Retrospective considerations do influence vote choice. But the manner in which voters incorporate retrospective evaluations is still an empirical question. This essay compares two voting rules: a traditional retrospective rule and a retrospective-prospective rule. The rules are differentiated based on the information costs associated with each. It is assumed that the traditional retrospective rule is less costly since voters need only make evaluations about the incumbent, whereas using a retrospective-prospective rule, voters must acquire additional information about the challenger. This essay argues that a choice of voting rule is a function of whether voters perceive the electoral environment as being stable or unstable (in terms of the issue positions candidates adopt over time). Laboratory experiments are conducted to test this proposition. In general, the results show that voters show a greater tendency to rely on a traditional retrospective rule when they perceive a stable electoral environment.
