Technical efficiency in agriculture: A decade-long meta-analysis of global research

Poverty and food insecurity remain a glaring reality, particularly in regions linked to dryland agriculture. Low agricultural productivity due to inefficient input use reduces the prospects of improving the circumstances many rural households face due to these challenges. This highlights the need to understand factors affecting efficiency indices from several published studies, including shedding light on whether efficiency has increased over time. In this comprehensive meta-analysis spanning a decade (2014-2023), we shed new light on the evolving landscape of global technical efficiency in the agricultural sector, assuming that the institutional environment has changed over time. The literature search yielded 289 published articles and 549 observations for analysis. The beta regression meta-analysis results indicate that the translog functional form leads to higher mean technical efficiency values than the cobb-Douglas and other functional forms, while panel data studies yield lower mean technical efficiency values than cross-sectional studies. This emphasizes the need to select the most suitable functional form based on the data and model specification tests. Measures should be taken to eliminate selection bias in efficiency-related cross-sectional studies. Technical efficiency in agricultural enterprises globally increases with publication year with an average mean of 0.72. Over the past decade, the major efficiency drivers include farm size, extension support, credit access, and education, influencing efficiency positively, while subsidies have had the greatest negative effect on efficiency. Policies that enhance extension support, credit access, participation of the educated in commercial farming and scaling down of subsidy programs should be strengthened to improve global agricultural productivity.