Write a PREreview

Conventional electoral theory posits that transparency enhances accountability by enabling voters to make informed, rational decisions. However, empirical evidence from diverse national contexts reveals a persistent disconnect: the disclosure of corruption frequently fails to precipitate significant electoral repercussions. This apparent paradox has led some scholars to posit voter irrationality. This paper challenges this interpretation, contending instead that the absence of electoral sanctions against corrupt politicians arises from the inherent complexities of strategic environments, characterized by information asymmetry and ambiguous signaling. We develop a formal model employing dynamic game theory to analyze the strategic interactions between politicians and voters who update their beliefs through Bayesian learning. Our model demonstrates that electoral accountability emerges incrementally, contingent upon sequences of credible signals across multiple electoral cycles. To mitigate the pervasive challenges of adverse selection and moral hazard, we propose the establishment of an independent Political Rating Agency (PRA). This agency would furnish credible, standardized, and readily interpretable information. The efficacy of such a PRA is critically contingent upon its perceived credibility and the electorate's propensity to integrate its ratings into their decision-making calculus.