Experimental and Survey Evidence on the Development of Preferences and Skills (dissertation)

Abstract

The focus of this research is on decisions related to human capital formation and preferences crucial for social interaction. These decisions are analyzed at different stages of the life-cycle, starting with other-regarding preferences in preschoolers, followed by skill formation in adolescents and preferences for honesty in adults. Answering the questions on skill formation and human capital accumulation was possible by exploiting the full potential the data provided, that is by combining different data types as well as making use of the data structure and analyzing paradata from the data collection process. Thus, this dissertation contributes to the literature by combining survey data, experimental data and paradata to analyze factors that influence individual decision-making. An important aspect this thesis focuses on are environmental factors that influence decision-making. These include family or community factors or situational factors in an experiment. Spite and cognitive skills in preschoolers: Other-regarding preferences in adults have been examined in depth in the literature. Research has shown that spiteful preferences play a crucial role in the development of human large-scale cooperation. However, there is little evidence of the factors explaining spiteful behavior in children. We investigate the relationship between children’s cognitive skills and spiteful behavior in a sample of 214 preschoolers aged 5-6 and their mothers. Here, other-regarding behavior in children is elicited through four simple allocation decisions. A key advantage of our study is that we have information about children’s cognitive and non-cognitive skills as well as maternal and household characteristics. We find that higher cognitive skills are associated with more spiteful behavior in children. This relationship is even more pronounced among boys. Moreover, we find further gender differences that depend on the measure of cognitive skills and the degree of spite displayed. Is it the family or the neighborhood? Evidence from sibling and neighbor correlations in youth education and health: In this paper we present sibling and neighbor correlations in school grades and cognitive skills as well as indicators of physical and mental health for a sample of German adolescents. In a first step, we estimate sibling correlations and find substantial influence of shared family and community background on all outcomes. To further disentangle the influence of family background and neighborhood, we estimate neighbor correlations. Our results show that for all outcomes, estimated neighbor correlations are clearly lower than estimated sibling correlations. However, especially for cognitive skills and mental health, neighbor correlations are still substantial in relation to sibling correlations. Thus, compared to existing results from other countries, the influence of the neighborhood is not negligible in Germany for these outcomes. Who lies and to whom? Experimental evidence on cheating in a household survey context: In this paper, I evaluate a cheating experiment implemented in the German Socio-Economic Panel (SOEP) Study. My results show no relationship between cheating and socio-economic characteristics, personality traits, or economic preference parameters, except between cheating and education. I also study a high- and a low-stakes treatment to test the robustness of cheating in the face of increased gains. The results indicate that people cheat more when potential gains from doing so are greater, but that education appears to be a mediating factor. Further, I carry out a closer analysis of the decision-making situation focusing on the relationship between interviewer characteristics and the probability of cheating. This analysis provides the novel result that some interviewer characteristics—such as gender—significantly increase the probability of cheating.