USD 2.866 billion requirement (40.3 per cent), South Sudan's USD 0.507 billion of USD 1.463 billion (34.7 per cent) and Somalia's USD 0.248 billion of USD 0.852 billion (29.1 per cent).[30] WFP needs a further USD 579 million for Sudan to October 2026 and projects 825,000 children under five in severe acute malnutrition during the year,[31] while its Somalia operation reports critical shortfalls against 6.5 million people in crisis and 3.9 million displaced.[32]

The shocks are already running: 2 million Somalis in Emergency in early 2026, about 50 per cent more than in October to December 2025[18] and on WFP's account double the early-2025 level;[32] 3.7 million Kenyans in Crisis or worse in mid-2026, up 32 per cent year on year, with FEWS NET flagging a credible scenario in which funding fails after September;[20][29] 1.14 million Burundians in Crisis after bean losses of 80 per cent in affected regions;[24] and 8.6 million people internally displaced in Sudan alone as of mid-May 2025.[17] FEWS NET puts displacement across its wider East Africa coverage, which includes Ethiopia, at more than 12.8 million.[27]

We applied an upward adjustment of three points for demonstrated substitution, the ground the methodology admits for documented wartime adaptation. Tanzania carries a structural maize surplus, exporting about 1.0 million tonnes in MY2025/26 against 7.5 million tonnes of production, and functions as the region's buffer stock rather than a market participant, supplying 93.6 per cent of Kenyan maize imports.[13][14] Sudan's 2024 harvest of 6.7 million tonnes, 60 per cent above 2023, was achieved in the second year of a civil war.[17] Household dietary breadth across five cereals, bananas, cassava, sweet potato and 4.19 million tonnes of pulses is genuine,[5] and pastoral herds convert to hard currency through established Gulf demand.[15] That is why the score is 31 rather than in the twenties.

Crisis purchasing power adds nothing positive. External debt service consumes 27.2 per cent of Kenya's export earnings and 13.9 per cent of Uganda's, a claim on foreign exchange that competes directly with the reserves a shock would require those two, the region's largest formal import economies, to spend outbidding other buyers.[12] The lighter burdens elsewhere buy no comfort: Somalia's 0.66 per cent, Djibouti's 2.5 and Sudan's 2.96 reflect arrears, debt relief and war-disrupted servicing rather than solvency, and Eritrea and South Sudan report nothing at all.[12] No member operates a documented national social protection system with the capacity to scale rapidly to a crisis caseload, so the humanitarian pipeline described above, funded at 29 to 40 per cent, is in practice the region's entire scalable safety net rather than a backstop to a domestic one.

Trajectory

Three levers would move this outcome, one of them already running backward.

The first is fertiliser and yield. The region applies between 0.18 and 43.5 kilograms of nutrient per hectare against a world average many times higher, so the cheapest protein available lies in the yield gap rather than in new land.[6] Closing part of it requires imported nitrogen at a price smallholders will pay, which is precisely what the 2026 Gulf disruption threatens.[33] The absence of any large domestic nitrogen project, against the Gode complex under construction next door in Ethiopia, is the most consequential gap in the region's industrial policy.[38][39]

The second is the fishery. Lake Victoria supports Africa's largest freshwater fishery and its highest-value stock has halved since 1990.[10] Aquaculture is the substitution route and it is working: Tanzanian farmed output rose from 10,166 tonnes in 2013 to 138,607 in 2024, a thirteenfold increase, and Rwandan from 1,165 to 8,392.[9] Because the species farmed are herbivores on plant rations, that growth is additive rather than drawn from the wild stock.

The third is peace. Sudan, South Sudan and Somalia hold 31 per cent of the region's protein consumption and essentially all of its famine risk.[3][12] A settlement in Sudan alone would restore a 51.7 million-person market that has already produced 6.7 million tonnes of cereals under fire,[17] reactivate the region's largest livestock export trade,[15] and remove the single largest claim on a humanitarian budget now covering under 40 per cent of assessed need.[30] Continued war on current funding would pull endurance towards the twenties and the overall score towards the mid-thirties.

What will not shift this outcome is the land expansion investors have pursued. Land Matrix records 1.98 million hectares of concluded large-scale acquisitions in South Sudan, dominated by a single 1.68 million hectare Emirati holding, and 802,628 hectares in Sudan with Gulf and Arab state investors prominent.[34] These are protein-sovereignty assets for the buyers, and on present evidence they have not raised protein supply in either host.

