Institutional Reform and Trade Liberalization 673 ing years,and,by 1939,a full quarter of dutiable imports were subject to quantitative limits.12 In the postwar era,of course,the most popular forms of the "new protection- ism,"such as voluntary export restraints(VERs),have severely restricted imports of major items including textiles and clothing,footwear,steel,and automobiles.13 Perhaps most important,the tariff estimates shown in Figure 1 are extremely sen- sitive to import price levels when the actual tariffs imposed are in the form of specific duties rather than ad valorem duties.If import price levels rise,fixed specific duties will yield declining estimates of average ad valorem tariffs.Since roughly two-thirds of dutiable imports to the United States were subject to specific duties in the 1940s, and import prices rose some 81 percent between 1945 and 1955,this represents a glaring problem.Douglas Irwin has calculated that over three-quarters of the decline in estimated tariff levels between 1934 and 1967 can be attributed to the dramatic rise in import prices in the postwar period-only the remaining one-quarter of the esti- mated drop in tariffs was actually caused by changes in policy.14 The common story told about a radical change in trade policy after 1934 thus needs to be drastically amended. Even after taking such qualifications into account,the 1930s and 1940s still appear to be a turning point of sorts in U.S.trade politics.Trade policy did shift in a more liberal direction,and the change broadly coincided with the passage of the RTAA. What happened?Without doing too much damage to the subtleties,we can distin- guish two major "magic bullet"theses in the literature:(1)members of Congress recognized that the existing rules produced sub-optimal results because of logrolling and chose to delegate authority to the executive branch to ensure more efficient policies;and(2)leaders of the Democratic party,hoping to make trade liberalization more lasting,chose to base it on reciprocal concessions that would solidify support from export interests.Each of these broad claims warrants scrutiny. The RTAA as a Solution to Logrolling The most common view that has emerged in all the various accounts of the RTAA is that the delegation of authority it entailed was the result of members of Congress recognizing that logrolling in the passage of tariff legislation produced highly protec- tionist outcomes in which all were left worse off.15 Tariff setting in a legislature is regarded as a classic case of"distributive politics"in the famous Lowi typology.16 Tariff benefits can be doled out to import-competing producers in one area without 12.Diebold1941.19. 13.Bhagwati 1988. 14.Irwin 1996.He develops an econometric model of the estimated average tariff rate from 1865 to 1973,then uses parameter estimates to distinguish the effects of import price inflation and changes in commercial policy.Notice that just as the price inflation of the 1940s and 1950s leads to a vast overstate- ment of tariff reductions in those years,the price deflation during the Depression years of the early 1930s leads to an overstatement of the tariff increase in the Smoot-Hawley Act of 1930. 15.See Baldwin 1985;Goldstein 1988;and Destler 1992. 16.L0wi1979,690.Institutional Reform and Trade Liberalization 673 ing years, and, by 1939, a full quarter of dutiable imports were subject to quantitative limits.12 In the postwar era, of course, the most popular forms of the "new protection￾ism," such as voluntary export restraints (VERs), have severely restricted imports of major items including textiles and clothing, footwear, steel, and automobiles.13 Perhaps most important, the tariff estimates shown in Figure 1 are extremely sen￾sitive to import price levels when the actual tariffs imposed are in the form of specific duties rather than ad valorem duties. If import price levels rise, fixed specific duties will yield declining estimates of average ad valorem tariffs. Since roughly two-thirds of dutiable imports to the United States were subject to specific duties in the 1940s, and import prices rose some 81 percent between 1945 and 1955, this represents a glaring problem. Douglas Irwin has calculated that over three-quarters of the decline in estimated tariff levels betwe& 1934 and 1967 can be attributed to the dramatic rise in import prices in the postwar period-only the remaining one-quarter of the esti￾mated drop in tariffs was actually caused by changes in policy.14 The common story told about a radical change in trade policy after 1934 thus needs to be drastically amended. Even after taking such qualifications into account, the 1930s and 1940s still appear to be a turning point of sorts in U.S. trade politics. Trade policy did shift in a more liberal direction, and the change broadly coincided with the passage of the RTAA. What happened? Without doing too much damage to the subtleties, we can distin￾guish two major "magic bullet" theses in the literature: (1) members of Congress recognized that the existing rules produced sub-optimal results because of logrolling and chose to delegate authority to the executive branch to ensure more efficient policies; and (2) leaders of the Democratic party, hoping to make trade liberalization more lasting, chose to base it on reciprocal concessions that would solidify support from export interests. Each of these broad claims warrants scrutiny. The RTAA as a Solution to Logrolling The most common view that has emerged in all the various accounts of the RTAA is that the delegation of authority it entailed was the result of members of Congress recognizing that logrolling in the passage of tariff legislation produced highly protec￾tionist outcomes in which all were left worse off.15 Tariff setting in a legislature is regarded as a classic case of "distributive politics" in the famous Lowi typology.16 Tariff benefits can be doled out to import-competing producers in one area without 12. Diebold 1941, 19. 13. Bhagwati 1988. 14. Irwin 1996. He develops an econometric model of the estimated average tariff rate from 1865 to 1973, then uses parameter estimates to distinguish the effects of import price inflation and changes in commercial policy. Notice that just as the price inflation of the 1940s and 1950s leads to a vast overstate￾ment of tariff reductions in those years, the price deflation during the Depression years of the early 1930s leads to an overstatement of the tariff increase in the Smoot-Hawley Act of 1930. 15. See Baldwin 1985; Goldstein 1988; and Destler 1992. 16. Lowi 1979,690