1. Institutional studies have done well in explaining the origins of public social provision and often take the form of case studies or close comparisons of the experiences of a few nations. Political theories have done well in explaining variation in social spending efforts among postwar capitalist democracies. For reviews, see Skocpol, Theda, Protecting Soldiers and Mothers (Cambridge, Mass., 1992)Google Scholar , introduction ; Huber, Evelyne, Ragin, Charles, and Stephens, John D., “Social Democracy, Christian Democracy, Constitutional Structure, and the Welfare State: Toward a Resolution of Quantitative Studies.” American Journal of Sociology 99 (1993): 711–49Google Scholar.

2. U.S. Committee on Economic Security, Report to the President (Washington, D.C., 1935), 9 Google Scholar . As late as December 24, 1934, Roosevelt wanted to combine the legislation for the Works Program and other economic security measures in one bill . Macmahon, Arthur W., Millett, John D., and Ogden, Gladys, The Administration of Federal Work Relief (Chicago, 1941), 26–27 Google Scholar ; Witte, Edwin, The Development of the Social Security Act: A Memorandum on the History of the Committee on Economic Security and Drafting and Legislative History of the Social Security Act (Madison, 1962), 77 Google Scholar ; Altmeyer, Arthur, The Formative Years of Social Security (Madison, 1966), 9 Google Scholar . “Social policy effort” is generally understood as the amount spent on social programs as a share of gross national product. On the U.S. efforts in the 1930s, see Amenta, Edwin, Bold Relief: Institutional Politics and the Origins of Modern American Social Policy (Princeton, 1998), introductionGoogle Scholar.

3. Dahl, Robert, Polyarchy: Participation and Opposition (New Haven, 1971).Google Scholar

4. David Mayhew refers to these as “traditional party organizations” and defines them as substantially autonomous, long-lasting, hierarchical, seeking to nominate candidates for a wide range of public offices, and relying substantially on material incentives . Placing Parties in American Politics (Princeton, 1986), 19–20 Google Scholar . For other notable arguments about “machine politics,” see Banfield, Edward C. and Wilson, James Q., City Politics (Cambridge, Mass., 1963)Google Scholar ; Shefter, Martin, “Party and Patronage: Germany, Italy, and England,” Politics and Society 7 (1977): 404–51Google Scholar ; Katznelson, Ira, City Trenches: Urban Politics and the Patterning of Class in the United States (New York, 1981)Google Scholar.

5. Mayhew also argues that patronage parties are unlikely to attract program builders, deter the participation of pro-spending groups like the labor movement, promote issue-less politics, and create a political culture of pessimism about the utility of government. Placing Parties, 292-94.

6. Stephens, John D., The Transition from Capitalism to Socialism (London, 1979)Google Scholar . For a review of arguments and evidence, see Esping-Andersen, Gosta and Kersbergen, Kees van, “Contemporary Research on Social Democracy,” Annual Review of Sociology 18 (1992): 187–208 Google Scholar . Others have amended this thesis by focusing on the discouraging impact on social policy of right-wing political parties and center-party routes to the adoption of social policy, suggesting that gains in U.S. public spending would result from the taking of power by the Democrats, considered a center party . Castles, Francis G. and Mair, Peter, “Left-Right Political Scales: Some ‘Expert’ Judgments.” European Journal of Political Research 12 (1984): 73–88 Google Scholar.

7. Esping-Andersen argues that a red-and-green coalition of workers and farmers was behind breakthrough Scandinavian social policies in the 1930s . Three Worlds of Welfare Capitalism (Princeton, 1990)Google Scholar , chap. 1. For a farmer-labor coalition argument in the American setting prior to the New Deal, see Sanders, Elizabeth, Roots of Reform: Farmers, Workers, and the American State, 1877–1917 (Princeton, 1999)Google Scholar introduction. Orloff argues that a coalition of the organized working class and policy experts was necessary for the passage of old-age pensions in the United States, Canada, and Great Britain. The Politics of Pensions: A Comparative Analysis of Britain, Canada, and the United States, 1880–1940 (Madison, 1993), chap. 2.Google Scholar

8. See, for instance , Piven, Frances Fox and Cloward, Richard A., Poor People's Movements: Why They Succeed, How They Fail (New York, 1977)Google Scholar , chap. 1 ; Sidney Tarrow, Power in Movement (New York, 1994)Google Scholar , chap. 10. Theda Skocpol argues that “widespread federated interests”—challenging organizations or reform groups organized across congressional districts—have the best chance of influencing U.S. social policy. Protecting Soldiers and Mothers, introduction.

