Rational herding in crowdfunding, social preferences under uncertainty and overplacementa behavioral aproach

Abstract

It has been proved in this dissertation how experimental economics continues to be a tool that, as it nourishes theoretical economics (and viceversa), it keeps producing results that are necessary to understand how to articulate social relations which, ultimately, have consequences oneconomic relations. Specifically, Chapter 1 has studied how the dissemination of new information affects crowdfunding markets and the generation of herding behaviour. Specifically, we provide evidence in that such behaviour is rational and could be well moulded through optimal choice under uncertainty with Bayesian review of beliefs. In this sense, the effect of the first sponsors on the agents# behaviour has been highlighted, through the change in the probabilities of success of the financing campaigns. Finally, another result shows how potential funders are more guided by the opinions of other individuals rather than those of experts. Chapter 2 of the dissertation has focused on analyzing how social preferences are affected in contexts of uncertainty. Taking an experimental (more conventional) laboratory design, a Dictator's Game was proposed in which it was verified to what extent the probability of being dictated (recipient) affected the distribution made by the agents according to the given initial endowment. The three conclusions obtained are: the level of generosity decreases as the probability of being dictator increases; being in the role of dictator reduces generosity; and starting making decisions with absolute certainty affects only those who are in the role of recipient. We did not find that variables such as sex, degree of risk aversion or cognitive abilities would affect the level of generosity of individuals. Finally, chapter 3 has focused on studying the existence -or not- of possible gender differences in the performance of tasks in which individuals must evaluate their performance. This evaluation was not only in reference to themselves (absolute), but also in relation to other reference groups (total, same sex, opposite sex). For this study, and as a control measurement, two tasks have been assigned that are different in the degree of the subjects# exposure to the performance of similar tasks (level of experience). Namely, the Raven test and the recording of a short video. The latter task adds the lack of previous experience in other similar tasks and the difficulty of self-assessment, since the evaluation was done by people who did not take part of the experiment. The results obtained do not support the hypothesis saying that men are more optimistic than women when performing the tasks with respect to others. In the Raven test no gender differences were found. No differences in both the terms of predictions and how it has been carried out with respect to others (overplacement), and in relation to the total or those of the same sex. However, we did find that women are more precise in adjusting their results compared to men.