Determinants of presidential position taking in Congress, 1949–1995

Abstract

In this paper we seek to explain the activity presidential position taking on roll call votes in the House. Position taking may help presidents pursue their policy agendas, but time and available resources constrain their ability to take positions. Even though position taking is a discretionary presidential action, it occurs in the legislative arena and, thus, presidents must consider elements in both institutions as well as outside conditions when making this decision. Accordingly, we posit a multiple perspectives approach to explain the number of presidential positions on votes in the House. Our multivariate two-stage least squares regression reveals that variables from all three environments (executive, legislative and exogenous) within our multiple perspectives approach are necessary to adequately explain presidential position taking overall and when divided according to domestic and foreign policy positions.

Acknowledgements

We wish to thank David A. Freeman, Editor, anonymous reviewers, and George C. Edwards III for comments on an earlier version of this paper. We are also grateful to Lyn Ragsdale for providing some of the data used.

Notes

1 Neustadt (1960, p. 65) does not believe that the Gallup measure adequately captures public prestige.

2 This paper is an extension of a larger project that provides fuller justification for the variables incorporated in the three environments. The authors will provide any desired information to interested readers.

3 Reasons exist why data for the Senate were not collected and used. In the upper chamber, many votes are non-germane amendments and their inclusion would be problematic as policy formulation. This difference makes similar analysis between the two chambers problematic.

4 Since Lyn Ragsdale (1996) does not provide the individual votes themselves categorized by policy area, caution must be exercised in interpreting these frequencies due to the differing number of votes by issue area for any given year. We combine her foreign aid, foreign trade, and defense into foreign and her economic management, civil rights-social welfare, resources, and agriculture into domestic.

5 Jack knifing provides average coefficient estimates for a series of regression coefficients that successively omit a different observation, thereby indicating the robustness of the coefficients across cases.