Comparative Advantage in a Multi-Factor World
探讨在多要素世界中，基于行业劳动生产率的古典比较优势理论是否仍能预测贸易模式，结论是平均而言有效但非完全准确。
This paper examines the question: Is there any reason to expect that a country's trade pattern can be accurately predicted on the basis of relative industry labor productivity differentials across countries (i.e., comparative advantage in its Classical or Ricardian sense), given that we inhabit a multi-factor world? The answer to this question is of some interest, since early attempts to test comparative advantage appeared to be quite successful,2 or at least as successful as tests of other theories based on more plausible assumptions. Although there is some dispute over the validity of these tests,3 and later efforts have not been as successful,4 there still appears to be a fairly widespread belief that comparative advantage retains some explanatory power.5 This paper examines the extent to which such a belief can be justified theoretically, given that the actual trade pattern is determined in a quite different manner (e.g., on the basis of factor endowment differences) than the Classical theory described. We are therefore not attempting to refute or verify the validity of the Ricardian model, but only to determine the extent to which its predictions of the trade pattern can also be expected to hold in more general models.6 Our general conclusion will be that while comparative advantage is unlikely to prove a wholly accurate predictor of the trade flow in each individual instance, on average it can be expected to perform