Jim Dew thinks term limits are one of the better institutions related to the presidency. He argues that power ‘corrupts,’ and therefore high turnover for elected officials is a good idea.
I disagree with this assessment. Term limits have few benefits. While I agree that political ‘corrupts,’ there is no reason to believe that ‘corruption’ increases the longer a person stays in power. If the amount of power a person holds we should expect the level of ‘corruption’ to remain the same, so there is little reason to suppose that high turnover for elected officials will decrease undesirable behavior from them. Term limits are not costless either.
In the final term of service, term limits eliminate the incentive for officials to be competent, non-corrupt and non-ideological that elections create. The instant a sitting president is re-elected, he is free from having to consider whether the public approves of him. Barry Weingast discusses this argument and gives examples in his recent EconTalk podcast.
Term limits do not have informed support. People who observe term limits in action are apparently much more likely to oppose term limits than those who do not have an opportunity to observe them in action (link).