Regression coefficient estimates of exchange rate total effects on aggregate demand are broken into separate income and substitution effects. Total effect (substitution and income) regression estimates can seem contrary to theory and common sense. Separating them into their two components shows this is not the case. The separation method also provides a simple test to determine if imports are normal or inferior goods. The paper finds consumer imports are normal goods, investment imports are inferior goods. The paper shows that if import total effects exceed domestic total effects, imports are a normal good. If smaller, they are inferior goods.
John J. Heim. "A Method for Separating Income and Substitution Effects of Exchange Rate Changes." Proceedings of the New York State Economics Association. vol. 4, September 2011, p. 80-87
BibTeX entry download