Abstract
This paper identifies export instability as an additional cause of changes in GDP. Instability is measured by deviations of exports from their trend values. A multivariate error-correction model that incorporates export instability is tested on individual time series data (1967–1990) for twenty-one African countries. The results indicate reduction in export-earnings but not any significant decline in GDP growth due to instability. Half of these countries experienced export-to-growth causality, with a reverse causality in three-fourths of them. The results are sensitive to both the omission of variables and the test procedure used.