Country notes

Kenya. The most import-exposed large economy: protein self-sufficiency 70.0 per cent, cereal import dependency 53.7 per cent, wheat effectively wholly imported with Russia at 81.5 per cent, and inland catch down 42 per cent since 2013.[3][13][14][10] Offsetting that, the region's largest reserves at USD 12.4 billion, the heaviest fertiliser use and 93.6 per cent of maize from Tanzania.[12][6][14] Still 3.7 million people in Crisis or worse in mid-2026.[20]

Tanzania. The strongest member and the region's buffer: protein self-sufficiency 96.0 per cent, cereal import dependency 1.8 per cent, 39.1 million cattle, 12.8 million tonnes of cereals forecast for 2026 and the only fertiliser nutrient produced regionally.[3][4][7][21][6] Only 5 per cent of the analysed population is severely food insecure through January 2027.[21] Wheat is the gap, 1.55 million tonnes imported with Russia at 66.1 per cent.[13][14]

Uganda. Near self-sufficient at 92.6 per cent protein, with the region's highest fish consumption at 15.4 kilograms a head in 2023, 538,160 tonnes of capture and 103,420 of aquaculture, and a maize surplus sold into Kenya and South Sudan.[3][8][9][13] At 2.0 kilograms of nutrient per hectare its yield headroom is the largest outside South Sudan.[6] It hosts more than 2 million refugees and asylum seekers, about 1 million of them from South Sudan, of whom 712,000 were in Crisis or worse to February 2026.[22]

Rwanda. Protein self-sufficiency of 83.7 per cent on 0.083 hectares of arable land per person, through intensive beans and aquaculture that grew sevenfold in a decade.[3][12][5][9] Animal protein of 6.3 grams a day is second lowest in the region.[3] The 2025 harvest of 943,000 tonnes was 12 per cent above average, FAO calls the country generally food secure, and 4.2 months of import cover is the region's best current reading.[23][12]

Burundi. The weakest member on protein adequacy and among the weakest in the world: 38.9 grams a day with 3.0 animal, and 1.23 million cattle for 14.4 million people.[1][2][7][12] The 2026A season brought maize losses near 50 per cent and bean losses of 80 per cent in affected regions, with 1.14 million people in Crisis.[24] Food imports equal 81 per cent of merchandise exports and import cover was 0.73 months.[4][12] Beans are effectively the whole protein system.[5]

Somalia. Structurally the most exposed: no wheat grown, 1.15 million tonnes imported plus 650,000 of rice, cereal import dependency 92.2 per cent against 147,700 tonnes of output.[13][18] Undernourishment 56.5 per cent, 6.5 million in Crisis or worse with 2 million in Emergency.[4][18] Food imports cost 266 per cent of merchandise exports and no reserves exist.[4][12] Livestock is the one real asset, USD 426 million of live small ruminants sold to Saudi Arabia and Oman in 2024.[15]

Djibouti. The purest importer in the index outside the Gulf: cereal import dependency 100 per cent, protein self-sufficiency 12.7 per cent, arable land 0.003 hectares per person.[4][3][12] Protein supply is the region's highest at 75.5 grams because it can pay, on GDP per capita of USD 3,906, 1.7 times Kenya's and the highest in the region, funded by Ethiopian transit rents and foreign bases.[3][12][35] Income is a service rent, import cover is 0.85 months, and it sits on Bab el-Mandeb.[12][25]

Eritrea. The region's statistical blank, scored with the least confidence. FAO's September 2025 brief carries no production, import or food security figures, and the last import-cover reading is 0.85 months in 2000.[26][12] What is measurable is discouraging: 1.87 kilograms of nutrient per hectare with no phosphate or potash use, and 2,459 tonnes of capture plus 26 tonnes of aquaculture from a 1,200-kilometre Red Sea coastline.[6][8][9]

South Sudan. The most extreme case of protein that exists on paper: animal protein of 30.3 grams a day, the region's highest, from 14.3 million cattle.[3][7] Alongside that, 57 per cent of the population in Crisis or worse in the 2025 lean season with about 83,000 in Catastrophe, and credible famine risk in Akobo, Nyirol, Nasir and Ulang.[19][40] It has the best water endowment at 2,359 cubic metres per person, uses 0.18 kilograms of nutrient per hectare and irrigates 0.8 per cent of its arable land.[12][6][36] Reserves were USD 16 million in 2024.[12]

