Productivity & Yields

Key evidence

• Coffee: yields per tree rose in the endlines and midlines that measured them, e.g. 1.73 to 3.73 kg cherry/tree in TRACE Kenya (self-reported), 2.3 to 3.8 kg/tree for Ugandan small farms in the Coffee Resilience programme, and 4.49 to 10.73 quintals per household in OCDDP; Tanzania (2.5 kg) and PACE (3.4 kg) stayed below targets of 4–5 kg/tree. TRACE Kenya · Endline · 2025PFC · Endline · 2022OCDDP · Endline · 2021PACE · Midline
• Dairy: the Kenyan FOSEK pilot with milk coolers raised milk delivered per farmer from 3 to 10 litres a day within a year, but in F4APK dairy farmers gained only about 0.75 litres per cow per day from 8.5 litres, no different from control farmers; CROSL farmers in Ethiopia started at 2.9 litres per cow per day, with 25.1% owning improved breeds. FOSEK · Endline · 2021F4APK · Endline · 2020CROSL · Baseline · 2028
• Maize: Bungoma farmers started at 10.3 bags (90 kg) per acre and produced at a loss despite 95% using fertiliser; by Year 2 all CSV respondents had adopted at least one technology (84% certified seed) and 87% expected higher production, though harvests were not yet in. In Acting Now Kenya, 63% of treatment maize farmers reported lower yields, mainly from drought. CSV · BaselineCSV Maize · Midline · 2024AN · Baseline · 2025
• Horticulture: H4G farmers widely used improved seed (72.2%) and fertiliser (70%), but the share growing the project's target crops fell (e.g. beans 49% to 21.2%) and per-acre costs rose sharply; F4APK found no positive effect on horticulture yields except French beans and bananas before COVID-19. Avocado yields in the Tanzanian Southern Highlands ranged from 2,675 kg (Mbeya) to 180 kg (Songwe) per household. Horticulture for Growth (H4G) – Promoting Horticulture Enterprise Development in Kenya · EndlineF4APK · Endline · 2020C&B · Baseline
• Tea: yields were 1,968 kg/ha in Kenya and 1,509 kg/ha in Uganda in 2021 against a potential of 4,745 kg/ha; Ugandan out-growers averaged 1,464 kg of made tea per hectare against up to 4,000 kg on estates. RS! · Thematic studyRS! · Baseline · 2025
• Oil palm: Kalangala growers averaged 6.98 tons per acre per year, ranging from 2.0 to 16.0 tons, a wide yield gap among growers. NOPP · Baseline
• Livestock (beef): the Loita project distributed 5 Sahiwal bulls, 20 rams and 20 bucks for breeding, which evaluators judged low for the expected impact; 82% of respondents had access to veterinary services. Promotion of Sustainable Beef Production through Development of Field to Fork Value Chain Approach · Endline · 2015
• Cotton: 2,608 smallholders adopted better practices in line with CmiA (104% of target), and a low-cost IPM method is now replicated across Ethiopia. Bottom Up · Endline · 2023
• Gold: artisanal methods recover only about 25–40% of the gold, leaving the rest in tailings that could yield up to 90% with better processing. RS! · Thematic study
• Leather: existing MSMEs given business training and improved machines reported production increases of 50% to 300%; Addis Ababa Abattoirs reported skin defects falling from 15,564 kg to 7,794 kg after training. Leather Initiative for Sustainable Employment and Job Creation (LISEC) Green Jobs Project · MidlineGTI · Endline · 2021

In figures

Change over time

Recommendations for programming

• Build seedling and breeding systems run by cooperatives, lead farmers or groups: coffee nurseries, and decentralised community breeding with more bulls, bucks and rams. OCDDP · Endline · 2021P2P · BaselinePromotion of Sustainable Beef Production through Development of Field to Fork Value Chain Approach · Endline · 2015
• Expand soil-testing capacity and water management (irrigation, rainwater harvesting, water storage for livestock). CSV Maize · Midline · 2024PACE · MidlineHorticulture for Growth (H4G) – Promoting Horticulture Enterprise Development in Kenya · EndlineCROSL · Baseline · 2028
• Support investment in modern processing technology for artisanal gold and train miners to test and process tailings. P2P · BaselineRS! · Thematic study
• Close tea yield gaps through value-chain upgrading and attention to large regional productivity differences. RS! · Thematic study
• Make credible counterfactual designs part of evaluations so productivity effects can be attributed. F4APK · Endline · 2020

Evidence gaps & caveats

• Only one study (F4APK) used a control group; other changes are before–after comparisons, mostly self-reported.
• Each commodity uses its own unit, so results across commodities cannot be added up or ranked.
• Tea, oil palm, cotton, gold and livestock have baseline or output data only; no endline yet measures their yield change.