9. Schlesinger, Arthur M. Jr, The Age of Roosevelt: The Crisis of the Old Order, 1919-1933 (Boston, 1957), 390–95Google Scholar ; Davis, Kenneth S., FDR: The New York Years, 1928-1933 (New York, 1979)Google Scholar , chaps. 3, 8 ; Schwarz, Jordan A., The Interregnum of Despair: Hoover, Congress, and the Depression (Urbana, III., 1970)Google Scholar , chap. 7 ; Brock, William R., Welfare, Democracy, and the New Deal (Cambridge, 1988)Google Scholar , chap. 4.

10. U. S. Bureau of the Census, Historical Statistics of the United States from Colonial Times to 1970 (Washington, D.C., 1976), series Y 84-134, Y 135-186, pp. 1075, 1077.Google Scholar

11. This group also includes Republicans from states dominated by traditional, patronage party organizations, Republicans from underdemocratized polities, and Democrats or Republicans affiliated with conservative third parties. There were few such legislators, however.

12. , Macmahon et al., Federal Work Relief, chaps. 1–3 Google Scholar ; Rosenman, Samuel I., ed., The Public Papers and Addresses of Franklin D. Roosevelt (New York, 1938-1935 Google Scholar , Items 45, 54, and 89 ; Howard, Donald S., The WPA and Federal Relief Policy (New York, 1943)Google Scholar , chap. 1 ; , Sherwood, Roosevelt and Hopkins, 65–71 Google Scholar ; Mcjimsey, George, Harry Hopkins: Ally of the Poor and Defender of Democracy (Cambridge, Mass., 1987)Google Scholar , chaps. 5-6 ; Charles, Searle F., Minister of Relief: Harry Hopkins and Depression (Syracuse, 1963), chaps. 6-7Google Scholar ; Badger, Anthony, The New Deal: The Depression Years (New York, 1989), 200–215 Google Scholar . For data on work program and social spending efforts, see U.S. National Resources Planning Board (NRPB) Committee on Long-Range Work and Relief Policies , Security, Work, and Relief Policies (Washington, D.C., 1942), 560–61Google Scholar ; , Bureau of the Census, Historical Statistics, 224, 1114 Google Scholar.

13. U.S. Federal Works Agency, Final Report on the WPA Program (Washington, D.C., 1946), 23–26.Google Scholar

14. , Howard, The WPA, 491, 517–20Google Scholar ; Federal Works Agency, Final Report, 23-25,37-41.

15. Mcjimsey, Harry Hopkins, 81–83. According to a schedule devised by the WPA, there were relatively large variations in monthly wages across five skill categories and between wage employees and supervisory employees, as well as slight variations across four regions of the country and across county size to take into account differences in the cost of living. Federal Works Agency, Final Report, 23–26, 37-41; Howard, The WPA, chap. 6.

16. Patronage politics also seeped in through the requirement that almost all WPA workers be taken from general assistance rolls, over which the WPA or the federal government had little say . Amenta, Edwin, Benoit, Ellen, Bonastia, Chris, Cauthen, Nancy K., and Halfmann, Drew, “Bring Back the WPA: Work, Relief, and the Origins of American Social Policy in Welfare Reform” Studies in American Political Development 12 (Spring 1998): 1–56 CrossRefGoogle Scholar.

17. The 1936 Democratic Party platform was explicit about public employment: “Where business fails to supply such employment, we believe that work at prevailing wages should be provided.” By contrast, the Republican platform criticized New Deal relief policies. Porter, Kirk H. and Johnson, Donald Bruce, National Party Platforms, 1840-1968 (Urbana, III., 1970), 360–63Google Scholar , 366 [quote on 362] ; Sherwood, Robert E., Roosevelt and Hopkins: An Intimate History (New York, 1948), 81–86 Google Scholar ; Schlesinger, Arthur M. Jr The Age of Roosevelt: The Politics of Upheaval (Boston, 1960), chap. 33Google Scholar.