Sudan. The largest single risk and the reason endurance scores where it does. It holds 31.3 million cattle, produced 5.39 million tonnes of sorghum in 2024, supplies 45.3 per cent of Saudi Arabia's live small-ruminant imports and imports 2.7 million tonnes of cereals, about 90 per cent of it wheat.[7][5][15][17] The 2024 harvest of 6.7 million tonnes, 60 per cent above 2023, was real resilience under fire.[17] It coexists with 22.0 to 22.99 million people needing assistance, 8.6 million displaced, 825,000 children projected in severe acute malnutrition, and a response plan 40.3 per cent funded.[28][17][31][30]

Somaliland. Listed as a member of this entity and assessed without a separate score. It reports to no international statistical agency, so its production and trade sit inside Somalia's figures where they are recorded at all, and the Somalia assessment above carries it. One point bears on the region's scoring: the live small-ruminant trade that earns the Somali pastoral economy USD 426 million from Saudi Arabia and Oman moves substantially through Berbera,[15] so the single genuine protein asset in Somalia's assessment lies largely in territory Mogadishu does not administer. Berbera's throughput is not quantified at primary-source quality.

Country notes

Kenya
The region's most import-exposed large economy and the one that most resembles a middle-income importer. Protein self-sufficiency is 70.0 per cent and protein import dependency 24.9 per cent (2022), cereal import dependency 53.7 per cent in MY2025/26, and wheat is effectively wholly imported at 2.9 million tonnes against 220,000 tonnes of production, with Russia supplying 81.5 per cent of it in 2024. Renewable internal freshwater is 382 cubic metres per person, well inside water stress, and inland capture fisheries have fallen 42 per cent since 2013. Offsetting that, Kenya holds the region's largest reserves at USD 12.4 billion and 4.0 months of import cover, applies the most fertiliser (43.5 kg nutrient per hectare) and sources 93.6 per cent of its maize imports from Tanzania. It nonetheless had 3.7 million people in Crisis or worse in April to June 2026, up 32 per cent year on year, and undernourishment at 34.9 per cent is the region's second highest.
Tanzania
The strongest member and the region's buffer. Protein self-sufficiency is 96.0 per cent, protein import dependency 6.6 per cent and cereal import dependency 1.8 per cent on FAO's measure. It holds the largest cattle herd in the region at 39.1 million head, produces 1.46 million tonnes of dry beans and, on USDA's MY2025/26 balance sheet, 11.2 million tonnes of cereals, exports about 1.0 million tonnes of maize a year and is the only member producing any fertiliser nutrient at all (18,028 tonnes of P2O5). Aggregate 2026 cereal output is forecast near average at 12.8 million tonnes and only 507,000 people, 5 per cent of the analysed population, face severe food insecurity through January 2027. Its weaknesses are wheat, which is almost entirely imported at 1.55 million tonnes with Russia at 66.1 per cent, and stale macroeconomic data: the last reported import cover is 5.4 months from 2018.
Uganda
Near protein self-sufficiency at 92.6 per cent with the region's highest fish consumption at 15.4 kilograms per person in 2023 (15.9 kilograms in 2022), 538,160 tonnes of capture fisheries and 103,420 tonnes of aquaculture. Cereal import dependency is 15.3 per cent and it exports maize into Kenya and South Sudan. It applies almost no fertiliser, 2.0 kilograms of nutrient per hectare, so yields are far below potential and there is more nutrient headroom here than anywhere else in the region outside South Sudan. It hosts more than 2 million refugees and asylum seekers, about 1 million of them from South Sudan, of whom 712,000 were in Crisis or worse between August 2025 and February 2026, and 84.3 per cent of its wheat came from Russia in 2024. Import cover is 2.4 months.
Rwanda
Protein self-sufficiency of 83.7 per cent on 0.083 hectares of arable land per person, the second tightest land constraint in the region, achieved through intensive beans (478,000 tonnes) and a fast-growing aquaculture sector (8,392 tonnes in 2024, up from 1,165 tonnes in 2013). Animal protein supply is only 6.3 grams a day in 2022, second lowest after Burundi, and milk consumption is 12.9 kilograms per person against Kenya's 88.5. Cereal import dependency is 35.3 per cent and protein import dependency 25.5 per cent. The 2025 cereal harvest was 943,000 tonnes, 12 per cent above the five-year average, the country is described by FAO as generally food secure, and import cover of 4.2 months is the region's best current reading.
Burundi
The weakest member on protein adequacy and among the weakest in the world. Protein supply is 38.9 grams per person per day with only 3.0 grams of animal protein, milk consumption is 5.4 kilograms a year and the cattle herd is 1.23 million head for 14.4 million people. Arable land is 0.097 hectares per person. The 2025 cereal harvest was 789,000 tonnes, 5 per cent below average, and the 2026A season brought losses of about 50 per cent for maize and 80 per cent for beans in affected regions, with 1.14 million people projected in Crisis from January to March 2026. Food imports equal 81 per cent of merchandise exports and import cover was 0.73 months in 2023. Beans at 505,000 tonnes are the entire protein system.
Somalia
