Within the past decade, an extraordinary debate has developed in the United States regarding the merits and demerits of lengthy legislative service at all levels of government. Fueled by a growing public perception that government in general was unresponsive and that legislators in particular were out of touch with constituents and overly influenced by careerist considerations, a movement rapidly developed to limit the number of consecutive terms that could be served by members of Congress and state legislatures.

Efforts to impose term limits on members of Congress gained considerable momentum in the early 1990s while proceeding along two different tracks, one calling for state legislation to limit Congressional service, the other for a national constitutional amendment to achieve the same result. However, a derailment on the first track occurred when the U.S. Supreme Court in U.S. Term Limits, Inc. v. Thornton (1995) declared unconstitutional an Arkansas statute denying positions on the ballot to Congressional members seeking to serve more than the allowable number of consecutive years (six years for House members; 12 years for Senate members). The drive for a constitutional amendment has likewise faltered, there now being little prospect for the two-thirds majorities in Congress needed for initiation there. Ironically, the unexpected Republican successes in the 1994 Congressional elections probably contributed to that fade. Many members of the GOP had come to believe that term limits were needed to break the control of the House of Representatives enjoyed by Democrats since 1954. Their success in winning both the House and Senate in 1994--and again in 1996--undercut that reasoning while leaving them without sufficient votes to initiate the constitutional amendment for term limits that they championed.

Proponents of term limits for state legislators have been much more successful in their crusade. By early 1997, according to Karen Hansen's account in State Legislatures (June 1997), a total of 20 states had adopted some form of limits on the consecutive terms of their state legislators--occasionally by statutes enacted by the legislature itself, more often by popularly initiated laws or constitutional amendments. In a sizable number of these states, legal attacks based on the state or federal constitution have been launched with varying degrees of success. For example, California state courts found term limits imposed on Assembly members by popular initiative to be acceptable under the state constitution, but just recently a federal district judge there declared them a violation of the Fourteenth Amendment to the federal constitution. While the idea of term limits for state legislators has been most popular west of the Mississippi--where a sizable majority of states has so acted--it has had some eastern success as well, with Florida, Maine, Massachusetts, Michigan, and Ohio already committed to such restrictions.

Out of a decade of debate and contention regarding term limits, two very different perspectives have emerged, one emphasizing the need for experience and institutional memory and the dangers of too much amateurism in the legislative process, the other emphasizing the need for a constantly renewed stream of fresh ideas and approaches and the undesirable consequences of long-serving and deeply entrenched memberships. These two perspectives reflect very different assumptions about the nature and role of legislative bodies and cannot be easily reconciled. In evaluating them, however, it is helpful to have accurate knowledge of the facts regarding tenure and turnover in the representative assembly in question. This article offers such data with respect to the Virginia General Assembly, expecting them to be helpful should the issue of term limits for Assembly members become more salient. They may also help us to understand why attempts to give that issue more salience in Virginia have, to date, fared poorly.

In discussing tenure in legislative bodies, the choice of measures is crucially important. Unfortunately, one measure frequently encountered in debates over term limits is unhelpful and misleading, namely, the percentage of those members seeking re-election who win another term. Though useful for judging the competitiveness of general elections, that measure is quite useless when taken as an indication of the amount of turnover in a legislative body. Much of the drive for Congressional term limits was generated by reaction to the fact, widely reported in the late 1980s and early 1990s, that well over 90 percent of U.S. House members who seek re-election regularly win another term; in some recent years, as many as 98 percent have done so. In such reports, the frequent failure to emphasize--or even to include--the words just italicized led many people to conclude that virtually all House members continued from one term to another. If true, that indeed would have signified a worrisome dearth of new blood. But that was never the case. Though only a few of the members seeking another term were defeated, the number not seeking another term was always substantially larger, producing a biennial turnover rate in the 15-20 percent range over the past quarter-century.

