In der ersten Abhandlung untersuche ich die Auswirkungen von Informationsfriktionen und Aufwandskosten auf die Effektivität von Steuervergünstigungen bei der Förderung von Spenden für wohltätige Zwecke. Mittels eines groß angelegten Umfrageexperiments und einem Feldexperiment prüfe ich empirisch, ob die Verringerung von Informationsfriktionen durch Hervorhebung von Steuervergünstigungen und Bereitstellung von Informationen über die Höhe von Steuervergünstigungen Spendenentscheidungen verändert. Meine Ergebnisse zeigen, dass die Spenden steigen, wenn Steuervergünstigungen hervorgehoben und Informationen bereitgestellt werden. Ich lege dar, dass der Hauptverhaltensmechanismus, der Informationsfriktionen zugrunde liegt, Unaufmerksamkeit ist. Die zweite Abhandlung ist eine gemeinsame Arbeit mit Steffen Altmann, Armin Falk, Paul Heidhues und Rajshri Jayaraman. In dieser Arbeit untersuchen wir die Auswirkungen von voreingestellten Beträgen auf Spenden für wohltätige Zwecke in einem groß angelegten Feldversuch auf einer Online-Spendenplattform. Wir beobachten eine starke Auswirkung von voreingestellten Beträgen auf das individuelle Verhalten, stellen jedoch fest, dass die Gesamtspende nicht von voreingestellten Beträgen beeinflusst wird. Mit einem Strukturmodell wird untersucht ob die Personalisierung von voreingestellten Beträgen die Spendeneinnahmen erhöhen kann. In der dritten Abhandlung, die in Zusammenarbeit mit Rajshri Jayaraman und Michael Kaiser verfasst wurde, untersuchen wir gemeinnützige Spenden für Naturkatastrophen auf einer großen Online-Plattform. Wir beobachten, dass der Großteil der Spenden für wohltätige Zwecke an einen winzigen Teil der Naturkatastrophen geht, bei denen es sich in der Regel um besonders schwere Katastrophen handelt, über die in den Medien berichtet wird. Unter Verwendung eines Ereignisstudiendesigns finden wir Hinweise, die mit zwei Erscheinungsformen der Spenderermüdung übereinstimmen, jedoch nicht mit einer dritten. / In the first essay, I study the implications of information frictions and hassle costs for the effectiveness of tax incentives in encouraging charitable giving. Empirically, I test whether mitigating information frictions by making tax incentives more salient and by providing information on the magnitude of tax incentives alters donation decisions by conducting a large-scale survey experiment, representative of the German adult population and I complement this with a large-scale field experiment on an online donation platform. My findings indicate that when tax incentives are made salient, and when information is provided, donations increase. I find that the main behavioral mechanism underlying information frictions is inattention. Analyzing who responds more to salience and information provision shows that it is predominantly individuals that face lower hassle costs and have high incomes. The second essay is joint work with Steffen Altmann, Armin Falk, Paul Heidhues, and Rajshri Jayaraman. In this paper, we study the effects of defaults on charitable giving in a large-scale field experiment on an online fundraising platform. We document a strong effect of defaults on individual behavior but nevertheless find that aggregate donation levels are unaffected by defaults. In contrast, co-donations increase in the default amount. We complement our experimental results with a structural model that investigates whether personalizing defaults based on individuals’ donation histories can increase donation revenues. In the third essay, which is joint work with Rajshri Jayaraman and Michael Kaiser, we investigate charitable donations to natural disasters on a large online platform. We document that the bulk of charitable donations go to a tiny fraction of natural disasters, which tend to be severe disasters that receive media coverage. Using an event study design, we find evidence consistent with temporal fatigue and donor fatigue, but not with crowding out.

The transport and urban economics literature applies different labor supply approaches when studying economic or planning instruments. Some studies assume that working hours are endogenous while the number of workdays is given, whereas others model only decisions on workdays. Unfortunately, empirical evidence does hardly exist on account of missing data. Against this background, we provide an assessment of whether general effects of transport policies are robust against the modeling of leisure demand and labor supply. We introduce different labor supply approaches into a spatial general equilibrium model and discuss how they affect the welfare implication of congestion policies. We, then, perform simulations and find that in many cases the choice of labor supply modeling not only affects the magnitude of the policy impact but also its direction. While planning instruments are suggested to be quite robust to different labor supply approaches, the way of modeling labor supply may crucially affect the overall welfare implications of economic instruments such as congestion tolls. Based on these findings it becomes clear which labor supply approach is the most appropriate given specific conditions. Our study also emphasizes the need for better micro labor market data that also feature days of sickness, overtime work used to reduce workdays, the actual number of leave days, part-time work, days with telecommuting etc.

The transport and urban economics literature applies different labor supply approaches when studying economic or planning instruments. Some studies assume that working hours are endogenous while the number of workdays is given, whereas others model only decisions on workdays. Unfortunately, empirical evidence does hardly exist on account of missing data. Against this background, we provide an assessment of whether general effects of transport policies are robust against the modeling of leisure demand and labor supply. We introduce different labor supply approaches into a spatial general equilibrium model and discuss how they affect the welfare implication of congestion policies. We, then, perform simulations and find that in many cases the choice of labor supply modeling not only affects the magnitude of the policy impact but also its direction. While planning instruments are suggested to be quite robust to different labor supply approaches, the way of modeling labor supply may crucially affect the overall welfare implications of economic instruments such as congestion tolls. Based on these findings it becomes clear which labor supply approach is the most appropriate given specific conditions. Our study also emphasizes the need for better micro labor market data that also feature days of sickness, overtime work used to reduce workdays, the actual number of leave days, part-time work, days with telecommuting etc.