Back to the Term Limits Debate

Several years ago, I published an article in these pages arguing that term limits are not the panacea that advocates of limited government seem to think that they are and that, furthermore, they may be counterproductive to that cause. In a rare moment of disagreement, Brian Underwood and I went back and forth over the issue in a recorded mini-debate. In the three-plus years since, we have revisited it in personal conversations several times, neither able to budge the other from his position. Writing, however, has a way of clarifying positions and illuminating the relative merits of arguments. Thus, I thought it best to revisit the subject with a more empirical tone than we had previously taken, looking at the best evidence available for some of my earlier claims. As I do so, I hope that the reader will keep in mind that both sides of this debate share the same ends and primary values, so this is strictly a debate over the best means of achieving limited government and a capitalist politico-economic system.

The first point that I made in my previous writing on the subject was the end-game problem or, as I termed it then, “fewer elections, fewer consequences.” In this scenario, term limits

“establish conditions in which a considerable number of congressmen and senators know at a given time that they will not  have to soon face the will of their constituents and, for the time being, whatever relations with interest groups they have established in that time will appear all the more enticing and void of consequence.

Similarly, lobbyists would likely come to view those same departing legislators as ideal patrons on whom their efforts would be best spent. Thus, instead of a crisis in which long-serving legislators have built networks of favors and pull over time, we would likely face one in which the costs to lawmakers of accommodating lobbyists falls considerably.”

I allowed that legislators would perhaps have some concern for their party’s prospects, thus limiting them to some extent in their behavior, but I didn’t estimate this to be sufficient to deter them from self-serving behavior. So, what is the evidence? As it turns out, this is one of the most common arguments against term limits. The empirical evidence, however, is very mixed, with some studies finding evidence of end-game problems and others finding evidence of better responsiveness and performance. One doctoral dissertation in the economics department at George Mason University found that the overall net effect of term limits in states that adopted them had been an increase in the size of government in those states. Erler (2007) argues as much, looking at states with term limits and those without and finding that those with term limits have notably higher government spending levels than those without—hardly the result that their advocates intended.

A larger state does not, however, translate to a stronger legislature. “Larger” may even be interpreted in many cases in the literature as “more executively driven.” An academic literature that emerged in the early 1990’s, when discussion of term limits skyrocketed to the forefront, was speculative but with certain prevailing lines of argument. Rosenthal (1992) and Beyle (1992) argued that state legislative term limits enhance the power of governors. Rosenthal (1992) and Capell (1993) echoed the concern that diminished legislative experience strengthens legislative staff, administrative agencies, and lobbyists, who then become the more tenured parties at state capitols. Empirical studies have corroborated some of their predictions, though.

A collaborative effort between scholars scattered across the United States, the Joint Project on Term Limits (JPTL), examined and compiled evidence from separate state legislatures beginning in 2000. Though in many cases they found little to no significant effect of imposing term limits, where effects were observed they often noted that the net effect of term limits in state legislatures had been to deprive legislatures of institutional memory and valued leaders, making legislatures weaker institutions, “creat[ing] chaos” when as many as 50 percent of legislators at a time can be incoming freshmen.

In comparing the professional legislatures of California, Illinois, and Ohio (CA and OH being term-limited, Illinois not being so), they found that party leaders appeared in Ohio to have lost power relative to rank-and-file legislators and that the committee system lost power relative to the general assemblies in California and Ohio. Non-partisan staff lost power, though this effect was considered ambiguous and possibly related to other changes. Lobbyists are reported to have gained influence after term limits were imposed in California but not in Ohio, making the effect again imperfect. Partisanship was found to have increased in all three states, making the effect of term-limits unclear on this dimension. Legislators’ civility, however, was observed to have declined in the term-limited states as legislators more frequently encountered members of the opposition party with whom they had developed no lasting working relationship (Farmer and Green, 2007). Thus, the only decisive effects in professionalized states appear to have been more turnover and less civility, with weaker evidence for a loss of power suffered party leaders, the committee system, non-partisan staff and for an increase in lobbyists’ power.

When Farmer and Green turn their attention to semi-professionalized state legislatures, they use Arizona and Colorado as their test cases and Indiana as a non-term-limited control. There, they observed increased turnover in the two limited states but not in Indiana. They observed an increase in “career-oriented members.” This was present in Indiana as well, so the effects cannot be deemed clearly the result of term limits, but it contradicts the typical line of term limits being designed to weed out “career politicians.” Most significantly, they observed an increase in gubernatorial power in Colorado that they considered to have likely been the direct result of term limits.

