Essays in Behavioral and Experimental Economics

Abstract

Behavioral economics has improved the understanding of economic phenomena by enriching the understanding of economic decision making with insights from psychology, sociology, and anthropology. Rigorous empirical investigations of individual behavior---which commonly involve the use of laboratory experiments---have been at the heart of behavioral economics and have lead to new theoretical accounts of decision making. The following three insights gave rise to influential branches of behavioral economics. First, the context in which individuals make decisions often unleash behavioral influences that go beyond those identified by standard economic theory. In particular, individual preferences commonly depend on contextual features, as has been highlighted by the literatures on reference-dependent preferences, and default effects. Second, individuals' perception and processing of information often does not live up to the high demands of standard economic theory. Instead, individuals seem to employ simple heuristics and attention-based decision rules in complex environments. Third, while standard economic theory typically constrains individuals' motives to pure self-interest, a more comprehensive view on individuals' behavior in social interactions uncovers that individuals often care directly about the well-being of others as well as about how they are viewed and treated by others. This thesis consists of three essays that each contribute to one of these three building blocks of research in behavioral economics. <br /> In my first essay, I investigate the consequences of social reference points for decision making under risk in a series of laboratory experiments. In the main experiment, decision makers observe the predetermined earnings of peer subjects before making a risky choice. I exogenously manipulate peers' earnings and find a significant treatment effect: decision makers make riskier choices in case of larger peers' earnings. The treatment effect is consistent with the predictions of a model featuring social-comparison--based reference points and loss aversion. In two control experiments, I demonstrate that nonsocial---e.g., expectations-based---reference points do not explain the treatment effect. <br /> In my second essay, I present novel results on individuals' intertemporal choices. The findings cannot be explained by exponential and hyperbolic discounting, the canonical approaches to intertemporal decision making in economics, but are consistent with an attention-based approach to intertemporal decision making that is based on concentration bias. In particular, the essay provides causal evidence from novel lab experiments that intertemporal choices are systematically affected by whether consequences of intertemporal choice are concentrated in few or dispersed over multiple periods: (i) Individuals are less patient in the case that the advantages of patient behavior are dispersed over many future periods than when they are concentrated in a single future period. (ii) Individuals are more patient in the case that the disadvantages of patient behavior are dispersed over multiple earlier periods than when they are concentrated in a single earlier period. Both findings demonstrate concentration bias in individuals' intertemporal choices. Our results are in line with recent theoretical models of attention-based decision rules.<br /> In my third essay, I study whether prior experience of unfair versus fair treatment affects how much individuals trust others. The essay provides causal evidence that trust is affected by prior personal experience of fair versus unfair treatment by an unrelated third party. I compare the willingness to trust of subjects in a lab experiment after they experienced either being paid or not being paid for a real-effort task by a peer subject. After being paid, subjects' willingness to trust is substantially higher relative to subjects who were not paid previously. Importantly, this treatment effect holds despite the fact that subjects knew the exact frequency with which subjects overall got paid or did not get paid, such that the personal experience of fair versus unfair treatment did not provide additional information regarding the subsequent interaction. Rational learning hence cannot explain the treatment effect on trust. By employing a control experiment, we show that the effect of experiencing fair versus unfair treatment on trust does also not result from income effects: when subjects were paid based on a coin toss, subjects' willingness to trust was similar to subjects who where not paid based on a coin toss.