18. , Schlesinger, The Politics of Upheaval, chap. 32 Google Scholar ; , Burns, Roosevelt, chap. 14 Google Scholar ; Milkis, Sidney M., The President and the Parties: The Transformation of the American Party System Since the New Deal (New York, 1993), 62–74 Google Scholar ; Ware, Susan, Beyond Suffrage: Women in the New Deal (Cambridge, Mass., 1981)Google Scholar , chap. 4 ; Weiss, Nancy J., Farewell to the Party of Lincoln: Black Politics in the Age of FDR (Princeton, 1981), chaps. 7, 9Google Scholar.

19. Bureau of the Census, Historical Statistics, series Y 84-134, Y 135-186, pp. 1075, 1077.

20. Inter-University Consortium for Political and Social Research (ICPSR); United States Congressional Roll Call Voting Records, 1789-1991 [Computer file] (Ann Arbor: Inter-University Consortium for Political and Social Research, 1992)Google Scholar ; , Patterson, Congressional Conservatism, 142–43Google Scholar.

21. U.S. Congress, Senate Special Committee to Investigate Unemployment and Relief, Unemployment and Relief: Report, 76th Cong. 1st sess., January 14, 1939, 3–6 Google Scholar ; , May, From New Deal to New Economics Google Scholar ; , Macmahon et al., Federal Work Relief, 140–42Google Scholar ; , Sherwood, Hopkins and Roosevelt, 91–99 Google Scholar ; Patterson, James T., Congressional Conservatism and the New Deal: The Growth of the Conservative Coalition in Congress, 1933-1939 (Lexington, Ky., 1967), 234–44Google Scholar ; , Rosenman, ed., Public Papers of Franklin Roosevelt, 1939 Google Scholar , Item 49; Federal Works Agency, Final Report.

22. U.S. NRPB Committee, Security, Work, and Relief Policies, 407 Google Scholar ; Federal Works Agency, Final Report, 8, 19 Google Scholar ; , Howard, The WPA, 361–68Google Scholar.

23. U.S. Bureau of the Census, Historical Statistics, 224 Google Scholar ; U.S. NRPB Committee, Security, Work, and Relief Policies, 558, 561 Google Scholar ; , Howard, The WPA, 854–57Google Scholar.

24. , Patterson, Congressional Conservatism.Google Scholar

25. In the Senate, an amendment immediately to restore the $150 million was defeated 47–46. Soon afterward Roosevelt asked Congress to appropriate the additional $150 million, but received only $100 million . New York Times, 6 January 1939, 1 Google Scholar ; 11 January 1939, 1 ; , Patterson, Congressional Conservatism, 294-97, 302–5Google Scholar.

26. Congress enacted these other important changes: (1) a requirement that local sponsors provide 25 percent of funding; (2) a $52,000 limit on nonfederal building projects; (3) a mandate to abolish geographical differentials in security wages not justified by differences in cost of living; (4) a formula for the distribution of WPA jobs to the states; (5) the elimination of the Federal Theatre Project and a requirement that other arts projects gain local sponsorship . New York Times, 15 June 1939, 1 Google Scholar ; 29 June 1939, 1; 1 July 1939, 1.

27. New York Times, 7 July 1939, 1 Google Scholar ; 11 July 1939, 7; 24 July 1939, 1; ICPSR, United States Congressional Roll Call Voting Records.

28. , Rosenman, ed., The Public Papers of Franklin D. Roosevelt, vol. 8, 1939 Google Scholar , Item 2, pp. 36-53 ; , Milkis, The President and the Parties, chaps. 5–6 Google Scholar ; , Macmahon et al., Federal Work Relief, chap. 6 Google Scholar.

29. Karl, Barry Dean, Executive Reorganization and Reform in the New Deal: Genesis of Administrative Management, 1900–1939 (Cambridge, Mass., 1963)Google Scholar ; Polenberg, Richard, Reorganizing Roosevelt's Government, 1936–1939 (Cambridge, Mass 1966), chap. 9 Google Scholar.

30. See, for instance , , Rosenman, ed., The Public Papers of Franklin D. Roosevelt, 1935 Google Scholar , vol. 9, Items 119, 120, 123, 125, and 128.

