Empirical Tests of the Factor Abundance Theory: What Do They Tell Us?

Leamer and Levinsohn (1995) have recently proposed a formula for empirical work in international microeconomics summarized by the injunction to “Estimate, Don’t Test!” This is based on a perception that much of the empirical work has failed to be influential as it tests propositions we know ex ante to be false. We argue that the empirical work has been more influential than they suggest. We use the example of recent work on the Heckscher-Ohlin-Vanek model of trade to show how a cumulation of tests may be influential, even as no one test on its own would be. 1 Leamer and Levinsohn also provide a caution the other direction -not to treat the theory too casually. Leamer’s (1980) own work on the conceptual problems underlying the Leontief paradox provide a striking object lesson in the dangers of turning to empirics without a clear tie to the theory. Empirical Tests of the Factor Abundance Theory: What Do They Tell Us? 1 Estimate, Don’t Test? In their masterful survey of recent empirical work on international microeconomics, Leamer and Levinsohn (1995) expound a distinctive view on the appropriate aims of empirical work in this area. In a phrase, they exhort researchers to “Estimate, don’t test!” The foundation for this advice is a view that empirical work on trade has had very little influence on our beliefs about the causes of trade. And they believe that this is true because empirical researchers have taken the theory too seriously. In Leamer and Levinsohn’s view, we should Understand that theorems are neither true nor false. They are sometimes useful and sometimes misleading. If we approach a data base with the contrary attitude, hoping to determine the verity or falsity of a theorem, we may statistically “reject” the theory, but leave it completely unharmed nonetheless. After all, we already knew it wasn’t literally true. [p. 1341] That the study of international trade has been dominated by theory is evident to anyone who turns the pages of the journals. Until recently, empirical work on international trade has been accorded little of the prestige bestowed on theoretical advances. And, no doubt, much of the existing empirical work fails to strongly influence the profession because it takes the theoretical caricature too literally. Yet we believe that Leamer and Levinsohn understate the true influence exerted by the existing empirical work. And we believe that the anti-testing perspective is not the appropriate remedy. We will pursue these themes within the context of a review of selected work on the