Structurally the most exposed member. It produces no wheat, imported 1.15 million tonnes of it in MY2025/26 plus 650,000 tonnes of rice, and its cereal import dependency is 92.2 per cent against total 2025 output of 147,700 tonnes. The 2025 deyr harvest was 75 per cent below average at 15,600 tonnes. Undernourishment is 56.5 per cent, dietary energy adequacy 80 per cent, and 6.5 million people (33 per cent) were in Crisis or worse in February to March 2026 with 2 million in Emergency. Food imports cost 266 per cent of total merchandise exports and there are no usable foreign reserves. Its one genuine asset is livestock: 82,023 tonnes of live sheep and goats to Saudi Arabia and 24,089 tonnes to Oman in 2024, worth USD 426 million combined, which is also why animal protein supply at 16.9 grams a day exceeds Kenya's.
Djibouti
The purest importer in the index outside the Gulf. Cereal import dependency is 100 per cent, protein self-sufficiency 12.7 per cent, protein import dependency 92.5 per cent, arable land 0.003 hectares per person and renewable internal freshwater 264 cubic metres. Yet protein supply is 75.5 grams a day, the region's highest, because it can pay: GDP per capita is USD 3,906, the highest in the region and 1.7 times Kenya's USD 2,363, funded by port and logistics rents from the Ethiopian transit corridor and foreign military bases. It also re-exports Horn livestock, sending 26,832 tonnes of live sheep and goats to Saudi Arabia in 2024. The vulnerabilities are that income is a service rent rather than production, import cover is 0.85 months, it sits directly on Bab el-Mandeb, and 230,000 people (20 per cent) still faced Crisis or Emergency in July to December 2025.
Eritrea
The region's statistical blank and scored with the least confidence. FAO's September 2025 country brief contains no production, import or food security figures; there is no protein supply figure in either FAO series; the last World Bank GDP per capita observation is 2011 and the last import cover reading 2000. What is measurable is discouraging: 1.87 kilograms of fertiliser nutrient per hectare, no phosphate or potash use at all, 3.1 per cent of arable land equipped for irrigation, 2,459 tonnes of capture fisheries and 26 tonnes of aquaculture from a 1,200-kilometre Red Sea coastline, and a 2.16 million head cattle herd for 3.6 million people. Arable land per person is a relatively comfortable 0.199 hectares. We treat Eritrea as a low-capacity, low-endurance member and flag the data absence.
South Sudan
The most extreme case of protein that exists on paper. Animal protein supply is 30.3 grams a day, twice the regional average and the highest in the region, from 14.3 million cattle and 197.5 kilograms of milk per person a year, and total protein supply is 74.0 grams. Alongside that, 7.7 million people (57 per cent of the population) were in Crisis or worse in the April to July 2025 lean season with about 83,000 in Catastrophe, maize and sorghum prices in Juba almost quadrupled between October 2024 and June 2025, and FEWS NET identifies credible famine risk in Akobo, Nyirol, Nasir and Ulang. It has the region's best water endowment at 2,359 cubic metres per person and 0.217 hectares of arable land per person, and uses 0.18 kilograms of fertiliser nutrient per hectare and irrigates 0.8 per cent of its arable land. Reserves stood at USD 16 million in 2024, 0.17 months of imports. Land Matrix records 12 concluded large-scale acquisitions covering 1.98 million hectares, dominated by a single 1.68 million hectare UAE holding.
Sudan
The region's largest single risk and the reason the endurance score sits where it does. Sudan has 31.3 million cattle, produced 5.39 million tonnes of sorghum in 2024 and supplies 45.3 per cent of Saudi Arabia's live sheep and goat imports, worth USD 727 million in 2024, while importing 2.7 million tonnes of cereals of which about 90 per cent is wheat. The 2024 harvest of 6.7 million tonnes was 60 per cent above 2023 and 7 per cent above the five-year average despite the war, evidence of genuine agricultural resilience. It coexists with 22.0 to 23.0 million people needing food assistance through September 2026, credible famine risk in seven North Darfur areas plus Dilling and Kadugli, 8.6 million internally displaced as of May 2025, 825,000 children projected to suffer severe acute malnutrition in 2026, and a 2026 response plan 40.3 per cent funded. Renewable internal freshwater is 81 cubic metres per person, the lowest in the region, fertiliser use is 7.0 kilograms per hectare, and Land Matrix records 31 concluded large-scale land deals covering 802,628 hectares with Gulf and Arab state investors prominent, including a 161,874 hectare UAE port-and-farmland project.
Somaliland
Listed as a member of this entity in the index's entity model and assessed without a separate score. Somaliland is not a UN member state and appears in no FAOSTAT, World Bank, USDA or UN Comtrade series as a reporter, so its production, trade and food security are recorded inside Somalia's figures where they are recorded at all, and its protein position is carried by the Somalia assessment above. One structural point matters for the region's scoring: the live small-ruminant trade that gives the Somali pastoral economy its hard-currency earnings, USD 426 million from Saudi Arabia and Oman in 2024 on mirror statistics, moves substantially through Berbera rather than through the ports controlled from Mogadishu, so the single genuine protein asset in Somalia's assessment sits largely in territory Mogadishu does not administer. Berbera's livestock throughput is not quantified from a primary source. The wider judgement, a pastoral system with real export value, negligible cereal production, no reserves and no published national accounts, applies to Somaliland as it does to Somalia.

