Essays in Experimental Economics

Abstract

This dissertation examines economic decisions through the lens of beliefs and mental models. Across three chapters, it experimentally investigates how they are formed and revised, as well as how they in turn shape economically relevant decisions. <br/> The first chapter (co-authored with Paul Grass and Philipp Schirmer) examines how mental models are revised in dynamic environments, in which previously omitted variables become available. Using a novel two-stage experimental design, it shows that models are sticky, as participants fail to sufficiently revise misspecified models, even when provided with complete data on all relevant variables. Revised models remain strongly influenced by earlier models formed using only a subset of variables. <br/> The second chapter (co-authored with Yana Radeva) then turns to policy preferences, asking how beliefs about the causes of the gender wage gap shape support for policies addressing gender inequality. Using an online survey experiment in the US, it shows that people substantially underestimate the contribution of children to the gender wage gap. Increasing the perceived importance of children in driving the gap changes the composition of policy support, shifting it toward child-related policies and away from discrimination-related ones. The effect on the overall level of support, however, depends on how fair people perceive the different causes to be. <br/> The third chapter (co-authored with Laura Ehrmantraut, Pia Pinger and Renske Stans) then looks at hiring and wage decisions, asking how employers evaluate candidates with different levels of educational attainment and how these evaluations relate to employers' underlying beliefs about productive traits. Using an online survey experiment with German HR managers, it shows that while finishing a master's degree increases candidate attractiveness, acquiring additional credits without obtaining the degree is perceived as less favorable than a bachelor's degree. The majority of these effects are explained by what completing or not completing a degree signals about cognitive and especially non-cognitive traits. <br/> Together, the three chapters demonstrate the following: Firstly, that people can hold different mental models, despite having received the same information, depending on how that information was encountered. Secondly, these models systematically shape policy preferences, as well as hiring and wage decisions.