The inadequacy of such a measure can be seen closer to home if we look at re-election success rates in the Virginia House of Delegates. As reported by Larry Sabato in his periodic editions of Virginia Votes, the success rate for incumbent delegates seeking re-election in recent years has been almost as high as that for the U.S. House, averaging around 92-93 percent for the 100-member House of Delegates. It was even higher in 1993, when only three incumbent delegates were defeated out of 90 who sought another term--a 97 percent success rate. But the retention rate that year was actually 85 percent, for in addition to those defeated, 10 incumbents did not seek re-election and two interim vacancies were filled by special elections, yielding 15 members who had not been in the House two years earlier. Thus, while one ought not to be complacent about the ability of incumbents to win re-election--it bespeaks a worrisome lack of competition in the electoral arena--it sheds little light on the amount of new blood entering the General Assembly.

To determine that, we need to ask how many new members take Assembly seats over the course of a single electoral cycle: two years for the House of Delegates, four years for the Senate. Some newcomers gain their seats by defeating incumbents, while others do so by winning open seats, either in the general election or in interim special elections to fill vacancies. The relative importance of each of these three electoral avenues, as well as the overall turnover rates, can be judged from Figures 1 and 2.

Figure 1, covering the House of Delegates, shows that over the seven electoral cycles from the opening of the General Assembly session in January 1984 through January 1996, a total of 95 delegates were newly elected, an average biennial turnover of 13.6 percent. Only 27 of those turnovers resulted from the defeat of incumbents; the remaining 68 stemmed from the filling of open seats caused by incumbent retirement, resignation, or death. Most (56) of those open seats were filled in general elections, but a significant number (12) involved special elections held in the interim between general elections.

Figure 2 covers turnover in the Senate over the four electoral cycles from the start of the 1984 session to the beginning of the 1996 session. It shows that 44 Senate seats changed hands, an average quadrennial turnover of 27.5 percent. In the Senate, incumbent defeat accounts for a greater proportion of turnovers than in the House. Specifically, such defeats totaled 18, with an identical number of open seats filled in interim special elections. The higher rate of turnover in the Senate compared to that of the House is, in part, an artifact of lengthier Senate terms, but the greater number of defeated incumbents makes it clear that, overall, electoral competition is keener in the upper chamber.

Useful as it is, the turnover rate does not tell us all we need to know about legislative service. Conceivably, the seats that change hands from one election cycle to another are mostly the same ones, yielding acceptable numbers of freshman entrants who nevertheless must contend with large majorities of long-serving and deeply-entrenched members. So we need some knowledge as well of the distribution of length of Assembly service. Such data, provided in Figure 3, are reassuring for the General Assembly. The House members who convened in January 1996 were rather evenly spread in terms of length of service. Only 15 percent had served 20 years or more, with another 26 percent whose service spanned 10 to 19 years. The largest category--37 percent--consisted of delegates who had served four through nine years. Those with less than four years of service--22 percent--made up the remainder.

A remarkably similar pattern prevailed in the Senate in 1996. Only 15 percent of the members had 20 or more years of service, along with 25 percent in the 10- through 19-year range. Those with four to nine years of service accounted for 35 percent of the members, and those with less than four years, the remaining 25 percent.

Judged both by the number of newcomers in each electoral cycle and by the distribution of members' length of service, the Virginia General Assembly appears to be in healthy condition. A Senate that averages a turnover of 27.5 percent of its members in each four- year period and with 60 percent of its members having served less than 10 years can hardly be said to lack new blood. While the average biennial turnover in the House is only half that of the Senate, it, too, has almost three-fifths of its members'59 percent'with less than 10 years of service.

Admittedly, the satisfactory functioning of a representative body involves far more than its mix of freshmen and longer serving members; these data tell us nothing about how well members represent their constituents, the effectiveness of legislative procedures in place, or how power is distributed and wielded. For a variety of reasons, including the fact that in both chambers seniority is often used to fill powerful positions such as committee chairmanships, longer-serving members of the Assembly have disproportionate influence. What we can say from this analysis is that contemporary patterns of tenure in the Virginia General Assembly do not make a case for term limits or other policies designed to speed up the turnover of members.

Clifton McCleskey is a professor of government and foreign affairs at the University of Virginia, where he formerly chaired the Department of Government and Foreign Affairs, directed the Institute of Government, and edited the University of Virginia News Letter. He is the author, co-author, and editor of a number of books, including his most recent one on Political Power and American Democracy, and has written numerous articles and book chapters. He has served as president of both the Southern Political Science Association and the Southwestern Political Science Association.