Heterogeneity Effects of Improved Tomato Farmers' Technical Efficiency in Southwest, Nigeria: An Unconditional Quantile Regression Approach.

Abstract Nigeria's present economic situation has raised production costs and directly affected the affordability of farm inputs, endangering both the sustainability of output and the improvement of farmers' efficiency. Furthermore, the consequences of climate change are becoming more noticeable, and one of the industry’s most at risk is agriculture. Therefore, to alleviate all of these problems, improved tomato varieties and sustainable farming techniques must be developed to boost farmers' output. Therefore, the study aimed to measure the efficiency of the farmers and to examine the effects of improved tomato varieties cultivated and sustainable agricultural practices on the efficiency of the farmers. The study was carried out in Osun and Ekiti States based on the density of improved tomato farmers. Data Envelopment Analysis (DEA) model was used to measure the technical efficiency of the farmers while Unconditional Quantile Regression (UQR)was used to determine the heterogenous effects of SAP and variety cultivated on the efficiency of the farmers. DEA model employed to assess the technical efficiency of tomato growers revealed that the Variable Return to Scale Technical Efficiency (VRSTE) output of the tomato farmers in the examined area achieved an average technical efficiency of 0.93, indicating a higher level of technical efficiency among them. The assessment of efficiency is based on analyzing the input amounts used by the tomato farmers about the volume of tomatoes produced and it became apparent that the farmers applied a greater quantity of fertilizer (mean slack of 1.64kg) to provide essential nutrients to the crops and enhance their yield. The model also indicated a mean slack of approximately 0.13ha of land among the farmers in their agricultural activities. The UQR findings revealed that the factors at various quantile levels suggested that there is heterogeneity among the quantiles, as the strength of the coefficients varied as well. Soil quality and income level had a significant impact across all quantiles. Family size, years of experience, land ownership, planting period, income level, and access to credit displayed negative coefficients across the quantiles, while age, marital status, gender, education level, access to extension agents, number of sustainable agricultural practices, and number of varieties grown presented positive coefficients across the quantiles. The implementation of SAP exerted a negative influence on efficiency at the τ.50 quantile, where it was not statistically significant. This suggested that a one-unit rise in the number of SAP adopted at the τ.50 quantile would lead to a 0.82% decrease in farmer efficiency. Furthermore, an increase of one unit in SAP adoption would result in efficiency improvements of 0.001% and 0.002% for farmers at the τ.25 and τ.75 quantiles, respectively, indicating a minimal effect on the efficiency of the local tomato farmers. The number of hybrid tomato varieties cultivated positively influenced efficiency at all quantile levels, and it reached statistical significance at all quantile levels. This indicated that for every additional unit of varieties grown, farmer efficiency is expected to rise by roughly 0.08%, 0.03%, and 0.05% at the τ.25, τ.50, and τ.75 quantiles, respectively. Keywords: Improved Tomato Varieties; Unconditional Quantile Regression (UQR); Efficiency; Data Envelopment Analysis (DEA); Heterogeneity effects; Sustainable Agricultural Practices (SAP).