Essays on Beliefs and Economic Behavior
Next to preferences over outcomes, people's subjective beliefs about unobserved states of the world are the central building block of economic models of decision-making. This thesis submits that a more nuanced account of the nature of subjective beliefs can improve the explanatory power of models of economic behavior. It revolves around the following questions. How do people incorporate new information into their beliefs? What are the cognitive mechanisms underlying different updating rules? To what extent is observed heterogeneity in beliefs and behavior predictable? Are beliefs shaped by people's individual experience? How do beliefs translate into economic behavior? The unifying approach of this work is cross-disciplinary, leveraging ideas from other fields such as cognitive science and psychology. The empirical motivation for Chapter 2: "Inattentive Inference" is the pervasive evidence for miscalibration to news in practice. In many situations, beliefs overreact, as if base rates are neglected or new information overweighted, and they often underreact, leading to information rigidities or conservatism in updating. Whether and how these conflicting updating patterns can be reconciled remains an open question. Chapter 2 builds on the observation that most information structures pose a signal extraction problem for belief updating: an agent wants to learn an unobserved state of the world X, but information about X is "compound" in that it also depends on other states Y. Inattention to other causes Y in the signal structure could create misattribution to X. A series of laboratory and online experiments show that most people do not take into account other causes (labelled feature neglect), leading to excessively sensitive and overprecise beliefs. There is pronounced heterogeneity, with up to 90% of beliefs corresponding to three updating rules: feature neglect, Bayesian updating and non-updates. Exploring the underlying mechanism, I find that unawareness about the necessity to factor in alternative causes drives feature neglect, whereas non-updating follows cost-benefit considerations. I propose a conceptual framework that accounts for the combined evidence. Moreover, learning is found to be limited by unawareness, but cues and hints debias by changing people's mental models directly. Chapter 3: "Heterogeneity of Loss Aversion and Expectations-Based Reference Points" examines how people's forward-looking beliefs, i.e., expectations, affect their decisions. A seminal insight from psychology is that we tend to evaluate outcomes relative to a reference point. In theories of reference-dependent decision-making, people code outcomes as gains or losses relative to some reference point. Yet, the location of this reference point is a critical degree of freedom. A recent theoretical advance characterizes the reference point based on people's expectations about their own future outcomes. In the past decade, empirical tests of this model yielded mixed results and there remains a lack of consensus on the location – and thus the empirical relevance – of reference points. Chapter 3 attempts to reconcile different approaches and findings. In this study that is joint work with Lorenz Goette, Charles Sprenger and Alexandre Kellogg, we developed a tightly controlled exchange experiment with two main innovations: First, the design recognizes that testing the role of expectations-based reference points requires experimental control of other plausible avenues of reference dependence, such as the status quo or personal experience. Second, it accommodates a critical confound related to the key behavioral parameter, loss aversion. Loss aversion captures that people dislike losses more than equal-sized gains. A growing body of evidence documents substantial heterogeneity in measured levels of loss aversion, with a substantial fraction of people being loss-neutral or even loss-loving. Different levels of loss aversion, however, lead to different signs of comparative statics. In our results, recognizing heterogeneity in loss aversion allowed us to reliably recover the central prediction of expectations-based reference points. Chapter 4: "Breaking Trust: On the Persistent Effect of Economic Crisis Experience" shifts the focus to a particularly important belief for economic transactions. Trust is the degree of belief in the benevolent intentions of another person. In the realm of economic behavior, trust plays a central role as a prerequisite for all forms of economic exchange: without a minimal amount of trust in the counterpart, no person would be willing to sign a contract. In fact, trust has been shown to affect economic outcomes at the individual, group and societal levels. However, much less is known about the origins of trust. Recent evidence documents that levels of trust vary substantially across locations and over time, but the determinants of this geographical and temporal variation are not well understood. In Chapter 4, which is joint work with Tom Zimmermann, we analyzed the economic implications of a breach of trust argument, positing that trust is not easily restored once it has been abused. Building on a nascent literature on the economic implications of people's lifetime experience, we hypothesized that trust is partially determined by the experience of catastrophic macroeconomic events. Using a variety of identification strategies in a large cross-country sample, we estimated a persistent and robust negative long-term effect of economic crisis experience on trust in other people. In line with the breach of trust hypothesis, the effect was specific to living through crises in trust- intensive domains, most of all banking crises. The effect was not driven by distrust in financial institutions but was accommodated by a lack of confidence in the political class, and operated via beliefs rather than changes in preferences. Chapter 5: "Negative Long-run Effects of Prosocial Behavior on Happiness" studies happiness. In recent times, measures of subjective well-being are increasingly viewed as relevant indicators of a society's welfare, and a rising number of countries have incorporated national happiness levels as a policy objective. This development concurs with a renewed scientific interest in the causes of happiness. Most prominently, recent studies contribute to a debate spanning more than two millennia on the hypothesis that prosocial behavior is a key to happiness. The existing causal evidence indeed confirms a positive influence of prosocial behavior on happiness, but is limited to the short-term effects of an enforced prosocial or selfish act. In Chapter 5, which is joint work with Armin Falk, we reconsider this hypothesis in a behavioral experiment that extends the scope of previous studies in various dimensions. In our Saving a Life paradigm, every participant either saved one human life in expectation or received one hundred euros, respectively. Using a choice between two binary lotteries with different chances of saving a life, we observed subjects' intentions at the same time as creating random variation in prosocial outcomes. We repeatedly measured happiness at different time horizons after the experiment. We confirmed the previous consensus finding of a positive short-term effect, but this effect quickly faded. As time passed, the sign of the effect even reversed, and we recorded significantly greater happiness associated with the selfish outcome than with the prosocial outcome one month later. Our findings hint at distinct sources of happiness as time passes. Chapter 5 provides a first piece of evidence that a comprehensive understanding of the effects of prosocial behavior on happiness requires a more nuanced view that accounts for delayed effects.