And in citizen legislatures, comparing term-limited Arkansas and Maine to non-term-limited Kansas, they again found clear effects. The number of female legislators grew in Arkansas, confirming theories that suggested that term limits might benefit women. Competition in that state is said to have increased, though with this effect being contrary to the trend in other states Farmer and Green hesitate to consider it the direct result of term limits. More assuredly, leadership was seen to have lost power in Maine but gained it in Arkansas, with no change observed in Kansas. Committee power again decreased, non-partisan staff power increased in Arkansas and Maine, and lobbyists lost power in both term-limited states. These results thus stand in contrast to the effects of term limits on professional and semi-professional legislatures. Both partisanship and civility are found to have decreased and gubernatorial power to have increased in term-limited states.

One overarching effect that the authors note is the way in which term limits accelerate existing trends. If there is a nascent trend toward caucus-based elections, term limits will accelerate it, expediting institutional change. Advocates of term limits would thus be well served to ask whether existing institutional trends—even nascent ones—are desirable and include that among their cost-benefit calculations.

“Many effects were common to all termed states,” Farmer and Green write. “Increased turnover among members was consistent across all levels of professionalism… Demographic changes were inconsistent…Campaigns were more caucus centered in all three professionalized states and more expensive in all three semi-professionalized states. Leadership power declined in four of the six termed states… Committees experienced decreased power in all six term limited states. Non-partisan staff lost power in all three professionalized states. However, they had an increase in power in the four less professionalized termed legislatures… The results for lobbyists were mixed… [in all six termed states] [c]ivility is decreasing as partisanship increases… The only consistent change in gubernatorial power was the increase found in the citizen legislatures” (Farmer and Green, 2007).

Thus, overall, the scorecard is mixed. More caucus-based elections and increased partisanship likely mean more power to parties and party leaders rather than less. This, however, is counterbalanced by diminished power to elected party leadership in the legislature. Depending on how one views committees, their reduction in power could be viewed as good or bad, but there is reason to believe that reduced committee power might mean reduced rent-seeking legislation. Non-partisan staff and lobbyists appear to win some and lose some. Lobbyists lobby both for and against taxes, subsidies, and protectionism, so the effects of those changes in any one case are ambiguous from the standpoint of wanting to reduce government control of the economy. However, it contradicts the standard argument for term limits as a decisive roadblock against lobbyist-legislator relationships. The claim that term limits reduce “career-oriented politicians” is likewise contradicted by observations in Arizona and Colorado but is perhaps mitigated in the longer term by turnover effects. Assessed in full, it is unsurprising that Weissert and Halperin (2007) find that states with term limits show markedly less enthusiasm about them, with knowledgeable but disinterested parties often wanting to repeal them on the grounds that they have allegedly weakened the institutional features of a state’s legislature.

A few arguments remain unaddressed by these findings and deserve mention. The first is that incumbency advantage is simply so great that political competition without term limits is unable to combat the effects of long-serving legislators’ complacency, cozy relationships with special interests, and war chests.

On the last of these points, the evidence for spending being a decisive factor in political campaigns is mixed, and it is often the case that the candidates who spend more are those who lose. As to cozy relationships with lobbyists and special interests, as I noted earlier, the evidence as to whether the performance-enhancing effects of term limits dominate the end-game problem is unclear. As to complacency or legislative drift from voters’ demands, that seems to be less a function of tenure than it is of the particular point in the election cycle at which legislators are casting their votes. Levitt (1996) presents empirical evidence that “senators alter their voting patterns as elections approach to better reflect the preferences of the median voter. Senators apparently consider voters to be myopic since most of the change in voting patterns is concentrated in the election year itself.”

He finds that this shift toward the median voter’s preferences comes at the expense of the party line. This suggests that party leaders probably give senators a bit more leeway in election years to deviate from the party line in order to secure reelection. He also finds that first-term senators place significantly more weight on the preferences of voters in their states and on their core support constituency, sacrificing some of their commitment to the party line and personal ideology. Levitt describes the implications for term limits as follows: “Term limits will have two direct effects: (i) increasing the number of first-term senators, and (ii) decreasing the number of standing senators seeking reelection. According to the estimates, those two factors will largely counterbalance since first-term senators and those seeking reelection exhibit similar voting patterns.” Thus, the net effect in terms of legislators’ accountability to voters is likely to be miniscule or even zero. In a survey of all 50 states, Carey et al (2006) actually find the accountability effect of term limits to be negative, reducing politicians’ adherence to constituents’ preferences. That finding may be explained by voters’ own diminished involvement after term limits are imposed; Nalder (2007) find that, contrary to term-limit-advocates’ predictions, voter turnout is notably diminished after term limits are imposed.

