Bilateral Treaties and The Most-Favored-Nation Clause: The Myth of Trade Liberalization in the Nineteenth Century

Abstract

Textbook accounts of the Anglo-French trade agreement of 1860 argue that it heralded the beginning of a liberal trading order. This alleged success holds much interest from a modern policy point of view, for it rested on bilateral negotiations and most-favored-nation clauses. With the help of new data on international trade (the RICardo database), the authors provide empirical evidence and find that the treaty and subsequent network of MFN trade agreements coincided with the end of a period of unilateral liberalization across the world. They also find that it did not contribute to expanding trade at all. This is contrary to a deeply rooted belief among economists, economic historians, and political scientists. The authors draw a number of policy lessons that run counter to the conventional wisdom and raise skepticism toward the ability of bilateralism and MFN arrangements to promote trade liberalization.

Something Rational in the State of Denmark? The Case of an Outsider in the Cobden-Chevalier Network 1860-1875

“…Recent studies of the post-1945 world trade regime established through the General Agreement on Tariffs and Trade (GATT) and the World Trade Organization (WTO) have shown that the benefits for the average member -in terms of increased trade flows -might be insignificant (Rose 2004), while the core members who established the rules of the system and retained important bargaining power during its evolution gained significantly (Gowa andKim 2005, Westermeier 2008). Similar results have been obtained in recent studies of the first round of the Cobden-Chevalier network (Accominotti andFlandreau 2008, Lampe 2009a). …”

The French (Trade) Revolution of 1860: Intra-Industry Trade and Smooth Adjustment

“…Two potential answers for the absence of a more immediate reaction are: the liberalization was not very significant; or the adjustment costs of the shock were politically and economically tolerable. The former could be consistent with Flandreau and Accominotti (2008). The second view has not been well studied.…”

How Multiple Imputation Makes a Difference

“…Naturally, this change is largest in cases of disappearance, averaging 552% for all coefficients and 2,201% for key coefficients. Note, however, that these averages are heavily influenced by two outliers in which both figures are more than two standard deviations above the mean (Accominotti and Flandreau 2008;Kurtz and Brooks 2008). Excluding these studies, the averages are 206% and 117%, respectively, similar to those in cases of strengthening (233% and 201%) and weakening (219% and 47%) but notably higher than those in cases of no change (135% and 74%).…”