Shunsuke Yoshimura

Research

Working Paper

Ambiguous Policy Results Indicate Pandering

-Presented at MPSA 2024, APSA 2025

Abstract: Executives sometimes pander, i.e., implement popular policies that will cause bad results to secure reelection, if the probability that voters learn the right policy before the next election is low. Yet, executives have some control over the probability when they can provide verifiable information on policy results. In such situations, pandering becomes impossible because high-ability executives have an incentive to provide as much information as possible, and low-ability executives must follow suit to pretend to be the high type, which makes pandering no longer work. Surprisingly, this intuition holds even if the revelation of information is unobservable to voters, where the low type's different choice does not instantly reveal her type, and indeed, information is not always revealed. The normative implication is that when executives are expected to have evidence of policy results, real voters could also prevent pandering by interpreting ambiguous policy results as a sign of pandering.

Work in Progress

Checks and Imbalances: When Separation of Powers Facilitates Extremism and Policy Polarization

-Will be presented at MPSA 2026, APSA 2026, Bocconi-Columbia Political Economy Conference

Teaching

• Mathematical Modeling (PhD Formal Theory I), Fall 2025, TA for Prof. Mark Fey  [Course Description] [TA Evaluation]

• Positive Political Theory (PhD Formal Theory II), Spring 2026, TA for Prof. Tasos Kalandrakis  [Course Description]