The recent federal corruption convictions of Sheldon Silver and Dean Skelos, longtime New York legislative leaders, have rightly led many to offer suggestions for preventing political corruption by elected officials. In two posts on this blog, Sarah suggested a mechanism for creating additional parties to make elections more competitive, and, in an earlier post, she proposed limiting New York legislators’ opportunity to take on additional employment. Others have suggested increasing legislator pay, amending campaign finance laws to close the “LLC loophole,” and increasing enforcement, including with independent ethics officers. This list is far from exhaustive.

One other “fix” that comes up again and again: term limits for legislators. Soon after the corruption scandal involving Silver and Skelos hit the news, a New York Post opinion piece called for term limits. And since Silver and Skelos were convicted, the calls have continued for term limits as part of a package of reforms (see, for example, here, here, and here). Although no one asserts that term limits are the silver bullet for ending corruption, many claim that term limits can play a constructive role as part of a comprehensive anticorruption package. But I am not convinced that term limits actually reduce the likelihood of corruption. Not only are term limits unlikely to be much help, but—as others have also argued (see here and here)—term limits might even increase corruption. Here’s why:

• First, when a legislator knows she will never again face reelection, the deterrent effect of the electoral sanction disappears, leaving fewer disincentives for a politician to engage in corruption. This paper found that the corruption was twice as prevalent among Brazilian mayors in their final term, compared to similarly situated mayors who could run for reelection. (A similar study found that increasing the maximum number of terms for Brazilian mayors from two to three was more effective in reducing corruption than increasing the mayor’s salary.) This result is actually quite intuitive. Indeed, if one believes that certain politicians are seeking office for personal enrichment, then when there is a definite end date for that person’s political career, they are likely to try to “cash in” before it is too late. Even a study arguing that term limits have positive effects on governance suggests that removing the possibility of reelection offsets term limits’ demand-reducing effect. This suggests that if states impose term limits, it will be important to balance the desire to keep entrenched legislators from becoming corrupt against the potential risk that otherwise upstanding legislators feel more comfortable engaging in corruption because they will never again face reelection.

• Second, term limits are anti-democratic, because they take from voters the power to elect the representative they choose. Jeanne recently discussed the anti-democratic implications of disqualification rules in an earlier post, and much of her analysis applies with equal force to term limits. Term limits are “overly blunt” because they sweep out even the most upstanding legislator, who also has enhanced institutional knowledge that helps the legislature operate effectively. In fact, some studies (helpfully listed in footnote 66 of this article) suggest that the decrease in institutional knowledge in the legislature that term limits impose transfers power to executive officials and professional staff–the latter a group that never faces possible electoral sanction. Thus, term limits may not only take power from voters to determine whom they want to represent them in the legislature, they may also risk transferring power (and opportunities to engage in corruption) to a less accountable group.

• Third, to the extent one is concerned that campaign finance gives special interests outsize power to influence policy, replacing long-serving members with new members might actually increase special interest influence. For example, if a long-serving member is electorally “safe,” it will be more difficult for private interests to influence her. If term limits force the incumbent legislator to retire, her replacement is unlikely to have the same ability to act counter to special interests without increased political risk. Indeed, the current system by which parties select candidates suggests that the very donors who could not “capture” the hypothetical long-serving legislator might be the people who choose the candidate who replaces her. Consistent with this, some research has found a correlation between increased campaign contributions and the adoption of term limits in Ohio. In short, term limits probably risk enhancing special interest powers that may breed the seeds of corruption.

Despite these risks, I do not mean to suggest that term limits always increase corruption; indeed, some of the research cited above suggests that term limits can be an effective piece of a comprehensive anticorruption strategy if they are implemented thoughtfully and in concert with other reforms, such as campaign finance restrictions, increased salaries for legislators, and broader enforcement powers for ethics officers. But the corruption risks associated with term limits are real, and it would be a shame if understandable disgust at recent revelations of misconduct by arrogant long-serving legislators leads to ill-considered and counterproductive reforms.