Forty-foureconomic, social, political, and policy variables are tested for their significance in discriminating among three groups of developing countries designated "unattractive,""moderately attractive," and "highly attractive" with respect to foreign investment in manufacturing.Among the six essential discriminators, the only policy variable to emerge is the corporate tax level. * Do the policies of host developing countries have a significant influence on their capacity to attract direct foreign investment in manufacturing? Specifically, do host country policies on corporate taxation, tax incentives, joint ventures, local content requirements, and limitations on foreign personnel discriminate between "unattractive," "moderately attractive," and "highly attractive" developing countries with respect to foreign investment? Of the many possible economic, social, political, and policy variables that may influence direct foreign investment in manufacturing, which variables are the best discriminators among the three groups of developing countries? To answer these questions, stepwise discriminant analysis was applied to a sample of 41 developing countries. Forty-four economic, social, political, and policy variables were tested for their significance in discriminating between three groups of these countries.1 Data on nonextractive direct investment inflows were compiled for 70 developing countries that were then classified into three groups according to their average annual per capita inflow of nonextractive direct investment over the period 1966-1970. The Unattractive Group received an average annual per capita inflow of one U.S. dollar or less. The median per capita inflow for the remaining countries was $4.10. Countries with inflows less than the median value (but more than one dollar) were classified as Moderately Attractive whereas those with inflows over $4.10 were classified as Highly Attractive. These dollar amounts constitute, therefore, the criterion variables used in this study. Because of data limitations on possible determinants (particularly on policy factors), only 41 countries were subsequently used in the discriminant analysis. Table 1 indicates the classification of the sample countries.
The choice of capital inflows rather than capital stocks as the dependent variable wasmade for two reasons. First, comprehensive data on stocks of direct foreign investment in developing countries were available only up to 1967. Moreover, informationwas not available on the accumulation periods of the pre-1967 investment stocks in many coun-* Franklin R. Rootis Associate Professorand Chairman of the Multinational Enterprise Unitof the WhartonSchool. He holds an MBAfromthe WhartonSchool and a Ph.D.fromthe University of Pennsylvania. His publications includetwo books: Strategic Planning for ExportMarketing and International Tradeand Investment (FourthEdition). He is currently involvedin research projects on multinational enterprise and publicpolicy and political risk management underthe auspices of the WorldwideInsti...
