Return to Presenters and Papers

Friday Presenters and Papers

Jon WoonAs We Are: Directions and Questions for Behavioral Models of PoliticsHow might the theoretical toolkits for modeling politics be expanded to include features of humans "as we are"? I discuss how models of decision making and social interaction might include beliefs, limited sophistication, learning, and dynamics.
Eric DicksonLearning About Legitimacy Using ExperimentsIn a series of laboratory experiments employing different institutional manipulations and different behavioral measures of legitimacy, we attempt to isolate non-material motivations underlying legitimacy and to study how these non-material motivations vary across different institutional (and other) settings. An ultimate goal of the project is to learn how to build a psychologically-realistic model of legitimacy into a behavioral game-theoretic framework.
David SiegelSimple Models for Complex ProblemsThere are multiple approaches available to us in dealing with substantively important deviations from standard game theory. One is to include complexity. But another is to reduce the complexity of the assumptions we employ. I discuss each approach broadly and offer an example as to how the latter might work.
Ryan Vander Wielen & Kevin ArceneauxDemocratic Accountability for Some: Individual Differences in How Partisans Process Political Information
Most political science models implicitly assume that individuals update their political attitudes in a uniform fashion, and therefore these models have powerful implications for democratic accountability in a polarized polity. In this paper, we demonstrate that individual differences in motivation shape how people update support for their party in light of new information. We conduct a web-based survey experiment and draw on observational panel data collected during the 2012 presidential election. Consistent with our theoretically derived expectations, we find that individuals who possess a high need for cognition (i.e., they enjoy effortful thinking) exhibit more plastic attitudes toward their party and are more likely to update their attitudes in the direction of negative information about their party. In contrast, individuals who are high in need for affect (i.e., they enjoy experiencing strong emotions) tend to be quite stable in their opinions and unwilling to punish their party.
Michelle TorresPersonality Stability and Politics: TIPI Variability
Researchers frequently claim that personality traits, as measured using the Big Five personality through the TIPI (Ten Item Personality Inventory) battery, affect Americans’ political attitudes and behaviors. Such studies often depend on two key assumptions: personality measurements display stability over time and variability in such measurements predates those of political behavior. In this paper we employ new panel survey data to test these assumptions. First, we find high levels of variability in responses to TIPI. Second, we associate this variability with not only socioeconomic and demographic characteristics, but also, and more concerning, political attitudes. The variability and associations of the instrument suggest that the relationship between personality and politics may be weaker than indicated by previous scholars and moreover should not be employed as a variable that predates political behavior. Finally, we address the consequences of these tests by applying our findings to previous research that hinges on the relationship between personality and political behavior. The results of this exercise suggest that ignoring the dynamic nature of personality could alter the interpretation of its relationship to political attitudes.
Salvatore NunnariDynamic Coalitions and CommunicationWe present a laboratory experiment to study the formation of durable coalitions in a dynamic bargaining setting where the current status quo policy is determined by the policy implemented in the previous period. Our main experimental manipulation is the ability of legislators to negotiate with one another, through unrestricted cheap-talk communication, before the proposal reaches the floor. We compare committees with no communication, committees where communication is public and messages are observed by all committee members, and committees where communication is private and any committee member can send private messages to any other committee member. We find that the ability to negotiate has a significant impact on outcomes and coalitions. Dynamic coalitions emerge more frequently and last longer when communication is allowed. When negotiations are public, committees quickly converge to universal coalitions, where every member in the committee receive a significant fraction of the budget and supports the agreement. On the other hand, when we allow for private communication, we observe a significant increase in the share of minimal-winning coalitions.
Jens GrosserTentative and definite policy proposals in committees: An experimental studyWe experimentally study a two-stage legislative bargaining game with incomplete information, consisting of one proposer and two voters. While the proposer’s ideal policy and the default policy if the proposal is eventually rejected are common knowledge, the two voters’ ideal policies are private information. We distinguish between first-stage proposals that are tentative (so voting decisions are “straw polls”) and definite as well as between simple-majority and unanimity voting. On average, observed proposals and voting decisions are usually close to the Bayesian Nash predictions. Moreover, the proportions of accepted proposals are the same for first-stage tentative and definite proposals under simple-majority voting, but the proportion is markedly lower for tentative proposals under unanimity voting. Our results have important implication for the design of committee decision-making.
