Term Limits and Parliamentary Behaviour: Evidence from a Natural Experiment in Chile

Alonso Román Amarales

Abstract

The debate over whether term limits for elected representatives achieve their intended outcomes has persisted for decades. Scholars have studied their effects on legislative behavior and parliamentary composition, yet data constraints and inadequate contexts have limited conclusive findings. This paper addresses these challenges by leveraging a natural experiment in Chile’s Congress in 2020, where an unexpected reform imposed immediate reelection limits, disqualifying 51 of 198 legislators. Using a difference-in-differences approach, the study tests four hypotheses regarding term limits’ effects: attendance at roll-call votes, the number of law proposals, maverick voting, and bipartisan collaboration. The results indicate that term limits do not significantly alter legislative behavior overall. However, disaggregated findings reveal heterogeneous effects across chambers. In the lower chamber, term limits reduce legislative productivity, with a 16% decline in law proposals and a 7% decrease in bipartisan initiatives. This suggests that term limits foster parliamentary disengagement, incentivizing legislators to be less productive and less inclined to collaborate across party lines. No significant effects are observed in the upper chamber. These findings challenge the expectations of term-limit proponents, who anticipate greater legislative efficiency and independence, while alleviating concerns about maverick behavior or dysfunction. By analyzing this reform in a national Congress—where electoral incentives and accountability mechanisms are stronger—this study offers new causal evidence on term limits’ consequences. It contributes to understanding term limits’ effects in non-consolidated democracies, where such reforms are frequently proposed as tools to improve legislative behavior.