Measuring the factor content of trade

by 1976- May, Sharon Lee

Can the Heckscher-Ohlin-Vanek Theorem explain observed commodity trade? My conclusion is that it can. When factors and factor contents are measured correctly, we see a relationship between endowments and factor content flows. I describe the appropriate way to use OECD input-output tables, bilateral trade data, and national income data for nine member countries to calculate the measured and predicted factor contents of trade under both the traditional country-by-country approach and the new integrated method. A simple relationship between the observed national production techniques does not exist. Letting the production data speak for itself, the factor content of output matches well with the national factor endowments. Under the traditional method, I show that neither adjusting for not-traded goods nor allowing for quasi-homothetic preferences significantly improves the performance of the theory. I recognize that factor services are not homogeneous across countries. The Moore-Penrose inverse is used to construct a logical conversion matrix that translates foreign factors into their domestic equivalents. A new measured factor content of trade based on this aggregation method first calculates the factor content of goods where they were produced then converts those factors into domestic ones before aggregating across countries. With this translation, the performance of the theory in the trade test improves. ii Using the new integrated approach, I compare the assumption of perfect homogeneity with its opposite, complete heterogeneity of factor services. While the summary results for homogeneous factors look appealing, the case for country-specific factors is compelling. Identifying factors by country results in cases of both too much and too little factor content of trade. Among the EU members, factor content exchange is predicted well. This suggests that as barriers to trade are reduced and globalization homogenizes tastes, the importance of endowments in explaining trade patterns will increase. In the meantime, proper incorporation of imported intermediate inputs and heterogeneous factors are important clues to solving the mystery of the missing trade. iii