A term limit is a statutory or constitutional restriction on the number of terms that an individual is allowed to hold a particular elected office. The term limit may or may not be grandfathered. Full grandfathering allows officeholders at the time that term limits are enacted to be exempt from term limits, whereas limited grandfathering allows such officeholders to be treated as if they have previously served no terms. Some term limits allow an individual who is forced out of office by term limits to hold the same office again after a specified period of time has elapsed while many do not. In recent years there has been heated debate over the desirability of term limits.
The concept of a term limit goes back at least to Aristotle, who praised the merits of a legislature of citizens with each holding office for a short time. This frequent succession of citizens in an office is referred to as rotation in office. Term limits were not uncommon in colonial America and, although not included in the United States Constitution, were debated at length during the Constitutional Convention. Indeed, George Washington was widely praised for retiring from the presidency after two terms. While voluntary rotation in office was commonplace in the U.S. Congress until the late nineteenth century, Polsby (1968) documents the rise of the career congressman and the institutionalization of the Congress that took place thereafter. Not only did the average number of terms served by congressmen steadily increase over time, but the Congress itself developed an increasingly complex committee system accompanied by growth in congressional staff and the rise of the seniority system.
Although today the prevalence of professional politicians indicates at best a modicum of enthusiasm on their part for rotation in office, term limits do exist for many state and municipal elected offices. Most states have term limits for governor. Term limits also have been imposed in a large number of states and municipalities, many in the 1990s, on the offices of state representative, mayor, and city councilman. At the federal level the twenty-second amendment to the U.S. Constitution limits a president to two elected terms.
The two elected federal offices that are notably free of term limits are U.S. senator and U.S. representative, and it is the question of the imposition of term limits for these offices that has generated the most controversy in recent years. In the early 1990s twenty-three states passed voter referenda or laws limiting the number of terms that their congressmen could serve. These referenda and laws subsequently were overturned by the U.S. Supreme Court in 1995 in U.S. Term Limits, Inc. v. Thornton. Although there is little question that public support for term limits exists, it is not obvious that this support is an explicit cry for term limits or an expression of frustration with the way that government has been operating.
The focus of debate in recent years has been on the likely effects of term limits imposed on congressmen. Some important issues raised include the impacts of term limits on the composition of the Congress, the type of legislator who will serve, the degree to which legislators will faithfully represent their constituents' interests, and the level of federal government expenditures or the size of government. The issues are not necessarily independent of each other. Analysis is difficult because the absence of term limits on congressmen throughout the history of the Congress means that there are no data on the behavior of congressmen under term limits. Consequently, analyses are normative, theoretical, or indirect when empirical.
Legislative tenure almost certainly will be reduced if term limits are imposed. Reed and Schansberg (1994) estimate that a six-term limit for congressmen will reduce the expected number of terms served by a congressman from 8.9 to 3.2 with the maximum being 6.0. Furthermore, there initially will be large influxes of freshman congressmen every twelve years (six terms), but the magnitudes of these influxes will dampen over time. There will be times after the imposition of term limits when the Congress will have historically high percentages of inexperienced congressmen.
It should be noted that these estimates of the effect of term limits on expected tenure are based on historical continuation rates (for the period 1985 through 1991) for congressman and do not take into account how the term limits themselves can alter behavior and therefore continuation rates. Grofman and Sutherland (1996) argue that strong challengers may postpone running for office until term limits force an open-seat election, while Francis and Kenny (1997) argue that term limits in a higher office (e.g., the U.S. Senate) may encourage more holders of lower office (e.g., the U.S. House) to run for higher office by generating more open-seat elections for higher office. More generally, the probability of reelection depends on the degree to which congressmen serve their constituents' interests, and term limits may alter congressmen's incentives to serve those interests.
A related implication of term limits is the likely demise of the seniority system, under which committee chairmanships are assigned on the basis of length of tenure in office, because the number of congressmen with the maximum possible seniority may well exceed the number of committee chairmanships. It is not clear what would replace the seniority system.
Finally, there are partisan implications of term limits. Gilmour and Rothstein (1994) conclude that term limits will impact the partisan composition of the Congress by favoring Republicans. This is because in recent years Democratic incumbents have had lower retirement rates and higher reelection rates than Republican incumbents and Republican challengers have been more likely to win open-seat elections than Democratic challengers following the retirement of incumbents from their respective parties. The relative magnitudes of the parties retirement rates, reelection rates, and open-seat election rates taken together favor Republicans under the forced-retirement regime of term limits.
Term limits will impose the rotation in office that had once been voluntary. It has been argued (Petracca, 1996) that rotation in office guards against concentration of political power in the hands of career legislators and professional politicians and is consistent with the belief that in a democracy governance should be carried out by amateurs who shortly return to their lives as private citizens. Presumably such citizen-legislators, understanding that they are free from the task of continually seeking reelection and that they shortly will return to the private sector to live under the rules contained in their own legislation, would be more inclined to represent the true interests of their constituents and to not pursue their private interests and would be less susceptible to influence from special interest groups and their promises of campaign contributions.
The logic of this argument is not convincing. Why should a mandated limited time in office, effectively the constraint that it is not possible to make a career of that office, cause these citizen-legislators to better represent their constituents interests and to not pursue private gain? This is a particularly relevant question because the threat of not being reelected constrains legislators to take the interests of unorganized constituents into account (Denzau and Munger, 1986). In order to make the claim that term limits will produce a legislature of citizen-legislators more inclined to represent constituent interests it would be necessary to make a convincing argument that term limits will influence the self-selection of citizen-candidates in this manner. No such argument has yet been made. Indeed, it has been argued that the interruption of a private sector career in order to serve a brief time in office under term limits may attract a higher proportion of independently wealthy individuals and opportunists than does the current system without term limits (Polsby, 1991).
