Signaling Valence in Primary Elections - Games and Economic Behavior, Vol. 126, March 2021 (1-32) -- (working paper version pdf)

This paper presents a model of two-stage (primary and general) elections in which primary election candidates differ in terms of a privately observed quality dimension (valence). I show that primary election candidates have the incentive to signal their valence by means of their policy platform choice. There can be two types of separating equilibria in primary elections, with opposite implications concerning the relationship between valence and policy extremism. In an extremist equilibrium valent candidates choose more extreme policies than non-valent ones, whereas in a centrist equilibrium valent candidates move close to the incumbent from the opposing party. As a result, primary elections can foster the adoption of extremist policies, but they can also have the opposite effect. This result allows the model to also shed light on the circumstances in which party voters are likely to benefit from the introduction of primary elections.

Flip-flopping and Electoral Concerns - Forthcoming (2022), The Journal of Politics (pdf)

I consider a model of political agency in which an incumbent politician takes two decisions in sequence, having private information about the correct decision in each period. I find that policy reversals signal that a politician is incompetent. This suggests that the stigma associated with flip-flopping can be rationalized as a simple heuristic to select competent politicians. Furthermore, the negative impact of policy reversals on reputation gives the incumbent politician an incentive to disregard new information and stick to his initial policy. When reputation concerns are sufficiently strong, these incentives inhibit policy responsiveness, resulting in an inefficient degree of policy persistence. Term-limited politicians, or politicians with particularly safe (or compromised) election prospects, are more likely to reverse decisions. Relatedly, politicians are only willing to reverse decisions after a sufficient amount of time has passed, or when there is sufficient evidence that the state of the world might have changed.


Polarization and Policy Design (with Chris Li) -- submitted (pdf)

Political polarization has been rising in the U.S. and elsewhere, and there is concern that it will lead to increasingly divisive policymaking. We investigate this issue in a distributive context and show that the effect of polarization on policymaking is nuanced. As voters become more polarized, politicians may design policies to be more equitable. This is the case even if politicians are partisan-minded. Political turnover and the gradual resolution of policy uncertainty are key for the results. The implications of alternative electoral systems are also examined. Contrary to conventional wisdom, proportional systems may exacerbate partisan policymaking compared to majoritarian

Scandals, Media Competition and Political Accountability (with Antoni-Italo de Moragas) -- submitted (pdf)

We present a media market model in which a set of news outlets receive some information on whether a political scandal took place and compete to break the news. Media outlets can decide whether to break the story immediately or wait and fact-check, considering that if another outlet breaks the news, the profit opportunity disappears. We show that, as competition increases, each outlet becomes more likely to publish unverified news. As a consequence of this, the amount of information provided by the media market, and hence readers’ welfare, can be non-monotonic in the degree of competition. What is more, we show that readers following a realistic behavioral heuristic to consult media end up with no information as competition grows large.


Strategic Extremism and Voters' Sophistication (with Elia Sartori) -- draft coming soon!
Experts and Political Accountability (with Antoni-Italo De Moragas and Philipp Denter)
News Filtering on Social Media (with Anna d'Annunzio and Giovanni Immordino)

Giovanni Andreottola,
Jun 21, 2021, 6:26 AM
Giovanni Andreottola,
Jul 29, 2021, 9:05 AM
Giovanni Andreottola,
Nov 9, 2020, 3:29 AM
Giovanni Andreottola,
Nov 9, 2020, 3:28 AM