The second argument worth addressing is also well introduced by Levitt, who attributes fifty to seventy percent of senators’ votes in his sample to senators’ individual ideologies, as opposed to 10-13% to voter preferences and anywhere from two to twenty-five percent to party line. This reaffirms a point that I initially made in the last article: on an overall basis, nothing compares to the effect of culture and ideology in influencing politics. The effects of term limits, though we can all cite some evidence in favor of them or in opposition, are thus likely to be in all cases dominated by the dominant ideology in America and especially on the American right. What’s more: so long as the redistributive and regulatory power of the state remains as it is and continues to be run by individuals with no moral objection to its current conduct, the effects of minor institutional changes such as term limits are likely to only demonstrate the creativity of rent-seekers in developing new ways to obtain undeserved privileges

Lastly, on the point of ideology, I maintain that for anyone who saw the trend in the Republican Party from 2010 to 2015 as a beneficial shift away from a party-driven status quo and towards genuine free market principles on the right, it should be recognized that the politicians who worked to promote that—Rand Paul, Ted Cruz, Mike Lee, Justin Amash, etc.—would as surely be ousted when their limits come due as those “Establishment” Republicans who are the intended targets. Given that they are a small minority, to see term limits as a likely net benefit one would have to assume that the next crop of legislators to be brought in after they are term-limited out would be just as good if not better. Given the history of animosity between some of those legislators named and party leaders as well as the efforts by GOP leaders to run an “Establishment” primary challenger against Amash for voting his ideology rather than the party line, a great deal of optimism is required to expect the ex post results to be satisfactory. I believe that the rapid shift of the GOP away from those values now that they have secured a majority of both houses and the White House gives us reason to be all the more protective of the few limited government legislators in Congress today and to not support institutional changes that are far more likely to revert to the ideological views of the Republican majority.

This is only a sampling of the arguments suggesting that term limits are not likely to be beneficial to the cause of limited government and may actually lead to weaker legislatures and more rapid growth of government. Ultimately, I continue to see them as, at the very best, having mixed results and, at worst, promoting the growth of government through a weakened legislature. I believe that short of limiting the power of the state in our economy, institutional changes such as term limits will only increase turnover and change the rules of legislating without changing the state’s footprint on society. I beieve that were term limits to be implemented the nascent core of good, pro-market legislators out there would likely be the first to expire and be replaced by the kind of legislators who constitute the norm in both state and federal governments.

Finally, I will remind the reader that term limits are ultimately less a restriction on legislators (who would be just as powerful for their short tenures in office as they are for long ones) than they are on voters. When legislators truly deviate from a majority of their constituents’ preferences to such an extent that voters deem it worthy to do so, they vote politicians out of office. So long as they have not chosen to do so, we must interpret the actions of their elected officials as being within the realm of behavior considered acceptable by their constituents. If we do not like voters’ choices with respect to their politicians, then our only option is to influence the culture through philosophy and a better understanding of economics and the workings of capitalism. Until that is achieved, we cannot assume that the next person they elect will be any better than the current one.

Beyle, Thad L. “Term Limits in the State Executive Branch” In Limiting Legislative Terms, eds. Gerald Benjamin and Michael J. Malbin. Washington: CQ Press (1992).

Capell, Elizabeth. “The Impact of Term Limits on the California Legislature: An Interest Group Perspective.” Peper presented at the annual meeting of the Western Political Science Association, Pasadena, CA.

Carey, John M., Richard G. Niemi, Lynda W. Powell, and Gary F. Moncrief. “The Effects of Term Limits on State Legislatures: A New Survey of the 50 States.” Legislative Studies Quarterly 31, no. 1 (2006): 105-34. http://www.jstor.org/stable/40263375.

Erler, H. Abbie. “Legislative Term Limits and State Spending.” Public Choice 133, no. 3/4 (2007): 479-94. http://www.jstor.org/stable/27698198.

Farmer, Rick and John C. Green. “Introduction: Accelerating Change with Term Limits.” from Legislating Without Experience: Case Studies in State Legislative Term Limits. Eds. Farmer, Rick, and Christopher Z. Mooney, Richard J. Powell, and John C. Green. Lanham: Lexington Books (2007).

Levitt, Steven D. “How Do Senators Vote? Disentangling the Role of Voter Preferences, Party Affiliation, and Senator Ideology.” The American Economic Review 86, no. 3 (1996): 425-41. http://www.jstor.org/stable/2118205.

Nalder, Kimberly. “The Effect of State Legislative Term Limits on Voter Turnout.” State Politics & Policy Quarterly 7, no. 2 (2007): 187-210. http://www.jstor.org/stable/40421578.

Rosenthal, Alan. “The Effects of Term Limits on Legislatures: A Comment.” In Limiting Legislative Terms, eds. Gerald Benjamin and Michael J. Malbin. Washington: CQ Press (1992).

Weissert, Carol S., and Karen Halperin. “The Paradox of Term Limit Support: To Know Them Is NOT to Love Them.” Political Research Quarterly 60, no. 3 (2007): 516-30. http://www.jstor.org/stable/4623849.

Leave a Reply

Your email address will not be published. Required fields are marked *

%d bloggers like this: