Measuring the ideological preferences of U.S. presidents: A proposed (extremely simple) method

Abstract

Several observations can be made concerning the data presented in the preceding tables. First, as one might expect, not all Democratic or Republican presidents are alike. As shown in Table 3, John Kennedy has a higher cumulative pro-ADA rating than does Lyndon Johnson than does Jimmy Carter. On the Republican side, Dwight Eisenhower comes in at roughly the middle of the ADA liberal/conservative ideological spectrum and considerably to the “left” of both Ronald Reagan and George Bush. Second, while there appear to be significant differences between the ideological preferences of presidents from the same party, the correlation between a president's annual or cumulative pro-ADA percentage and a dummy variable reflecting the party of the president (Democrats equal unity, Republicans equal zero) is still fairly high. For example, the correlation between the nine post-World War II presidents' cumulative pro-ADA percentages (Table 3) and a party dummy variable is 0.93. The correlation between presidents' annual pro-ADA percentages (Table 2) and a party dummy variable is 0.87. In calculating the annual correlation, President Johnson's 1963 and President Ford's 1974 pro-ADA percentages are included as separate observations (i.e., separate from President Kennedy's 1963 and President Nixon's 1974 pro-ADA records). Thus, while a dummy variable reflecting the party of a president may not be the most preferred measure of presidents' ideological preferences, such a variable may be a reasonably accurate substitute for the pro-ADA percentages reported in Tables 2 and 3. Third, as are the ratings of congressional representatives published annually by ideological watchdog groups such as the ADA, the presidential ratings reported in Tables 2 and 3 are sensitive to the issues reaching (or not reaching) a vote on the House/Senate floor. President Bush's low 1989 pro-ADA percentage, for example, may partly reflect the fact that (Congressional Quarterly Weekly Report, December 30, 1989, p. 3545): “all the 1989 tests of his [Bush's] clean-air proposal took place in committee.” President Johnson's higher pro-ADA percentages in 1964 and 1965 relative to 1966–68 partly reflect the focus of Congress on debate and passage of Johnson's Great Society programs during the early years of his administration. The extent to which the issues reaching a vote in Congress influence and are influenced by the ideological preferences of the president (at least as the preferences are measured by my proposed methodology) is an important question for future research. The correlation between the Table 2 presidential pro-ADA percentage and the mean (analogously-constructed) pro-ADA percentage for the Senate (see the Appendix) is 0.04. Finally, the presidential ADA ratings developed in this paper do not appear to evidence a bias toward the middle (i.e., a 0.5 rating) as a function of the percentage of times a president takes a position on the ADA-selected issues. One could argue, for example, that presidents who are less likely to take positions on ADA-selected issues are also less likely to want to be labeled as “extremists” and are thus more likely to position themselves as the middle of the ADA ideological spectrum. The correlation between a variable reflecting the distance (expressed in absolute value terms) between a president's annual pro-ADA percentage and 0.5 and the percentage of times the president takes a position on the issues selected by the ADA for the same year is 0.06. In conclusion, the methodology and data presented in this paper are intended to be a first step toward testing the explanatory power of the congressional dominance theory relative to other, broader conceptions of bureaucratic/regulatory policy outcomes. After developing measures of the independent ideological preferences of bureaucrats, I plan to examine a specific case of bureaucratic policymaking as a function of the interests of overseeing legislators, the president, relevant interest groups, and bureaucrats themselves. In the course of the investigation, I plan to draw on the approach developed in Kalt and Zupan (1984 and 1990) for distinguishing between the independent preferences of legislators and the interests of legislators' constituents in explanations of congressional policy outcomes. The approach to be taken in analyzing bureaucratic policy outcomes ultimately also may be useful in explaining judicial decisionmaking. For example, on an issue like abortion or civil rights, to what extent do the independent preferences of Supreme Court justices such as Scalia, Brennan, and Warren matter? By undertaking a content analysis of the editorials in several of the nation's leading newspapers written in response to Supreme Court nominations, Segal and Cover (1989) develop one possible measure of justices' independent ideological values from Earl Warren to Anthony Kennedy (1953–88). And to what extent do the positions taken by such justices reflect the interests of overseeing/appointing legislators, presidents, and relevant interest groups?