A meta-analysis of the impact of agricultural extension services

Abstract

PurposeThis study aims to address two research questions. First, do the agricultural extension services have an impact on the potential outcomes considered in the primary studies, and to what extent? Second, how sensitive is the reported impact to the study-specific characteristics in the primary studies?Design/methodology/approachThe paper synthesizes 45 studies that assessed the causal impact of agricultural extension services published in 2004–2021, using meta-regression analysis. It considers three measures of effect sizes – Cohen’s, Hedges and principal correlation coefficient (PCC) – to standardize the reported impact of agricultural extension services in the primary studies.FindingsThe empirical results show that, on average, agricultural extension services have statistically significant and positive impacts on the potential outcomes identified in the primary studies. However, the magnitude of the impact is considered medium-sized. Other results show that the effect size estimates of agricultural extension services' impact significantly vary with the data type (cross-sectional data vs. panel data), research design (non-experimental vs. experimental design) and econometric methods employed in the primary studies.Practical implicationsOne can argue that the medium-sized impact we estimated indicates evidence of a moderate, weak relationship between agricultural extension services and the potential outcomes considered in the primary studies. This means that agricultural extension services need to be restructured in the current form to stimulate change in the agricultural sector globally. In addition, the sensitivity of effect sizes to study attributes (i.e. data types, research design and econometric methods) shows that researchers and academicians need to pay attention to these attributes to provide more reliable estimates for policy purposes.Originality/valueThis is the first study that attempts to shed light on the overall performance of agricultural extension services using a meta-regression analysis approach.