미국의 서비스무역 결정요인 분석

초록

This paper analyzes the determinants of US services trade using traditional gravity model. We particularly examine the effectiveness of controlling sample selection bias by employing the Heckman Sample Selection Model to empirically establish the theoretical expectations of gravity model. Based upon the results obtained in the estimation of the US services trade as a whole, lamda(λ) controlling sample selection bias is statistically significant, and all the explanatory variables seem to be statistically consistent with the theoretical model. This contrasts with those research outputs without controlling sample selection bias such as Freund and Weinhold(2002). Meanwhile, estimation outputs of sub-industries such as financial, transportation and insurance under US services industry are not entirely compatible with the theoretical anticipations even with controlling sample selection bias. These mixed outputs indicate that introducing lamda(λ) only is not enough to reduce the gap between theoretical and empirical forecasts on the determinants of services trade,which seems to stem from the heterogeneity of services industry. It is worth noting that foreign investments turn out to be a critical factor in promoting services trade, irrespective of recent rapid development of IT technology. It means that commercial presence rather than cross-border supply still works as a focal facilitator of services trade through localization, i.e., physical proximity of suppliers and users of services.

키워드