Existing theories of congressional retirement have developed in the context of the House of Representatives and thus do not consider important differences between the Senate and the House. Our empirical analysis of Senate retirement from 1962 to 2000 demonstrates that retirement decisions of Senators are not affected by the same things that affect their House colleagues' decisions; electoral safety and the value of formal institutional positions influence Senators less than House members. Senate retirement decisions are affected by the age of the member and majority-- party status. We discuss implications of our results for the Senate's operation and place in the constitutional system.
When considering membership change in Congress, most attention is focused on election contests. The spectacle of two candidates campaigning tirelessly for office, combined with the potential drama of surprising wins and losses, draws more interest than decisions of members to retire. Nonetheless, the greatest source of membership change in Congress is not electoral defeat but instead is voluntary departure due to retirement or seeking higher office (Ornstein, Mann, and Malbin 2000). Given this, the importance of studying retirement becomes clear.
White we know much about the individual-level motives behind retirements in the U.S. House, research on career decisions in the Senate is sparse. Given the importance of retirement in determining the Senate's membership, it is surprising that so little research has been done on individual-level determinants of Senate retirements (see Livingston and Friedman 1993 for an exception). In this study, we attempt to remedy this deficiency by studying retirement from the Senate between 1962 and 2000. We argue that much of our understanding of legislative retirement developed in the context of the House. While this literature does well in explaining House retirements, its explanations do not extend to the Senate. Significant differences between the House and Senate require further exploration of the determinants of Senate retirement.
We begin with a discussion of the importance of the study of legislative retirement and with a brief review of the differences in House and Senate retirement rates. Next, we review the literature on House retirements. We then evaluate the significant House-Senate differences that cast doubt on the utility of the House retirement literature for our present purposes. Following this, we suggest hypotheses for the determinants of Senate retirement and empirically evaluate such a model. We conclude with a discussion of some of the implications of our findings.
THE IMPORTANCE OF LEGISLATIVE RETIREMENT
Legislative retirement matters. First and foremost, individual-level retirement affects congressional policymaking. Since most members of Congress remain stable in their policy preferences during the course of their careers (Asher and Weisberg 1978), membership change becomes a clear path to policy change in Congress (along with change in party leadership or in the executive). Moreover, retirement provides a significant source of competition in congressional elections. When retirement occurs, competition for the seat increases (Jacobson 1997 on the House; Abramowitz and Segal 1992 on the Senate). Thus, in years with more retirements, we will see more competitive elections for the vacated seats than would have occurred had the incumbents sought reelection.
Understanding retirement and its determinants also tells us about what members value in their congressional service and how the institutional design of a legislature affects their satisfaction. For example, Hibbing's (1982b) explanation of higher House retirement rates in the 1970s rests on the decade's weakened seniority norm; members chose to leave when they learned that their years of service would no longer automatically translate into committee chairships. This tells us that members value guarantees of formal positions of influence and find congressional service less satisfying when they do not receive them. …