31. Bureau of the , Census, Historical Statistics, series Y 84-134, Y 135–186, pp. 1075, 1077.Google Scholar

32. New York Times, 5 January 1940, 8.Google Scholar

33. New York Times, 21 May 1941, 1 Google Scholar ; 14 June 1941, 1.

34. , Amenta et al., “Bring Back the WPA,” Federal Works Agency, Final Report, p. v Google Scholar ; Katz, Michael, In the Shadow of the Poorhouse: A Social History of in America (New York, 1986), 224–34Google Scholar.

35. Other major New Deal social spending programs follow a similar trajectory—initiation during the reform-oriented regime of 1935, expansion and institutionalization in the mid- to late 1930s, but suffering retrenchment rather than elimination in the mid-1940s . , Amenta et al., “Bring Back the WPA.”Google Scholar

36. See Leman, Christopher, “Patterns of Policy Development: Social Security i n the United States and Canada,” Public Policy 25 (1977): 261–91Google Scholar ; Block, Fred, “The Ruling Class Does Not Rule,” Socialist Revolution 33 (1977): 6–28 Google Scholar.

37. See Chandler, Lester V., America's Greatest Depression, 1929-1941 (New York, 1970), 21.Google Scholar

38. For comparisons among industrial countries, see Gourevitch, Peter, Politics Hard in Times: Comparative Responses to International Economic Crises (Ithaca, N.Y. 1986)Google Scholar . For a comparison of the United States and Great Britain in the adoption of Keynesian deficit spending techniques, see Weir, Margaret, “Ideas and Politics: The Acceptance of Keynesianism in Britain and the United States,” in The Political Power of Economic Ideas: Keynesianism Across Nations, ed. Hall, Peter A. (Princeton, 1989), 53–86 Google Scholar . On British social policy, see Fraser, Derek, The Evolution of the British Welfare State: A History of Social Policy Since the Industrial Revolution, 2d ed. (London, 1984)Google Scholar , chap. 8. For a comprehensive analysis of the adoption of core social insurance programs in seventeen industrial countries in the 1930s, see Hicks, Alexander, Social Democracy and Welfare Capitalism: A Century of Income Security tics (Ithaca, N.Y., 1999)Google Scholar , chaps. 3 and 4. For changes in U.S. and British social spending efforts in the 1930s, see , Amenta, Bold Relief, chap. 4 Google Scholar.

39. , Rosenman, ed., Public Papers of Franklin D. Roosevelt, vol. 4, 1935, Item pp. 19–20.Google Scholar

40. For a discussion of choosing roll-call votes, see Katznelson, Ira, Geiger, Kim, and Kryder, Daniel, “Limiting Liberalism: The Southern Veto in Congress, 1933–1950,” Political Science Quarterly 108 (1993): 283–306 Google Scholar . We selected all close votes, excluding the trivial and redundant, and then brought back a few votes that were not extremely close but seemed important. For the Senate, a close vote was one for which the margin of victory was twenty votes or less and for which two-thirds (63) or more of the Senators cast votes. (Our computations include legislators' indications of support or opposition through pairings and announcements recorded in the Congressional Record.) This yielded twenty votes. When multiple votes were taken the same day on a single issue, we chose only the closest vote with the most Senators voting and excluded the others. This led to three exclusions. We then examined the Congressional Record and newspaper accounts in order to exclude votes that were either misleading or trivial. We also excluded four votes in which Republicans insincerely joined with progressives chiefly for the purpose of embarrassing the Roosevelt administration. These included two early votes on prevailing wages, votes 74–24 and 74–34- Roosevelt initially opposed prevailing wages because they were predicated at first on a full work week and would thus increase the WPA budget by about 50 percent. After a prevailing wage amendment (vote 74-24) passed, the administration had the bill recommitted to committee and the language deleted, offering assurances that the president would set pay rates that would “not affect adversely or otherwise tend to decrease the going rates of wages.” The administration's language was substituted by a vote of 83-2. We similarly excluded a 1939 vote specifying that heads of families with the greatest need should be the first hired and the last to be laid off (vote 76-59) and the 1938 vote on civil-service status for WPA employees (vote 75-172). We also excluded two trivial votes: a 1935 losing proposal requiring that $500 million be spent on school construction (vote 74–40) and a June 1938 amendment to establish a minimum monthly wage of $40 for WPA work (vote 75-171). We exclude the latter because the proposal would have had little practical effect–a conference committee was meeting on a wages and hours bill that would apply a comparable wage to WPA work. We also included four budgetary votes to gain leverage on this issue across time: a 1935 vote to reduce the WPA appropriation (vote 74–38), a 1939 vote to decrease the deficiency appropriation (76–29), and a 1941 vote to increase the appropriation (vote 77–60). The last (vote 76–66) was to commit an amendment to the Committee on Education and Labor, which was expected to act in favor of prevailing wages. In any case, we analyzed both the original 20 votes and our smaller modified set of 15, finding little difference in the results (not shown, but available). By contrast, of the 33 House votes only three had a margin of less than 100, and each took place in 1941 or 1942. The votes were taken from ICPSR, United States Congressional Roil Call Voting Records Google Scholar.