Daniel ButlerLegislative Holdouts
Holding out occurs when a legislator votes against a policy that is closer to her ideal point than the status quo. Our original survey of state legislators shows that a large number, over a quarter, indicate that they would vote against a proposal even though it is closer to their ideal policy than the status quo. Following our pre-analysis plan, we examine a number of possible factors that could explain why these legislators hold out. Our data indicate that Republicans, legislators in the majority, and those who fear that their constituents will punish compromise are most likely to hold out. Our results show one way legislative gridlock can occur even when a supermajority of legislators could be made better off by policy change.
Sean Gailmard & John PattyPreventing Prevention
Most conceptions of electoral accountability rely upon voters’ abilities to process and act upon information about the performance of incumbent politicians in a rational manner. Yet recent empirical evidence suggest that voters do a particularly poor job of this in the context of disasters, failing to reward incumbents for beneficial actions and punishing incumbents for events arguably beyond their control. We develop a simple model of elections in the context of disasters, which shows that (at least some of) these empirical regularities are consistent with rational choice by voters uncertain about politicians’ motivations. In many cases, the unique equilibrium of the model entails suboptimal disaster prevention policy by all incumbents, even though voters are rational and make the best possible use of their blunt tool of political accountability.
Christopher LiElectoral Control with Behavioral Voters
We present a model of electoral control with behavioral voters. The model is intended to capture the main regularities of voting behavior found in empirical studies. Specifically, the voters' propensity to keep an incumbent in office is governed by a stochastic reinforcement process instead of strategic reasoning. The likelihood of a positive feed-back for a voter depends on the effort level exercised by the public official. We show that despite the lack of rational responses by voters, the electoral control of public officials can be substantial. Indeed, electoral control is the highest when voters are most forgetful. Moreover, our model generates comparative statics that are consistent with the main empirical regularities of electoral accountability.
Dominik Duell & Dimitri LandaDeterminants of Discrimination in Strategic Settings
n a laboratory experiment in a strategic principal-agent setting, we analyze how principals' elicited assessments of the causes of agent performance vary depending on whether they share the social identity with the agents. We isolate the effect on subjects' beliefs and choices of the strategic environment as such and provide a direct test of the strategic theory of statistical discrimination. We find that when principals use the sanctioning tools at their disposal in an outcome-contingent way, principals' and agents' choices may sustain a pattern of beliefs that is observationally equivalent to ``ultimate attribution error'': upon observing good outcomes, principals attribute them more readily to their agents' effort and reward their agents more frequently when they share a social identity; and in turn, agents who share a social identity with their principals tend to invest more into effort in expectation of principals' reward choices. However, when principals do not use the sanctioning tools outcome-contingently, and when they do not have access to sanctioning tools, they do not hold such beliefs. This and other evidence we report suggests that in strategic settings, principals' prejudicial treatment of agents may be a product of the strategic nature of the environment, rooted in asymmetric but correct beliefs. Prejudice and discrimination may owe more to the strategic nature of the environment than previously recognized.
Scott PageWhen Order Affects Performance: Institutional Sequencing, Cultural Sway, and Behavioral Path Dependence
How does the order that laws and other institutions are introduced affect their performance, and under what conditions will sequencing matter? To gain a foothold on these questions, we construct a formal model of institutional sequencing that includes cultural sway in the form of behavioral spillovers. We derive two broad categories of results: First, we characterize the relationships between the extent of cultural sway, payoff structures, and path dependence. We find that path dependence increases and then decreases in cultural sway and that optimal payoff structures in novel games require weak punishment regimes. Second, we characterize the optimal sequencing of institutions. We show that optimal sequences satisfy a property we call multi-incrementalism that combines initial institutional diversity with gradual movements to reduce inefficient spillovers. Such sequences have the counterintuitive property that they enable the potential for path dependence in order to avoid its realization.