Asymmetric effects of drought on wheat yield in North Africa: evidence from panel ARDL and NARDL models

Purpose This study shows that drought exerts asymmetric and temporally differentiated effects on wheat yields in North Africa. Focusing on Algeria, Egypt, Libya, Mauritania, Morocco, and Tunisia, it evaluates short- and long-run drought impacts over the period 1961–2021 in a context of rising climate variability and food security risks. Design/methodology/approach Panel Autoregressive Distributed Lag (ARDL) and Panel Nonlinear ARDL (NARDL) models are employed to capture dynamic adjustment processes and asymmetric yield responses to drought, distinguishing short-run shocks from long-run equilibrium effects. Findings Drought significantly reduces wheat yields in the short run. Nonlinear estimates reveal asymmetric long-run effects, with wetter-than-normal conditions generating persistent yield gains, weaker long-run losses from negative drought shocks, substantial cross-country heterogeneity, and sizable revenue losses under extreme drought. Research limitations/implications The use of aggregate national data may mask sub-national variation in drought exposure and resilience, motivating future research using higher-resolution data. Originality/value The study provides new evidence on asymmetric drought–yield dynamics and associated economic losses in North Africa, offering policy-relevant insights for climate-resilient agricultural planning.