This study aims to reveal how an endogenous change in political trust affects the performance of a representative democracy. To this end, we construct a twoperiod political agency model wherein voters face uncertainty about the distribution of politicians' types (model uncertainty) as well as each individual politician's type. Such model uncertainty allows political trust to endogenously change over time, whereas political trust is invariant without model uncertainty. We show that model uncertainty substantially increases corruption. Furthermore, it generates self-fulfilling multiple equilibria: a high-accountability equilibrium and a lowaccountability equilibrium coexist. In countries experiencing democracy only for a short time, model uncertainty would be severe. Our results indicate that democratic performance tends to be low and even similar countries could experience different performances depending on citizens' expectations in such new democracies. By extending the model, we also discuss the relationship between political trust and the rise of an outsider candidate.

No abstract

This paper aims to investigate the possibility that electoral campaigning transmits truthful information in a situation where campaigning has a direct persuasive effect on a subset of the electorate called “naïve voters.” To this end, we construct a multi-sender signaling game in which an incumbent and a challenger decide whether to focus on policy or ability in electoral campaigning, and a media outlet then decides whether to gather news. Voters are divided into sophisticated and naïve voters. We demonstrate that a candidate's strategy regarding their issues of focus (campaign messages) can signal his or her private information. Specifically, negative campaigning against the incumbent's ability signals the incumbent's low ability in all separating equilibria. It is also noteworthy that separating equilibria exist only when sophisticated and naïve voters coexist. This implies that a fraction of naïve voters has a non-monotonic effect on the possibility of transmitting truthful information.

We study how politicians' reputation concerns affect taxation in the presence of intergovernmental competition. To this end, we construct a two-country asymmetric tax competition model in which the residents in one of the two countries face a domestic political agency problem: the residents do not know whether the incumbent politician is benevolent or leviathan. To separate themselves from leviathan politicians and attract voters' support, benevolent politicians argue for excessively low taxation, which can be regarded as anti-taxation populism. This anti-taxation populism exhibits the following two properties. First, the populist country's low production technology relative to the other country facilitates the emergence of anti-taxation populism. Second, antitaxation populism can improve welfare in terms of either the populist country or the whole world, depending on the asymmetry of the production technology.Résumé. Concurrence fiscale et problèmes d'action politique. En présence de concurrence intergouvernementale, nous étudions la fac ¸on dont les préoccupations de la classe politique en matière de réputation peuvent avoir une incidence sur la fiscalité. À cette fin, nous développons un modèle de concurrence fiscale asymétrique à deux pays. Dans l'un d'eux, les résidents sont confrontés à un problème d'action politique puisqu'ils ignorent si le politicien sortant est du type «bienveillant» ou «léviathan» (c'est-à-dire en faveur d'une optimisation des recettes fiscales). Afin de se dissocier des politiciens «léviathans» et s'assurer du soutien des électeurs, les politiciens «bienveillants» plaident pour une fiscalité excessivement faible. Cette approche s'apparentant à une forme de populisme «anti-fiscalité» présente deux propriétés: elle est d'abord favorisée par les