It also is not clear that term limits actually will lead to a legislature of amateurs, let alone amateurs who efficiently and faithfully represent their constituents. It is more likely that term limits will not discourage politically ambitious people from running for office but will instead channel their ambitions in different directions. In lieu of the prospect of a potentially long career in Congress, the congressman will consider how to optimally manage a career that may include a succession of different political offices or periodic stints working in the private sector, perhaps as a lobbyist for firms with interests before the Congress (Garrett, 1996). While the congressman's decision-making calculus will give weight to the interests of his constituents, it necessarily will give weight to the interests of his potential future constituents or private-sector employers as well. Furthermore, the limited time horizon in office under term limits implies that the congressman has both less time to learn and less incentive to invest time in learning the substance of the issues being considered by the Congress because a shorter time horizon means a shorter time to collect the returns from such an investment.
There has been considerable debate on whether and to what extent term limits will alter the level of shirking by congressman. “Shirking” refers to a congressman's failing to fully represent the interests of his constituents. A congressman whose interests do not coincide with his constituents' interests has an incentive to shirk. The threat of not being reelected by the voters because of this shirking acts to reduce shirking. However, a “last-period problem” arises when a congressman, knowing that he is in his last term in office and therefore does not have to face reelection, can costlessly shirk as much as he likes and will do so if his interests do not coincide with his constituents' interests. A last-period problem does not arise for a congressman who is defeated for reelection because this congressman is not aware that he has served his last term in office until after defeat, whereas a potential last-period problem does arise for every congressman who retires from office. If voters can detect shirking earlier in a congressman's career, then it is likely that the last-period problem will not be important because congressmen who are inclined to shirk will be defeated for reelection before their voluntarily leaving office. Studies by Lott (1987, 1990) and Van Beek (1991) indicate little difference between congressmen's voting on legislative bills in their last term before retirement and their voting on legislative bills in their previous terms served, thereby suggesting that voters do tend to remove shirking congressmen from office prior to voluntary retirement.
Term limits, which would force many congressmen to leave office prior to electoral defeat, guarantee that the last-period problem will arise more frequently. The relatively short three-term and six-term limits that have been suggested for congressmen likely will cause the degree of shirking associated with the last-period problem to increase nontrivially by not giving voters enough time to remove shirking congressmen prior to their last terms before forced retirement under term limits. These term limits also will prematurely remove from office those congressmen who are inclined to shirk little and whom the voters would like to retain in office.
Focusing only on the last period is, however, an oversimplification. Although shirking by a congressman will be greatest in the last period because the reelection constraint is no longer binding, shirking likely will occur throughout a congressman's career.
Dick and Lott (1993) argue that tenure itself can be a source of shirking. A congressman with longer tenure has greater seniority on committees and likely has developed greater ability to utilize the rules of the Congress and greater skill at logrolling. These factors will increase the ability of the congressman to transfer wealth to his district from the rest of the country relative to a congressman with shorter tenure. Voters are therefore reluctant to remove from office a congressman with relatively long tenure. This allows the congressman to shirk with respect to other interests of the voters without unduly risking electoral defeat. The longer a congressman's tenure, the greater is the potential amount of shirking by a congressman without endangering reelection. By forcing all congressmen to leave office after a specified number of terms, term limits make it less costly for voters to discipline their congressmen at the polls. Dick and Lott believe that the resulting reduction in shirking will more than offset the increase in shirking generated by the last-period problem occurring more frequently under term limits.
While Dick and Lott make valid points that shirking can increase with tenure because the ability to transfer wealth increases with tenure and that term limits can reduce shirking from this source, their contention that any increase in shirking induced by term limits will be only in the form of more last-period shirking reflects too narrow a focus. Bender et al. (2001) argue that given a finite time horizon in office a congressman will optimally shirk throughout his congressional career (i.e., there is a time path of optimal shirking). In this dynamic context optimal shirking means that in each term the congressman's degree of shirking reflects a tradeoff between the benefit of additional utility from additional shirking in the current term and the cost of the reduction in expected present value of the utility received in future terms from the reduction in probability of reelection at the end of the current term induced by additional shirking in the current term. The less time remaining in the horizon, the smaller is the cost of electoral defeat and therefore the larger is the optimal amount of shirking. Term limits reduce the time horizon of the congressman and therefore raise his optimal amounts of shirking in each of his terms in office. The threat of electoral defeat becomes less effective in controlling shirking when term limits are present because the return to not shirking is diminished by the reduced time horizon under term limits. This impact on shirking is magnified to the extent that the reduced return to lower shirking causes the pool of potential candidates to consist of candidates who are on average more inclined to shirk. Term limits not only increase shirking by existing congressmen but also encourage potential shirkers to run for office.
Over the years both the average tenure of congressmen and the level of government expenditures, a proxy for the size of government, have been increasing. Some proponents of term limits argue that they will reduce the size of government based on a presumed positive and causal relationship between tenure and spending. It has been hypothesized that the longer that congressmen are in office the more likely they are to be drawn into a culture of spending or shirk more by supporting legislation for more spending than their constituents desire or develop more expertise at logrolling. However, Reed et al. (1998) test these hypotheses and find at best weak evidence that a relationship between congressional tenure and support of government spending exists and consequently little empirical support for the proposition that term limits will reduce the size of government.
The list of issues above is not exhaustive. Certainly the impact of term limits on the influence of lobbyists and on the relative power of the executive and legislative branches of government could be added to the list. The unavailability of data on the behavior of congressmen under a term-limits regime precludes direct empirical testing of all of the issues. This inability to test directly the hypotheses about the impacts of term limits may continue to leave the desirability of term limits an open question.