Quantile regression analysis of extension program heterogeneous effects on farm income risk management and portfolio optimization

Abstract

This study employed quantile regression analysis to examine the heterogeneous effects of agricultural extension programs on farm income risk management and portfolio optimization across different income quantiles. Using cross sectional data from 480 smallholder farmers in Northern Ghana collected between January 2023 and December 2023, 
the research applied quantile regression models at the 25th, 50th, and 75th percentiles to analyze differential impacts. Results revealed significant heterogeneity in the effects of the extension program, with the highest-income farmers (75th percentile) experiencing a 34.2% reduction in income variance, compared to 18.7% for the lowest quantile. Technical efficiency scores ranged from 0.612 at the 25th percentile to 0.784 at the 75th percentile, indicating substantial efficiency gains in the upper quantiles. Crop diversification index showed progressive improvement across 
quantiles (0.347, 0.521, and 0.679, respectively), while risk management capacity increased from 2.84 to 4.23 on a 5 point scale. The study confirms that extension programs exhibit diminishing marginal returns in lower-income quantiles but accelerating benefits for higher-income farmers, suggesting the need for targeted intervention strategies. These findings contribute to optimizing extension service delivery through quantile-specific approaches that address heterogeneous farmer characteristics and resource endowments.