Enterprise Hall, 318
February 10, 2008, 07:00 PM to 07:00 PM
This paper assesses the implications of spatial dependence of outward FDI flows to clarify the importance and prominence of complex foreign direct investment by multinational enterprises, which are neither purely horizontal nor vertical integration strategies. Empirical tests compare the ordinary least squares, fixed effects, and spatial autoregression techniques for two different models to identify and estimate complex foreign direct investment. The spatial autoregression method directly address spatial correlated residuals while the fixed effect estimation procedure and a fully specified ordinary least squares produce similar regression coefficients. Both models indicate the presence of export platform multinationals while each model has limitations. Then a modified gravity model first assesses the empirical evidence for complementarity between FDI and trade then indicates the predominance of multinational organization strategy. The asymmetric impacts of exports and i mports on foreign direct investment are examined, then the effects of both FDI and trade on long-run growth are used to estimate for the effectiveness of trade to increase income. The direct investment channel is found to be a strong and growing channel of importance for developement. Finally, the regional impacts of trade on long-run growth indicates that the spatial advantages of the gravity model can be further exploited using innovations such as an economically-weighted distance variable and regional sub-samples. The findings support broader uses of geographic specifications beginning with a re-estimation of the effect of trade on income. The model is relatively more robust for intra-region trade?underscoring the role of competition in limiting globalization and producing varying effects for the estimates of trade on income.