Sources

  1. Daily per capita protein supply: values for Burundi, Somalia, Sudan and South Sudan retained from the FAOSTAT release superseded on 28 October 2025 (FAO review ongoing) · Our World in Data (2025 (data 2022-2023))
  2. Daily protein supply from animal and plant-based foods: animal-protein values for Burundi, Somalia, Sudan and South Sudan retained from the superseded FAOSTAT release · Our World in Data (2025 (data 2022-2023))
  3. Food Balance Sheets, Africa (data) · FAOSTAT (2025 (data 2010-2023))
  4. Food Security and Nutrition Indicators (data) · FAOSTAT (2026 (data to 2025))
  5. Production: Crops and Livestock, Africa (data) · FAOSTAT (2026 (data 2024))
  6. Inputs: Fertilizers by Nutrient (data) · FAOSTAT (2025 (data 2023))
  7. Production: Crops and Livestock, Africa (data), live animal stocks; cattle, goats, sheep and camels by member state · FAOSTAT (2026 (data 2024))
  8. Capture fishery production (FAO FishStat) · Our World in Data (2026 (data 2024))
  9. Aquaculture production (FAO FishStat) · Our World in Data (2026 (data 2024))
  10. Global Production dataset (GlobalProduction_2025.1.0), capture and aquaculture by species and area, 1950-2023 · FAO Fisheries and Aquaculture Division (2025 (data 2023))
  11. The State of World Fisheries and Aquaculture 2024: the status of fishery resources · FAO (2024 (2021 assessment))
  12. World Development Indicators: population, population growth, arable land per person, renewable internal freshwater, reserves, agriculture value added, GDP per capita, external debt service (% of exports) · World Bank (2026 (data 2022-2025))
  13. Production, Supply and Distribution database: grains and pulses · USDA Foreign Agricultural Service (2026 (MY2024/25 and MY2025/26))
  14. UN Comtrade: wheat, rice, maize, soybean meal, fishmeal, milk powder, nitrogenous fertiliser and refined petroleum imports by Kenya, Tanzania and Uganda, 2024 · United Nations Statistics Division (2026 (data 2024))
  15. UN Comtrade: live sheep and goat imports by Saudi Arabia and Oman, and Indian rice and soybean meal exports, 2024 (mirror statistics for non-reporting members) · United Nations Statistics Division (2026 (data 2024))
  16. UN Comtrade: Nile perch fillet exports by Uganda and Tanzania, 2024 · United Nations Statistics Division (2026 (data 2024))
  17. GIEWS Country Brief: Sudan, 27 May 2025 · FAO (2025)
  18. GIEWS Country Brief: Somalia, 11 March 2026 · FAO (2026)
  19. GIEWS Country Brief: South Sudan, 17 September 2025 · FAO (2025)
  20. GIEWS Country Brief: Kenya, 4 May 2026 · FAO (2026)
  21. GIEWS Country Brief: United Republic of Tanzania, 10 July 2026 · FAO (2026)
  22. GIEWS Country Brief: Uganda, 8 May 2026 · FAO (2026)
  23. GIEWS Country Brief: Rwanda, 12 November 2025 · FAO (2025)
  24. GIEWS Country Brief: Burundi, 5 February 2026 · FAO (2026)
  25. GIEWS Country Brief: Djibouti, 9 July 2025 · FAO (2025)
  26. GIEWS Country Brief: Eritrea, 3 September 2025 · FAO (2025)
  27. East Africa Food Security Outlook, June 2026 - January 2027 · FEWS NET (2026)
  28. Sudan Food Security Outlook, June 2026 - January 2027 (data as of 30 June 2026) · FEWS NET (2026)
  29. Kenya Food Security Outlook, June 2026 - January 2027 · FEWS NET (2026)
  30. Financial Tracking Service: 2026 humanitarian response plan funding for Sudan (plan 1514), South Sudan (1512) and Somalia (1516), as of 31 July 2026 · UN OCHA (2026)
