| Abstract: | One of the important trade effects on growth is technology diffusion through learning by doing. Chuang [1998] proposed a trade-induced learning theory in which the nature of traded goods and the trading partners are two key factors determining the effectiveness of the trade-induced learning. The former conveys the characteristics that a country can learn; the latter determines the level of technology from which a country can learn. Using cross-country data, this article constructs a set of the trade-induced learning variables by taking into account trading partners and the characteristics of the traded goods and further tests the trade-induced learning hypothesis. The results show that holding other variables constant, trade-induced learning has a positive and significant effect on growth and the estimated effect implies that a one-standard-deviation increase in the trade-induced learning variable is estimated to generate an effect of between 0.4 to 1.0 percentage points on the annual growth rate. Robustness test shows that the trade-induced learning variable passes the extreme-bound analysis and also outperforms other conventional trade variables advocated in the literature. |
