Traditional comparative advantage theory employs static analysis to explore the relationship between international trade and a nation’s factor endowment. The dynamic changes in comparative advantage cannot, however, be explained through static analysis, and hence it is difficult to use the theory of comparative advantage as a forecasting tool. By inventing the stage approach, Professor Bela Balassa has devised a way to solve this problem. Basically, in the stage approach, each country is treated as being located at a certain point on a line of continuous comparative advantage movement. This analytical method allows for the prediction of each country’s future comparative advantage based on its own factor endowment growth; the comparative advantages of individual industries within a country can also be predicted.
The stage approach calls for two procedural stages of regression. First, for each country, the relative comparative advantage (RCA) of manufacturing industries is regressed on industrial characteristics. The estimated regression coefficient of industrial characteristics is regarded as an indicator of comparative advantage for each country. Second, this indicator of comparative advantage is used as a new dependent variable and regressed on the country’s factor endowment. Through these two procedures, commodity space can be converted into country space, and then this result can be used to validate the Heckscher-Ohlin (H-O) theory.
Using 1976 and 1981 data on industrial characteristics in Taiwan, this study employs Balassa’s stage approach to test the H-O theory and to predict future comparative advantages of industries in Taiwan.
Form the first procedural stage of the stage approach, we find that:
(1) After using stock measurement to calculate capital intensity, our results in the first stage are different from the findings in Balassa’s study, which uses US data on industrial characteristics, but are similar to Kim’s study of the Korean case.
(2) Human-capital intensity seems to be a better explanatory variable of the overall trade pattern than physical-capital intensity. This result is again similar to both Balassa’s and Kim’s studies. The result might also reflect that physical capital is more mobile than human capital.
The results from the second stage support the H-O theory and indicate that, following the development of both physical-capital and human-capital endowments, the pattern of comparative advantage tends gradually to move in a more capital-intensive direction.
The paper also compares the predictive ability of the stage approach with that of the “one-pass” method. We find that the stage approach is more powerful and more consistent in predicting future comparative advantage.
According to the empirical results presented in this study, assuming a constant factor-endowment growth rate, Taiwan’s export pattern (based on comparative advantage) after five years should be similar to Ireland’s pattern in 1981, and after fifteen years it should be similar to the 1981 US pattern.
