Impact of agricultural extension services on the technical efficiency of adopters and non-adopters of genetically modified maize in the Eastern Cape province.

Abstract

Maize (Zea Mays) is one of the most widely grown cereals and is consumed worldwide. In South Africa, maize is grown for various purposes, which include animal feed such as silage and grains, poultry and pig feed, and human consumption, as grains, sweet maize, and grain maize. Despite its importance, maize productivity remains low due to poor seed quality, pest and disease susceptibility, climate variability, limited access to high-quality inputs, inequitable allocation of resources, inadequate extension services, and suboptimal farm management. Genetically modified (GM) maize seeds have been developed to improve production, resilience, and overall farm performance; however, their adoption remains inconsistent. Smallholder farmers often prefer the use of traditional seed varieties, which tend to perform poorly under the harsh conditions of the Eastern Cape. Hence, this study evaluated the impact of agricultural extension services on the technical efficiency of adopters and non-adopters of genetically modified maize across Chris Hani District and OR Tambo District Municipality using a quantitative approach. A multistage sampling method was used to collect data from 306 maize farmers using a semi-structured questionnaire. The descriptive statistics of the overall sample revealed a predominance of female participants (56.9%), with an average age of 61 years, 7 years spent at school, managing about 8 hectares, and an average income of ZAR 6637.52. The average farm size reported in this study reflects the total land that farmers had access to for cultivation rather than land ownership. This includes land that was borrowed or leased from relatives and other landowners, as well as larger landholdings accessed by beneficiaries of the government's land reform programme. Only 6.2% were formerly employed, 19.3% self-employed, 34.0% unemployed, and 40.5% retired, with an average household size of 9.14. The probit model indicated several significant variables that affected GM adoption, such as gender, years spent in school, marital status, household size, household monthly income, land size, access to extension services, and amount of credits, which positively influenced the adoption of GM maize seeds by farmers, while distance to markets was negatively associated with the adoption of GM seeds. The truncated regression model revealed that household size, household income, land size, and amount of credits positively influenced adoption intensity; distance to markets was negatively associated with the intensity of adoption of GM maize seeds by farmers. Using the stochastic frontier analysis, Meta-Frontier, and Endogenous Switching, the study further evaluated the productivity effects of agricultural extension on smallholder maize efficiency. The findings revealed that agricultural extension services significantly enhanced maize productivity, with an average yield gain of 0.81 tonnes per hectare. Farmers with access to extension operated closer to the production frontier and exhibited higher technical efficiency than non-participants, with a mean efficiency level of 77%, indicating room for improvement. The study further delved into unpacking Smallholder Maize Farmers’ Preferences for Genetically Modified Maize Seeds in the study area. The findings revealed that farmers’ seed choices included GM (HT/BT) at 38.6%, GM (BT) at 26.8%, and traditional maize at 35.0%. While Multinomial Regression Logistic results revealed that household age, education, income, years of maize farming, and organizational membership positively influenced both GM (HT/BT) and GM (BT) selection. In contrast, extension services and credit access influenced GM (HT/BT) adoption only, and this might be because it offers combined herbicide tolerance and insect resistance, making it a more comprehensive technology package. Furthermore, the study looked at the productivity-profitability trade-offs and the welfare effects of GM Maize in the study area. The propensity score matching revealed that adopters consistently achieved higher maize yields compared to non-adopters of GM technology, with increases of 2.1 tons/ha (Nearest Neighbor) and 2.25 tons/ha (Kernel). Lastly, the study evaluated the adoption of genetically modified maize seed on input elasticities and productivity among Smallholder Farming Systems. The Cobb-Douglas Stochastic Production Frontier (SPF) model revealed that SPF results indicated that adopters revealed significantly higher input elasticities for seed (0.40, p < 0.01) and fertiliser (0.20, p < 0.01), while non-adopters showed limited responsiveness to modern inputs. Technical efficiency was markedly higher among adopters (TE = 0.75) than among non-adopters (TE = 0.60), with adopters operating under constant returns to scale and non-adopters under decreasing returns to scale. These findings demonstrate that GM maize adoption enhances input efficiency, boosts productivity, and provides a pathway for sustainable intensification of smallholder maize systems. Based on these key findings, the study recommends that strengthening agricultural extension systems through mobile-based platforms, tailored knowledge dissemination, and inclusive targeting of resource-constrained farmers can significantly narrow efficiency gaps and promote sustainable intensification. Furthermore, the study emphasises introducing context- and variety-specific policies, strengthening extension services and farm organizations, engaging experienced farmers as lead farmers, and developing subsidies and microcredit schemes to enhance the choice of GM maize seed selection by farmers. Lastly, the results provide policymakers and agricultural extension personnel with evidence to design effective input support, extension, and technology dissemination programs. The results provide policymakers and agricultural extension personnel with evidence to design effective input support, improve access to inputs, strengthen extension services, and enhance technology dissemination programs. This should involve the inclusion of public-private partnerships (PPP) models to address the high extensionto-farmer ratios and issue of funding, whereby government collaborates with private seed companies, non-governmental organisations, research institutions, and farmer organisations to co-finance and deliver extension services.