Discussion papers // Ibero America Institute for Economic Research 112
This study examines the export-led growth hypothesis using annual time series data from Chile. It addresses the problem of specification bias under which previous studies have suffered and focuses on the impact of manufactured and primary exports on the economic growth. In order to investigate if and how manufactured and mining exports affect economic growth via increases in productivity, the study uses the Toda and Yamamoto (1995) procedure for testing for Granger non-causality in Vector Autoregressive models that involve variables that are integrated of an arbitrary order and that are possibly cointegrated. The estimation results support the export-led growth hypothesis for Chile and at the same time point out to the differentiated impact of manufactured and primary exports on the economic growth.