Effect of input choices and use on productivity among maize farmers in Uganda

Abstract

Sustained improvements in maize productivity among smallholder farmers require not only access to modern inputs but also their correct use. Drawing on data from 2,243 households collected under Uganda’s World Bank-funded program, this study investigates (i) the factors influencing the adoption of different combinations of inputs (improved maize seed and inorganic fertilizer), (ii) the yield effects of these input-choice combinations; and (iii) the drivers of fertilizer misapplication and its effects on maize yield. A multinomial endogenous switching regression model showed that the likelihood of adopting only improved maize seed increased with access to credit, extension services, livestock ownership and drought experience while it decreased with larger land size. In contrast, the adoption of both improved maize seed and inorganic fertilizer was more likely among farmers with higher education levels, access to extension services, higher household incomes and larger land sizes. However, experiencing drought and having secure land tenure significantly reduced the likelihood of adopting both inputs. Access to education, engagement in farm employment, agronomic training, access to credit, market information and extension services, higher household incomes, larger land sizes and membership to a farmer organization significantly reduced the likelihood of adopting nonmodern maize inputs. However, the productivity analysis reveals that the use of modern inputs significantly improves maize yields, with the greatest gains achieved through the combined use of improved maize seed and inorganic fertilizer. Farmers who adopted both improved maize seed and inorganic fertilizer experienced a 185.5% increase in yield, with an average treatment effect on the treated (ATT) of 735 kg/ha. Those who adopted only improved maize seed or only inorganic fertilizer achieved yield gains of 138.5% (ATT = 698 kg/ha) and 137.8% (ATT = 359 kg/ha), respectively, compared to non-adopters. The effect of genetics appears to be greater than that of fertilizers, possibly due to the misapplication of fertilizers. However, with improved application practices, the impact of genetics is even likely to be greater, as the interaction between genetic traits and the environment can produce an additive effect that further enhances yield. Results from the multivariate probit model indicated that households with larger household sizes, older household heads, male household heads, access to extension services and secure land tenure were less likely to misapply fertilizer. In contrast, drought conditions, intercropping with legumes and higher household income increased the likelihood of fertilizer misapplication. Notably, land size and access to agronomic training showed mixed effects lowering the probability of certain types of misapplication while increasing others. Inverse Probability Weighted Regression Adjustment (IPWRA) estimation

results showed that fertilizer misapplication using wrong application methods, applying below the Abuja Declaration target of 50 kg/ha and below the recommended optimal rate of 125 kg/ha led to significant yield reductions of 21%, 28% and 35%, respectively. Moreover, combined forms of misapplication further intensified yield losses. Applying fertilizer only once using a wrong method led to a 21% reduction in yield. Similarly, applying fertilizer below the optimal rate of 125 kg/ha using a wrong method reduced yield by 24%. Applying fertilizer only once and below the Abuja target also resulted in a 28% yield reduction. The greatest yield loss of 34% occurred when fertilizer was applied below the Abuja Declaration target of 50 kg/ha using wrong methods. Notably, the results indicated that applying fertilizer within the recommended midrange (50 –125 kg/ha), meeting or exceeding the Abuja target significantly increased maize yield by 16%. This suggests that the Abuja Declaration target is appropriate, as yield increases were observed both at that level and when approaching the recommended optimal rate. These findings underscore the need for policy interventions that promote comprehensive input choices to harness complementarities, develop vibrant agro-input markets and expand agro- input dealer networks, and strengthen institutions such as farmer organizations and agricultural credit systems. In addition, there is a pressing need to enhance agricultural services, including access to credit, extension and insurance, which are crucial for guiding farmers in adopting and effectively using modern inputs. Improving farmer knowledge particularly on the prudent and efficient use of fertilizers should be central to agronomic training and advisory efforts. These services are essential to ensure that inputs, especially fertilizers, are not only adopted but also applied correctly to maximize their effectiveness. Ultimately, investing in institutional support, promoting integrated input use and improving the efficiency of input application will enhance maize productivity and help close persistent yield gaps among smallholder farmers in Uganda.