My Research

Evasive Radicalism

What it's about:  I construct a model of electoral accountability wherein the incumbent can include a challenger in governance in order to disclose her relative quality to the voter. The mechanism uncovered offers novel rationale for why increases in voter disaffection with the mainstream political class are often met with an influx of radical challenges: a radical challenge increases the incumbent's disclosure cost. For an inferior quality moderate, a radical challenge is always \textit{evasive}; the challenger forgoes inclusion as a moderate to run as a radical in an effort to remain untested and climb higher. When the incumbent includes moderate challenges but always excludes a radical one simple comparative statics show that the mechanism of evasive radicalism is more likely to be active when politicians are relatively office motivated; office rents are shared with friction; and the challenger would be influential if included. Thus, the model sheds light on empirical analyses of \textit{cordon sanitaire} policies to combat extremist parties. 

Personalized Commerce and Vote Selling (w/ Meiqing Zhang)

What it's about: Digital platforms (e.g., Youtube or Tik Tok) earn revenue through targeted advertising on behalf of their commercial interests. However, a common concern is that digital platforms may also target individuals in an attempt to distort the democratic process to their own ends. We consider a monopolist platform that responds to a citizen's privacy choice with a product recommendation. Only the platform knows the mapping between preference types and the citizen's ideal product. In equilibrium, the citizen may choose a lower privacy level in order to increase her consumption utility at the expense of increasing her probability of a voting error. This mechanism is consistent with the empirical fact that voting errors are concentrated among individuals with low levels of wealth.  The model also yields implications for the platform-advertiser relationship when the platform has a political objective; and for the existence of a societal level trade-off between consumer welfare and common interest electoral outcomes.

Pandering & Civic (Dis)Engagement

What it's about: Canonical models of electoral implicitly assume that voters come pre-equipped with the information needed to parse their leaders' signals. I instead assume that a voter chooses how much effort to spend investigating the circumstances behind a policy—i.e., a level of civic engagement—before deciding whether to retain the leader. In equilibrium, engagement has value if and only if policy is partially informative about the leader's type, and so it collapses at two informational extremes: under pandering, policy choice is independent of type and there is nothing to learn; and under separation, policy choice is fully revealing and the voter again disengages. Separation arises only when the precision of the leader's information falls below a confidence threshold: a leader who lacks confidence in his signal falls back on his policy bias, and sincere play becomes transparent. Thus, when the return to selecting good leadership is greatest the voter makes no effort. Notably, anticipated engagement does not discipline pandering; on the pandering path the voter rationally disengages, insulating the equilibrium from the engagement technology. Within the informative equilibrium, anticipated engagement raises the electoral reward to conformity precisely for the leader who believes that no common-interest policy exists. The takeaway is that anticipated scrutiny disciplines dissent, and that civic engagement incentivises leaders to use socially valuable, private information.

Climate Denial & Renegotiation Failure

What it's about: I explore a dynamic bargaining model within the context of an emerging, international crisis. The leaders of two countries know the magnitude of the impending crisis, but their representative constituents do not. Each leader chooses between informing her constituent, or understating the cost of a crisis. Leaders then bargain to partition the costs of the preventative action level consistent with their constituents' induced beliefs. The leader that understates the crisis relative to her opponent appropriates a larger bargaining share today. But, she also reduces the level of preventative action, and risks re-negotiation failure tomorrow–her constituents may regard any further action as unnecessary. Thus, the model offers a divergence in cultural, climate beliefs as plausible rationale for the international community's failure to build on the Paris Accord.