This paper challenges the common view that exports generally contribute more to GDP growth than a pure change in export volume, as the export-led growth hypothesis predicts. Applying panel cointegration techniques to a production function with non-export GDP as the dependent variable, we find for a sample of 45 developing countries that: (i) exports have a positive short-run effect on non-export GDP and vice versa (short-run bidirectional causality), (ii) the long-run effect of exports on nonexport output, however, is negative on average, but (iii) there are large differences in the long-run effect of exports on non-export GDP across countries. Cross-sectional regressions indicate that these cross-country differences in the long-run effect of exports on non-export GDP are significantly negatively related to cross-country differences in primary export dependence and business and labor market regulation. In contrast, there is no significant association between the growth effect of exports and the capacity of a country to absorb new knowledge.
Export-led growth Developing countries Panel cointegration