41. Some qualifications about the meaning of roll-call votes are in order. Power i s often exercised at earlier stages of the legislative process, such as in congressional committees. Legislators from the underdemocratized South were rarely defeated and thus accumulated seniority that allowed them to chair approximately one-half of Senate committees during the New Deal and influence legislation that they opposed. See Shelly, Mack C. II, Permanent Majority: The Conservative Coalition in the United States Congress (Birmingham, Ala., 1983)Google Scholar ; Key, V. O., Southern Politics in State and Nation (New York, 1949)Google Scholar . WPA bills began in the appropriations committees, chaired by southerners Senator Carter Glass (D-Va.) and Representative James Buchanan (D-Tex.), for much of the WPA's existence. The appropriations subcommittees on work relief were also chaired by southerners, Senator James F. Byrnes (D-S.C.) and Representative Clifton A. Woodrum (D-Va.), who no doubt influenced the content of Roosevelt's proposals.

42. To be counted as having a strong labor movement, a state needed to score high on two measures: union members in 1939 as a share of the nonagricultural employed in 1940, and the income of the state federation of labor from 1938 to 1940 as a percentage of the nonagricultural labor force, with the latter employed to indicate political mobilization. For union membership, see Troy, Leo and Sheflin, Neil, U.S. Union Sourcebook (West Orange, N.J., 1985), 7–8 Google Scholar ; for state federation of labor income, see Amenta, Edwin and Poulsen, Jane D., “Social Politics in Context: The Institutional Politics Theory and State-Level U.S. Social Spending Policies at the End of the New Deal,” Social Forces 75 (1996): 33–60 Google Scholar ; for the nonagricultural labor force, see U.S. Bureau of Census, Statistical Abstract of the United States (Washington, D.C., 1948), 194 Google Scholar , 196. The fifteen states, which constitute a conservative estimate, are Arizona, California, Colorado, Illinois, Indiana, Iowa, Minnesota, Montana, Nevada, Oregon, Pennsylvania, Utah, Washington, Wisconsin, and Wyoming.

43. This small number corresponds to the well-recognized but by 1939 tiny group of progressive Republican senators, who were mainly from the West and often in states where labor had made great advances. On th e decline of progressive Republicans, see Weed, Clyde P., The Nemesis of Reform: The Republican Part: During the New Deal (New York, 1994), chap. 6Google Scholar.

44. For discussions of this index, see Rice, Stuart A., “The Behavior of Legislative Groups: A Method of Measurement,” Political Science Quarterly 40 (March 1925)Google Scholar ; idem , Quantitative Methods in Politics (New York, 1928)Google Scholar ; Riker, William H., “A Method for Determining the Significance of Roll Calls in Voting Bodies,” in Wahlke, John C. and Eulau, Heinz, eds., Legislative Behavior: A Reader in Theory and Re search (Glencoe, III., 1959)Google Scholar.

45. There are three plausible and complementary explanations. First, Katznelson and colleagues in “Limiting Liberalism” argue that southern Democrats from the underdemocratized part of the polity were fairly reliable early supporters of New Deal social spending initiatives, but only if they were written to appease them. Very possibly the negative impact of underdemocratized Democrats on social spending was registered at the budget-proposal stage, before votes were taken. Second, early southern Democratic votes in favor of the WPA budget may have been a result of a weak strategic position of the anti-social-policy grouping in the face of large pro-spending contingents in Congress after 1934- The underdemocratized Democrats may have been able to act on their opposition to the WPA only after the losses suffered by the pro-spending groupings in 1938 ended their majority in Congress. Third, southern Democrats may have turned against the WPA in later years because the administration was increasingly making steps to ensure its permanence. All in all, the evidence suggests that Democrats from the underdemocratized part of the polity were not opposed to redistributive national government spending programs, so long as they provided a no-strings-attached regional redistribution of income. But southern Democrats did not want national control over spending, or national standards in it, concerned as they were about the impact of spending on the operations of elites who paid low wages.

