¿Does the level of development matter in determining the marginal propensity to import? An empirical comparison between Ecuador, Chile and the United States

Abstract

The objective of our investigation is to determine the incidence of income and the real exchange rate on imports from Ecuador, Chile and the United States during the period 1980-2014. For this we rely on the Blanchard model (2012) and use time series data from the World Bank (2017). The results show that the study we do partially agrees with the theory based on the fact that it is true for the relationship between income and imports in the three countries. The positive relationship of the real exchange rate with imports only occurs in the United States, in addition to being significant for Ecuador and Chile. In the case of the United States we introduced a dichotomous variable, and applied a distributed lag of the first order, to correct heteroscedasticity and serial autocorrelation. For a better explanation of imports we include a set of control variables: inflation, unemployment, total income of natural resources and tariffs. In the case of Ecuador, the significant variables are income and the real exchange rate, while for Chile, income, real exchange rate, inflation are significant; And for the United States income, inflation and unemployment. For the United States model we include a dichotomous variable to correct the serial autocorrelation. In general, it does not matter the level of development in determining the marginal propensity to import. Policies restricting imports and promoting domestic industry can reduce the high marginal propensity to import.