Testing the Marshall-Lerner Condition: Cointegration in Models with Regime Shifts

Testing the Marshall-Lerner Condition:
Cointegration in Models with Regime Shifts

Fang Dong, Ph. D.Assistant Professor
Department of Economics
Providence College
Sullivan Hall 104
Providence College 549 River Avenue
Providence, RI 02918,

U.S.A.

Abstract

Many industrial countries face current account problems and the United States is not an exception. To improve current account balance in the short run, one would expect that the Marshall-Lerner condition holds. That is, all else equal, a real depreciation improves the current account if export and import volumes are sufficiently elastic with respect to the real exchange rate change. This paper tests this condition on the bilateral trade balance of the United States with Canada, France, Germany, Japan, Mexico and the United Kingdom. By assuming that there is one cointegration relation among real U.S. exports to a foreign country, the foreign country’s industrial production and the real exchange rate, we estimate Vector Error Correction Models, which produce income elasticity of U.S. exports, price elasticity of U.S. exports and the dynamic adjustments of the U.S. exports to the foreign country. Similarly, by assuming that there is one cointegration relation among real U.S. imports from a foreign country, the U.S. industrial production, and the real exchange rate, we estimate Vector Error Correction Models, which produce income elasticity of U.S. imports, price elasticity of U.S. imports and the dynamic adjustments of the U.S. imports from the foreign country. In general the Marshall-Lerner condition is only partially satisfied in the sub-sample periods. We then proceed to test the stability of the cointegrating relation by applying Gregory and Hansen (1996)’s residual-based tests for cointegration models with regime shifts. By taking into account of a possible structural shift, we hope to identify the episodes in which the Marshall-Lerner condition is satisfied and to analyze their implications for actual current account responses to real exchange rate changes. (JEL codes: F14, F15)