  31. Sudan emergency · World Food Programme (2026)
  32. Somalia country page · World Food Programme (2026)
  33. Persian Gulf crisis impacting food security, FAO warns (statements by Maximo Torero, FAO Chief Economist, 26 March 2026) · UN News (2026)
  34. Large-scale land acquisition database: concluded deals in Sudan, South Sudan, Kenya, Tanzania, Uganda, Rwanda, Burundi and Eritrea (accessed July 2026) · Land Matrix (2026)
  35. Djibouti country overview · World Bank (2026)
  36. Inputs: Land Use (data), arable land and land area equipped for irrigation, including the 2000-2024 trend · FAOSTAT (2026 (data 2000-2024))
  37. Navigating troubled waters: impact to global trade of disruption of shipping routes in the Red Sea, Black Sea and Panama Canal · UNCTAD (2024)
  38. Official statement: Ethiopian Investment Holdings and Dangote Group sign shareholders' agreement for the USD 2.5 billion, 3 million tonne Gode urea complex (announced 28 August 2025; EIH 40 per cent, Dangote 60 per cent; completion targeted within 40 months of commencement) · Ethiopian Investment Holdings (2025)
  39. Dangote calls Ethiopia most attractive investment destination in Africa at Gode fertiliser plant groundbreaking (ceremony reported 3 October 2025; capacity 3 million tonnes urea a year) · Ethiopian News Agency (2025)
  40. South Sudan Food Security Outlook, June 2026 - January 2027 (data as of 24 June 2026): credible risk of Famine persists in Akobo, Nyirol, Nasir and Ulang · FEWS NET (2026)
  41. Cost and Affordability of a Healthy Diet (CoAHD), July 2026 update (SOFI report release): prevalence of unaffordability, number of people unable to afford a healthy diet, cost of a healthy diet · FAOSTAT (2026 (data 2017-2025))
  42. Sudan Market Monitor, April 2026 (data to March 2026): WFP Local Food Basket cost, staple food and livestock prices, exchange rates · World Food Programme (2026)
  43. Global Price Watch, June 2026 · FEWS NET (2026)
  44. Consumer Price Indices and Inflation Rates, June 2026 (and November 2025) · Kenya National Bureau of Statistics (2026)
  45. Investment: Credit to Agriculture (data), share of total credit and agriculture orientation index · FAOSTAT (2026 (data 1991-2024))
  46. World Development Indicators: population, 2000 baseline for per-capita land trend · World Bank (2026 (data 2000))
  47. Re-engineering of Galana Kulalu Food Security Project to Maximize its Potential · Kenya Institute for Public Policy Research and Analysis (KIPPRA) (2021)
  48. How irrigation is changing the once deserted land in Galana Kulalu · People Daily (2025)
  49. Satellite imagery reveals how Sudan's war scorched its 'breadbasket' · Al Jazeera (2026)
  50. The environmental costs of the war in Sudan · Conflict and Environment Observatory (CEOBS) (2025)

What we could not measure

Sudan, Somalia and Eritrea, about a quarter of the region's people, have no published affordability estimate, so the reported 76 per cent unaffordability figure is a floor that almost certainly understates the true picture, since Sudan and Somalia already carry the region's worst hunger and currency figures. FAOSTAT's protein supply data also excludes those two states plus Burundi and South Sudan, so the self-sufficiency reading rests on data now several years stale for over a third of the region.

Published August 2026 ·How scores are produced