46. These votes are on a June 1937 measure to set local contributions at 25 percent (75–42), which was opposed by the WPA; a June 1937 measure to turn work relief administration to the states (75-44), which was opposed by the WPA; a February 1938 measure to repeal the Woodrum amendment requiring the WPA to avoid shortfalls (75–119), the repeal being supported by the WPA; and a June 1938 measure to prohibit the use of work relief to influence elections (75-166), which was opposed by the WPA. The latter measure was a cleverly worded vote of no confidence in the program and predictably ran along party lines.

47. See Congressional Record, vol. 83, part 2, pp. 2285-98.

48. In all the votes, the open, democratic left (34) and the some-forces (32) groups supported the WPA, with the underdemocratized part of the polity providing weak opposition (-4). The no-forces group, mainly consisting of Republicans (-75), opposed the WPA.

49. Rosenthal, Howard L. and Poole, Keith T., United States Congressional Roll Call Voting Records, 1789-1987: Reformatted Data. [Computer file] (Pittsburgh: Howard L. Rosenthal and Keith T. Poole, Carnegie Mellon University, Graduate School of Industrial Administration [producers] Google Scholar . Ann Arbor: Inter-university Consortium for Political and Social Research [distributor], 1991), vote 77-38. A proposal to devolve relief administration to states and localities, which was recurrently offered and typically failed overwhelmingly, was defeated in June 1942 by only 184–139. Southern opposition was much weaker (14) than that of other Democrats (84), vote 77–132. The only other House vote with a margin less than 100 was a vote on the same day (vote 131), prohibiting new WPA projects, with a similarly close and losing result. A relatively close vote (77-49) indicated an instance of coalition between Republicans and southern Democrats–the June 1941 passage, 214 to 114, of an amendment by Representative Everett Dirksen (R-Ill.) to prohibit the use of any WPA funds to compensate the president of the Workers' Alliance, David Lasser. The amendment had strong support from Republicans (-80) and moderate support from southern Democrats (-52). It was opposed by nonsouthern Democrats (32).

50. It should be noted that at the beginning of 1939 Hopkins had taken over as Secretary of Commerce; he was being groomed at the time potentially to become Roosevelt's successor. Colonel Francis Harrington was named to head the WPA that year and thus senators were not literally voting with or against Hopkins—though he was still associated in the public's view with the WPA.

51. Senators Sheridan Downey of California, James O'Mahoney of Wyoming, and Alva Adams of Colorado voted with the WPA on all measures concerning the character of the WPA except the one to refer the issue of prevailing wages to the Committee on Education and Labor. O'Mahoney and Adams were members of the Appropriations Committee, which would lose control over the issue and possibly the WPA as a whole.

52. See Dorsett, Lyle W., Franklin D. Roosevelt and the City Bosses (Port Washington, N.Y., 1977)Google Scholar ; Erie, Stephen P., Rainbow's End: Irish Americans and the Dilemmas of Urban Machine Politics, 1840-1945 (Berkeley and Los Angeles, 1988)Google Scholar ; Charles, Searle F., Minister of Relief: Harry Hopkins and the Depression (Syracuse, 1963)Google Scholar ; Biles, Roger, Big City Boss in Depression and War: Mayor Edward. Kelly of Chicago (DeKalh, III., 1984)Google Scholar ; , Amenta, Bold Relief, chap. 5 Google Scholar.

53. Davis was also backed by the AFL . Keller, Richard C., “Pennsylvania's New Deal,” in The New Deal: The State and Local Levels, ed. Braeman, John, Bremner, Robert H., and Brody, David (Columbus, Ohio, 1975), 45–76 Google Scholar.

54. Congressional Quarterly, Guide to U.S. Elections (Washington, D.C., 1985)Google Scholar ; idem , Congressional Elections, 1946-1996 (Washington, D.C., 1998), 170–71Google Scholar.