ABSTRACT
This paper attempts to find out whether, and which, strategic patterns in international business—as represented by two strategies: product differentiation and product complementarity—drive countries' bilateral trade flows. Using a modified version of the gravity model along with features of Porter's National Competitiveness theory (1990), this paper investigates which of the two major trade theories—inter-industry or intra-industry—prevails in the different regions of the world during the 1970s, 1980s, and 1990s. Results show that trade transactions were most statistically influenced by product complementarity in the 1970s, but then switched to product differentiation in the 1980s and 1990s, most probably due to the increasing role of globalization. Most regions experienced a product differentiation pattern over all three decades, except for South Asia, Africa, East Asia, and the Middle East (the latter three only in the 1970s), which show a pattern of product complementarity. The estimation of certain pertinent regional blocs reveals that EU, ASEAN, SADC, and APEC seem to create trade, while ECOWAS, GCC, SAARC, and MERCOSUR seem to have insignificant results. Interesting policy implications are discussed.
Notes
Note: Sample size is 80 individual countries selected from six geographic regions. Standard errors are presented in parentheses. Coefficients are estimated using White's procedure (Citation1980). The dependent variable is “log imports.” ∗denotes less than 10% level of significance; ∗∗denotes less than 5% level of significance; and ∗∗∗denotes less than 1% level of significance.
Note: Coefficients are estimated using White's procedure (Citation1980). The dependent variable is “log imports.” ∗denotes less than 10% level of significance; ∗∗denotes less than 5% level of significance; and ∗∗∗denotes less than 1% level of significance.
Note: Coefficients are estimated using White's procedure (Citation1980). The dependent variable is “log imports.” ∗denotes less than 10% level of significance; ∗∗denotes less than 5% level of significance; and ∗∗∗denotes less than 1% level of significance.
Note: Coefficients are estimated using White's procedure (Citation1980). The dependent variable is “log imports.” ∗denotes less than 10% level of significance; ∗∗denotes less than 5% level of significance; and ∗∗∗denotes less than 1% level of significance.
Note: Coefficients are estimated using White's procedure (Citation1980). The dependent variable is “log imports.” ∗denotes less than 10% level of significance; ∗∗denotes less than 5% level of significance; and ∗∗∗denotes less than 1% level of significance.
Note: Coefficients are estimated using White's procedure (Citation1980). The dependent variable is “log imports.” ∗denotes less than 10% level of significance; ∗∗denotes less than 5% level of significance; and ∗∗∗denotes less than 1% level of significance.
Note: Coefficients are estimated using White's procedure (Citation1980). The dependent variable is “log imports.” ∗denotes less than 10% level of significance; ∗∗denotes less than 5% level of significance; and ∗∗∗denotes less than 1% level of significance.