670 International Organization districts.The second version emphasizes how empowering the president to negoti- ate trade agreements that elicited reciprocal tariff reductions from other nations gen- erated larger gains for,and thus more political support from,export interests.2 I argue that,though these accounts provide sound reasons why the RTAA might have helped to produce more liberal policy outcomes ex post facto,they fail to pro- vide an adequate explanation for just how the institutional innovation was achieved and sustained in the face of protectionist opposition.Instead,a closer examination of the changing nature of the underlying trade policy coalitions,and their relationships with the two parties,makes more sense of the American experience.The RTAA system was instituted by a Democratic majority with a core constituency of interests that favored more liberal trade policies and hoped to make them more resilient to Republican majorities in the future.But it narrowly survived the Republican majori- ties that did follow in the 1940s and 1950s only because growing divisions over the trade issue arose within the Republicans'electoral base.These changes reflected the dramatic,exogenous effects of World War II on U.S.export and import-competing industries as well as longer-term shifts in U.S.comparative advantage and in party constituencies.In this alternative account of the RTAA and U.S.trade liberalization, the historical coincidence of several profound economic and political changes are central to the eventual outcome.As a consequence,the RTAA story offers few if any lessons for trade policy reform in the current system.It does emphasize,how- ever,that since trade liberalization was driven by a particular alignment of societal coalitions in the past,it might well be stymied or reversed by a different alignment in the future.This possibility for change is ruled out by conventional accounts of the RTAA. In the next section I briefly discuss the origins of the RTAA and the evidence that it coincided with a dramatic shift in U.S.trade policy.The third section examines the claim that the reform was essentially a congressional act of self-restraint,aimed at eliminating logrolling in tariff legislation by delegating authority over policy to the president.The fourth section then considers an alternative claim that the reform was a Democratic ploy to swing the political balance in favor of free-trade interests by basing trade policymaking on reciprocal concessions negotiated by the executive branch.I show that neither view fits well with the record of congressional voting on trade bills during the last century.Moreover,both stories neglect important exog- enous changes in party constituencies and societal preferences that had crucial ef- fects on congressional votes to extend the RTAA authority and liberalize trade after 1945.In the fifth section I develop this point in more detail by outlining a simple model of distributive conflict over trade and its manifestation within Congress.The model emphasizes the role of dynamic shifts in the preferences of societal groups, and the positions taken by the parties,in the evolution of U.S.trade policy. 1.Lohmann and O'Halloran 1994. 2.Bailey,Goldstein,and Weingast 1997.670 International Organization districts.' The second version emphasizes how empowering the president to negoti￾ate trade agreements that elicited reciprocal tariff reductions from other nations gen￾erated larger gains for, and thus more political support from, export interests2 I argue that, though these accounts provide sound reasons why the RTAA might have helped to produce more liberal policy outcomes ex post facto, they fail to pro￾vide an adequate explanation for just how the institutional innovation was achieved and sustained in the face of protectionist opposition. Instead, a closer examination of the changing nature of the underlying trade policy coalitions, and their relationships with the two parties, makes more sense of the American experience. The RTAA system was instituted by a Democratic majority with a core constituency of interests that favored more liberal trade policies and hoped to make them more resilient to Republican majorities in the future. But it narrowly survived the Republican majori￾ties that did follow in the 1940s and 1950s only because growing divisions over the trade issue arose within the Republicans' electoral base. These changes reflected the dramatic, exogenous effects of World War I1 on U.S. export and import-competing industries as well as longer-tern1 shifts in U.S. comparative advantage and in party constituencies. In this alternative account of the RTAA and U.S. trade liberalization, the historical coincidence of several profound economic and political changes are central to the eventual outcome. As a consequence, the RTAA story offers few if any lessons for trade policy reform in the current system. It does emphasize, how￾ever, that since trade liberalization was driven by a particular alignment of societal coalitions in the past, it might well be stymied or reversed by a different alignment in the future. This possibility for change is ruled out by conventional accounts of the RTAA. In the next section I briefly discuss the origins of the RTAA and the evidence that it coincided with a dramatic shift in U.S. trade policy. The third section examines the claim that the reform was essentially a congressional act of self-restraint, aimed at eliminating logrolling in tariff legislation by delegating authority over policy to the president. The fourth section then considers an alternative claim that the reform was a Democratic ploy to swing the political balance in favor of free-trade interests by basing trade policymaking on reciprocal concessions negotiated by the executive branch. I show that neither view fits well with the record of congressional voting on trade bills during the last century. Moreover, both stories neglect important exog￾enous changes in party constituencies and societal preferences that had crucial ef￾fects on congressional votes to extend the RTAA authority and liberalize trade after 1945. In the fifth section I develop this point in more detail by outlining a simple model of distributive conflict over trade and its manifestation within Congress. The model emphasizes the role of dynamic shifts in the preferences of societal groups, and the positions taken by the parties, in the evolution of U.S. trade policy. 1. Lohmann and O'Halloran 1994. 2. Bailey, Goldstein, and Weingast 1997