Prospect theory introduces several anomalies in the behavior of rational agents, including loss aversion, the reflection effect, probability weighting, and the certainty effect. Loss aversion occurs relative to the current state of the world, called reference point. Being loss averse causes people to prefer the current state of affairs above and beyond the expected utility that comes from a risky political change, engendering a status quo bias. Yet, bias is asymmetric due to the reflection effect: people are too tepid toward advantageous platforms or candidates, whereas they are not critical enough of detrimental policies or bad politicians. Both rich and poor citizens take similar stances on nonpartisan issues (such as national defense): this happens because they evaluate uncertain policy changes relative to a reference point. Citizens welcome radical political platforms with greater enthusiasm than incremental proposals. Generally, under prospect theory societal conflict is smoother than under expected utility theory. Older societies are more prone to preserving the status quo than younger ones. These properties also affect the choice of voting rules. Loss aversion induces people to prefer more prudent voting rules and preserve the status quo. Hence, agents favor higher majority thresholds or even unanimity over simple majority in constitutional choice. The status quo bias supports the persistence of policy cycles, with prolonged drifts in one direction before a trend reversal. In sum, loss aversion and other anomalies pinpointed by prospect theory offer insightful predictions with which to study political phenomena.
Prospect Theory, Loss Aversion, and Political Behavior
Francesco Passarelli and Alessandro Del Ponte
Expected Utility and Political Decision Making
Expected utility theory is widely used to formally model decisions in situations where outcomes are uncertain. As uncertainty is arguably commonplace in political decisions, being able to take that uncertainty into account is of great importance when building useful models and interpreting empirical results. Expected utility theory has provided possible explanations for a host of phenomena, from the failure of the median voter theorem to the making of vague campaign promises and the delegation of policymaking. A good expected utility model may provide alternative explanations for empirical phenomena and can structure reasoning about the effect of political actors’ goals, circumstances, and beliefs on their behavior. For example, expected utility theory shows that whether the median voter theorem can be expected to hold or not depends on candidates’ goals (office, policy, or vote seeking), and the nature of their uncertainty about voters. In this way expected utility theory can help empirical researchers derive hypotheses and guide them towards the data required to exclude alternative explanations. Expected utility has been especially successful in spatial voting models, but the range of topics to which it can be applied is far broader. Applications to pivotal voting or politicians’ redistribution decisions show this wider value. However, there is also a range of promising topics that have received ample attention from empirical researchers, but that have so far been largely ignored by theorists applying expected utility theory. Although expected utility theory has its limitations, more modern theories that build on the expected utility framework, such as prospect theory, can help overcome these limitations. Notably these extensions rely on the same modeling techniques as expected utility theory and can similarly elucidate the mechanisms that may explain empirical phenomena. This structured way of thinking about behavior under uncertainty is the main benefit provided by both expected utility theory and its extensions.
Sunk Costs and Political Decision Making
Charles A. Miller
The “sunk costs fallacy” is a popular import into political science from organizational psychology and behavioral economics. The fallacy is classically defined as a situation in which decision-makers escalate commitment to an apparently failing project in order to “recoup” the costs they have already sunk into it. The phenomenon is often framed as a good example of how real decision-making departs from the assumption of forward-looking rationality which underpins traditional approaches to understanding politics. Researchers have proposed a number of different psychological drivers for the fallacy, such as cognitive dissonance reduction, and there is experimental and observational evidence that it accurately characterizes decision-making in certain contexts. However, there is significant skepticism about the fallacy in many social sciences, with critics arguing that there are better forward-looking rational explanations for decisions apparently driven by a desire to recoup sunk costs – among them reputational concerns, option values and agency problems. Critics have also noted that in practical situations sunk costs are informative both about decision makers’ intrinsic valuation for the issue and the prospects for success, making it hard to discern a separate role for sunk costs empirically. To address these concerns, empirical researchers have employed a number of strategies, especially leveraging natural experiments in certain non-political decision making contexts such as sports or business, in order to isolate the effects of sunk costs per se from other considerations. In doing so, they have found mixed support for the fallacy. Research has also shown that the prevalence of the sunk costs fallacy may be moderated by a number of factors, including the locus of decision-making, framing, and national context. These provide the basis for suggestions for future research.