An Aggregate Food Import Demand Function for the Caribbean a Cointegration Analysis

This article used cointegration analysis to examine aggregate food import demand in the Caribbean region between 1979 and 2018. The article examined the long-run and short-run relationship between food imports, relative prices, real per capita income, tourist numbers, export revenue from merchandise, and domestic food production using the autoregressive distributed lag model and bounds testing approach cointegration. The augmented Dicky-Fuller test confirmed that the variables used in the analysis were non-stationary, implying a possible long-run relationship between them. Once cointegration was established, an error-correction model was specified to separate food import demand's long-run and short-run dynamics. Real per capita income, tourist visits, and domestic food production were significant determinants of food import demand in both time horizons, while relative prices had little impact on food imports. Imported food was found to be price inelastic and a necessity for the study period. The adjustment parameter revealed that about 97.5% of short- to long-run disequilibrium is corrected yearly. Toda and Yamamoto test for causality was also used to assess causality between food imports and its determinants. It was found that mostly unidirectional causal relationships existed. Furthermore, it was determined that food imports were not the cause of domestic food production decrease within the Caribbean region over the study period. Finally, the paper also suggested some policy actions that can help to foster food and nutrition security within the region and reduce the ever-growing food import bill