TY  - JA
AU  - Ndegwa,M.K.
AU  - Nyangau,P.N.
AU  - Kariuki,S.W.
AU  - Debello,M.J.
AU  - Michelson,H.C.
AU  - Cairns,J.E.
TI  - Is newer better? The effect of varietal age on real-world maize yield in Kenya
PY  - 2025///
CY  - Netherlands
PB  - Elsevier, 
KW  - Entropy
KW  - AGROVOC
KW  - Smallholders
KW  - Maize
KW  - Yield gap
KW  - Senses
KW  - Fertilizers
KW  - Kenya
N1  - Preprint; Open Access
N2  - Maize varietal turnover is widely promoted across Sub-Saharan Africa to improve crop productivity and increase food security, yet its impact on yields remains poorly understood amid heterogeneous agroecological and socioeconomic conditions. This study quantifies the yield effects of varietal age in Kenya using a three-wave panel survey (2023-2024) of 4,160 smallholder households across Kenya. Using entropy balancing and weighted regression models to isolate the effect of varietal age on maize yield, we find a strong and consistent relationship between varietal age and yield. New varieties yield 147 kg/ha more than old ones in the long rains and 91 kg/ha more in the short rains. Finer age categorization reveals that switching to ultra-new varieties (0-5 years) delivers the highest gains-360 kg/ha over ultra-old varieties (21+ years) in the long rains and 269 kg/ha in the short rains. These findings suggest that slow varietal turnover carries significant opportunity costs in the form of forgone yield gains. While farmers generally perceive new varieties favorably-particularly for yield potential, early maturity and grain quality-concerns around labor intensity and resilience remain, potentially dampening adoption. Providing farmers with clear, locally relevant performance data and opportunities for on-farm experimentation can help shift perceptions and support wider uptake. Policies and programs that expand access to newer, better-performing varieties and strengthen seed quality assurance are essential for translating genetic gains into productivity improvements across Kenya's bimodal maize
UR  - https://hdl.handle.net/10883/36067
DO  - http://dx.doi.org/10.2139/ssrn.5345569
T2  - Social Science Research Network (SSRN)
ER  -