Environmental performance and cereal production in Ethiopia: evidence from an ARDL bounds testing approach
摘要
Climate change and environmental degradation pose significant challenges to agrarian economies, yet the relationship between aggregate environmental performance and staple crop production remains insufficiently quantified. This study investigates the association between environmental performance, measured by the Environmental Performance Index (EPI), and cereal production in Ethiopia over the period 2010–2023 (non-interpolated data) and 2002–2023 (with interpolation for robustness) using an Autoregressive Distributed Lag (ARDL) bounds testing approach. Arable land, fertilizer consumption per hectare, agricultural labor, and annual precipitation are included as control variables. The long-run results indicate that arable land has a positive and statistically significant association with cereal production, while fertilizer consumption exhibits a negative and significant long-run association, consistent with soil degradation dynamics. In the short run, fertilizer use positively associates with production, whereas environmental performance has a negative and significant association, suggesting short-run adjustment costs. The long-run EPI coefficient is statistically insignificant, indicating that no systematic long-run trade-off is detectable in the data. The error correction term (−0.51) implies that 51% of disequilibrium is corrected annually. Diagnostic and stability tests confirm model adequacy for the full sample, while robustness checks using a restricted non-interpolated sample, a non-agricultural EPI sub-index, and inclusion of precipitation support the main findings. The results suggest that environmental policies may impose short-run production costs but do not persistently hinder long-run agricultural performance. Policies should link fertilizer subsidies to lime application on acidic soils and provide temporary compensation for farmers affected by land set-asides.