Why are political leaders often attacked by their ideological allies? The paper addresses this puzzle by presenting a model in which the conflict between the incumbent and his allies is ideological, dissent is electorally costly, and voters are learning about their own policy preferences over time. Here, by dissenting against the incumbent (and thereby harming the party in the upcoming election), the allies can change his incentives to choose more or less extreme policies, which affects the amount of voter learning. This induces a trade-off between winning the current election and inducing the party leadership to pursue the allies’ all-things-considered more-preferred policy. Optimally balancing this trade-off sometimes involves active dissent that damages the party in the short-run. In equilibrium dissent arises precisely because it is electorally costly.
Do ideologically motivated parties have strategic incentives to lose? I present a model of repeated spatial elections in which the voters face uncertainty about their preferred policy and learn via experience upon observing their payoff realization. The amount of voter learning, I show, depends on the location of the implemented policy: the more extreme the policy is, the more information is generated. This, in turn, creates a trade-off for a party whose ideological stance is unpopular with the electorate, between winning the upcoming election so as to secure policy influence, and changing the voters’ preferences so as to win with a better platform in the future. Under some conditions the party gambles on the future: chooses to lose today, in order to change voters’ views and win big tomorrow.
What happens when voters become better informed about their representatives’ actions or performance? The empirical literature reports mixed findings on incumbents’ electoral fortunes for both good and bad news. We introduce a political agency model, meant to inform empirical research, in which voters are fully rational, unconstrained, and unbiased. We show that researchers should not expect information campaigns to always work or produce similar effects. Identical interventions in similar contexts can give different results. The same intervention measured in different ways can give different results. The analysis also highlights that the comparability of empirical estimates is non-monotonic in the geographical spread of observations. We offer recommendations to improve the accumulation of knowledge. By constructing variables related to governance outcomes and interacting them with treatment status in their regression analyses, researchers can recover fully comparable estimates of information treatment effects. Further, to avoid attenuation bias, good and bad news should be defined in absolute, not relative terms.
Do the right candidates for office choose to run at the right time? I analyze a model of repeated elections in which politicians differ in the probability of being competent. Voters update their beliefs about the office holder's ability upon observing his performance in office. In each period, the country faces either a safe situation or a crisis. A crisis has two key features: it exacerbates the importance of the office holder's competence and, as a consequence, the informativeness of his performance. I show that electoral accountability has the perverse consequence of discouraging good candidates from running in times of crisis. Precisely when the voter would need him the most, the politician who is most likely to be competent chooses to stay out of the race in order preserve his electoral capital. In contrast with results in the existing literature, this adverse selection emerges even if running is costless and if office is more valuable than the outside option.