Estimation of Maize Import Demand Function in Nepal: An ARDL-ECM Approach

Abstract

This study examined the long-run and short-run determinants of maize import behavior in Nepal, where maize is the second most imported cereal after rice. An error correction version of Autoregressive Distributed Lag (ARDL) model was employed using 28 years of time series data (1997-2024) collected from secondary sources. The stationarity of variables was tested using Augmented Dickey-Fuller (ADF) test, Phillips-Perron (PP) and Kwiatkoski-Phillips-Schmidt- Shin (KPSS) tests, confirming integration at either I (0) or I (1). The Bounds test verified the existence of long-run relationships among the variables (F=11.18, p<0.01). The error correction term (ECTt-1=-1.038) was negative and significant (p=0.0012), indicating rapid adjustment toward long-run equilibrium. The results revealed that remittance inflow (p<0.1), import price of substitute commodity rice (p<0.05) and consumer price inflation (p<0.01) significantly influenced maize imports in long-run, all exhibiting positive relationships. In the short-run, remittance inflow (p<0.01), price of substitute goods (p<0.1) and consumer price inflation (p<0.05) affected maize imports. However, unlike the long-run effect, increased remittance inflows were associated with a decrease in maize imports in the short-run, suggesting a temporary shift in consumer preference towards other commodities. Diagnostic tests confirm normality (Jerque-Bera Statistic X2=0.346, p>0.05 & Shapiro-wilk statistic w=0.975, p>0.05), absence of autocorrelation (Breusch-Godfrey LM=3.378, p>0.05) and heteroskedasticity (Studentized Bruesch-pagan BP= 17.198, p>0.05), and no model misspecification (D’Agostino skewness test statistic=2.267, z=0.658, p>0.05); thereby validating the robustness and reliability of model. The findings emphasize strengthening domestic maize production, promoting productive use of remittances, and maintaining stable inflation to reduce maize import dependency for policy implications.