Timothy R. Johnson and Jason M. Roberts
difficult time securing confirmation and will therefore be more likely to make public statements that support the nominee.
Second, to test whether the ideological distance between the Senate and the current Court (based on the eight remaining justices when a vacancy occurs) affects the president’s decision to go public, we calculate the absolute value of the difference between the filibuster pivot’s ideology and the ideology of the Court median. Finally, we calculate the absolute value of the distance between the pres- ident’s own ideology and the ideological score of the Senate filibuster pivot. To determine this distance we utilize DW-NOMINATE scores for the president and the pivotal Senator in the year that a nomination takes place.21 These variables should have significant and positive relationships with a president’s decision to pressure the Senate publicly during the confirmation process.
Beyond the spatial measures, we include two necessary control variables. First, we must consider whether the president’s public role in the confirmation process has changed over time. When we ran a GEC model for the president’s propen- sity to pressure the Senate publicly on year, we find a strong time trend. Con- trolling for no other factors, the year variable predicts that President Truman would make an average of 2.09 statements per nominee, whereas it predicts an average of 20.25 statements per nominee for President Clinton. Our model there- fore includes a measure for time as a control variable that is coded as the year of nomination. Second, the time period between a nomination and Senate action varies as well. Some nominations simply take longer than others due to schedul- ing problems or controversy. Thus, we include the natural log of the time (in days) between a nomination and Senate action to account for the theoretical maximum number of statements a president could make (see King 1988).
In general, we expect the president to go public when he is ideologically con- strained by the Senate. At the same time, we must also account for whether strong presidents publicly pressure the Senate at the same rate as weak presidents. To test whether a difference exists, we include three additional controls in the model that account for political resources held by the president at the time a nomina-
and Rosenthal (1997) for further details), whereas the ADA selects only 20 “key votes” for each calendar year. While the two sets of scores are highly correlated (r > .9 for most Congresses), DW- NOMINATE scores are less sensitive to missed votes and strategic manipulation. By design, the ADA chooses votes that will separate perceived “liberals” from “conservatives,” which Arnold (1990, 82) notes can lead to cases where legislators intentionally vote against the ADA position to ensure that they do not receive a “perfect” ADA score, which might attract potential challengers. The DW-NOMINATE procedure clearly precludes this type of behavior from severely contaminating the analysis. Note that using ADA scores rather than DW-NOMINATE does not alter any of our substantive findings.
21 Note that our hypotheses about these ideological distance variables are conditional on the nominee not being in the Senate’s winset. That is, if a president nominates someone who will move the Court in the direction of the Senate then the Senate should confirm the nominee without presi- dential prodding. However, the only nominee in our data that is in the Senate winset is Stephen Breyer—and he sits on the edge of the winset. We elected to include Breyer in the analysis for the sake of completeness, but the substantive results of the model do not change if he is excluded.