Small seed packs, big potential? Effect of seed packs on knowledge and adoption of improved crop varieties : technical report
Mulungu, K.H.
Small seed packs, big potential? Effect of seed packs on knowledge and adoption of improved crop varieties : technical report - [Place of publication not identified] : CIMMYT, 2025. - 28 pages
Open Access
Small seed packs are widely distributed to promote improved crop varieties, yet their effectiveness remains contested. This study evaluates how small seed packs influence knowledge and adoption of improved legumes, cereals, and vegetables among smallholder farmers in Tanzania and Zambia. Using mobile phone survey data from 1,577 respondents (840 in Tanzania, 737 in Zambia) and entropy balancing and machine learning to address self-selection, we uncover evidence of seed pack effectiveness across different crops. Seed packs significantly increased farmers technical knowledge scores for biofortified iron beans, cowpeas and improved traditional African vegetables by in both countries. Notably, while seed packs showed minimal impact on drought-tolerant maize (DTM) knowledge—reflecting high baseline awareness—they still significantly increased DTM adoption, revealing distinct knowledge-transfer and experiential-learning pathways through which seed packs influence farmer behavior. Adoption impacts were substantial across all crops: seed packs increased adoption likelihood by 10-16 pp for legumes and 9-13 pp for cereals in Tanzania, with slightly smaller but significant effects for vegetables (5-9 pp). Similar adoption gains were observed in Zambia (9-11 pp for legumes, and about 12 pp for DTM). These consistent positive effects across different crops and contexts, robust to alternative estimation approaches, confirm that small seed packs represent a promising demand creation mechanism for improved germplasm, with particularly strong impacts for crops where farmers have limited prior exposure, but persistent adoption benefits even for familiar technologies but with low baseline adoption.
Text in English
Smallholders
Demand
Improved germplasm
Machine learning
Entropy
United Republic of Tanzania
Zambia
Small seed packs, big potential? Effect of seed packs on knowledge and adoption of improved crop varieties : technical report - [Place of publication not identified] : CIMMYT, 2025. - 28 pages
Open Access
Small seed packs are widely distributed to promote improved crop varieties, yet their effectiveness remains contested. This study evaluates how small seed packs influence knowledge and adoption of improved legumes, cereals, and vegetables among smallholder farmers in Tanzania and Zambia. Using mobile phone survey data from 1,577 respondents (840 in Tanzania, 737 in Zambia) and entropy balancing and machine learning to address self-selection, we uncover evidence of seed pack effectiveness across different crops. Seed packs significantly increased farmers technical knowledge scores for biofortified iron beans, cowpeas and improved traditional African vegetables by in both countries. Notably, while seed packs showed minimal impact on drought-tolerant maize (DTM) knowledge—reflecting high baseline awareness—they still significantly increased DTM adoption, revealing distinct knowledge-transfer and experiential-learning pathways through which seed packs influence farmer behavior. Adoption impacts were substantial across all crops: seed packs increased adoption likelihood by 10-16 pp for legumes and 9-13 pp for cereals in Tanzania, with slightly smaller but significant effects for vegetables (5-9 pp). Similar adoption gains were observed in Zambia (9-11 pp for legumes, and about 12 pp for DTM). These consistent positive effects across different crops and contexts, robust to alternative estimation approaches, confirm that small seed packs represent a promising demand creation mechanism for improved germplasm, with particularly strong impacts for crops where farmers have limited prior exposure, but persistent adoption benefits even for familiar technologies but with low baseline adoption.
Text in English
Smallholders
Demand
Improved germplasm
Machine learning
Entropy
United Republic of Tanzania
Zambia