680 International Organization does appear that export interests now have a greater impact on those voting decisions,as Michael Gilligan's study of twelve trade bills pre-and post-RTAA indicates.36 But if pressure from import-competing interests has also risen,the net political effect of the change,and the extent to which it can be regarded as self-perpetuating,is problematic. To support their claim that all political actors became less protectionist as a result of the RTAA,Bailey,Goldstein,and Weingast point out that Republicans voted in greater numbers for liberalization between 1934 and 1962.This happy picture is darkened considerably if we look past 1962.Figure 2 shows the percentages of party members voting for trade liberalization-or against new protection-on major trade legislation in Congress since 1870.37 While it seems clear that Republican prefer- ences have indeed become more liberal on trade,the preferences of Democrats have shifted in the opposite direction.The result has been the creation of significant divi- sions and tensions within the parties over the trade issue.This outcome is not at all consistent with a claim that trade has created its own supporters and eliminated its opposition.Something else has been going on. Trade Policy Coalitions,Partisanship,and the RTAA The dominant"magic bullet"arguments about the success of the RTAA reform,by emphasizing the benefits derived from delegating authority to the executive(to avoid the evils of logrolling)or from entrenching the principle of reciprocity in policymak- ing (to court the support of export industries),provide reasons why the reform might have aided trade liberalization.But they are far less successful in providing a coher- ent account of this institutional innovation that also explains how it was engineered and sustained in the face of protectionist opposition.Most importantly,neither thesis can explain the decline in partisanship in votes on trade legislation in Congress since the 1930s and the growing divisions over trade within the parties. A stronger explanation for the RTAA and later liberalizations in U.S.trade policy emerges from evidence of a profound shift in the preferences of the different constitu- encies of each party.The story is best told with the aid of a simple model of the U.S. economy and decision making by members of Congress(see Appendix II for the formalization).Begin with the standard general-equilibrium model in which the re- turns to owners of"specific"factors of production are tied closely to the fortunes of the industry in which they are employed.38 Factors specific to export industries re- ceive a real increase in returns due to trade liberalization,whereas those employed in import-competing industries lose in real terms.39 Owners of these different types of 36.Bailey,Goldstein,and Weingast offer comparatively little evidence on this front:they examine House votes on trade bills in 1953 and 1962,showing that a district measure of export dependence has significant effects on legislative votes for liberalization. 37.These votes are described in more detail in Appendix I. 38.Jones 1971.See also Mussa 1974 and 1982. 39.Returns on mobile factors,of course,rise relative to the price of the imported good,but fall relative to the price of exports,so that the income effects of trade for owners of this factor depend on patterns of consumption.680 International Organization does appear that export interests now have a greater impact on those voting decisions, as Michael Gilligan's study of twelve trade bills pre- and post-RTAA indicates.36 But if pressure from import-competing interests has also risen, the net political effect of the change, and the extent to which it can be regarded as self-perpetuating, is problematic. To support their claim that all political actors became less protectionist as a result of the RTAA, Bailey, Goldstein, and Weingast point out that Republicans voted in greater numbers for liberalization between 1934 and 1962. This happy picture is darkened considerably if we look past 1962. Figure 2 shows the percentages of party members voting for trade liberalization-or against new protection-on major trade legislation in Congress since 1870.37 While it seems clear that Republican prefer￾ences have indeed become more liberal on trade, the preferences of Democrats have shifted in the opposite direction. The result has been the creation of significant divi￾sions and tensions within the parties over the trade issue. This outcome is not at all consistent with a claim that trade has created its own supporters and eliminated its opposition. Something else has been going on. Trade Policy Coalitions, Partisanship, and the RTAA The dominant "magic bullet" arguments about the success of the RTAA reform, by emphasizing the benefits derived from delegating authority to the executive (to avoid the evils of logrolling) or from entrenching the principle of reciprocity in policymak￾ing (to court the support of export industries), provide reasons why the reform might have aided trade liberalization. But they are far less successful in providing a coher￾ent account of this institutional innovation that also explains how it was engineered and sustained in the face of protectionist opposition. Most importantly, neither thesis can explain the decline in partisanship in votes on trade legislation in Congress since the 1930s and the growing divisions over trade within the parties. A stronger explanation for the RTAA and later liberalizations in U.S. trade policy emerges from evidence of a profound shift in the preferences of the different constitu￾encies of each party. The story is best told with the aid of a simple model of the U.S. economy and decision making by members of Congress (see Appendix I1 for the formalization). Begin with the standard general-equilibrium model in which the re￾turns to owners of "specific" factors of production are tied closely to the fortunes of the industry in which they are employed.38 Factors specific to export industries re￾ceive a real increase in returns due to trade liberalization, whereas those employed in import-competing industries lose in real terms.39 Owners of these different types of 36. Bailey, Goldstein, and Weingast offer comparatively little evidence on this front: they examine House votes on trade bills in 1953 and 1962, showing that a district measure of export dependence has significant effects on legislative votes for liberalization. 37. These votes are described in more detail in Appendix I. 38. Jones 1971. See also Mussa 1974 and 1982. 39. Returns on mobile factors, of course, rise relative to the price of the imported good, but fall relative to the price of exports, so that the income effects of trade for owners of this factor depend on patterns of consumption