Analyzing the distributions of wages for whites, blacks and Hispanics reveals the existence of a wage gap throughout the distribution. There are also clear cognitive and noncognitive skill differences across groups. Do differences in the distributions of these skills explain differences in the distributions of wages? Do predicted distributions of wages resulting from rewarding blacks and Hispanics as if they were white help explain the observed wage gap? Using data from the NLSY79, I look at the impacts of noncognitive skills on wages for blacks, Hispanics and whites. I estimate the entire distribution of wages conditional on skills for blacks and Hispanics to see if there is a difference in wages individuals with the same level of cognitive and noncognitive skills. I find that all cognitive and noncognitive measures examined are important in explaining the wage penalty paid by blacks and Hispanics and that, for blacks, predicting their wages conditional on skills approximates the distribution of actual wages. Do employers recognize noncognitive skills at the onset (interview) or is there a learning process? How does learning about these noncognitive skills occur over time? This paper uses data from the NLSY79 to incorporate measures of noncognitive skills into a model of employer learning described originally by Altonji Pierret (2001). Measures of noncognitive skills include the Rosenberg Self Esteem Score, the Rotter Locus of Internal Control Score, the Coding Speed Score, and the CES-Depression Scale. I find that employers observe an initial signal of self esteem and schooling and that, over time, employers learn about cognitive skills and motivation, placing less emphasis on these initial observations. Does learning transfer perfectly across employers or is there a degree to which learning resets as employees change jobs throughout their careers? In this paper, I use data from the NLSY79 to look for evidence of asymmetric employer learning. I use tests developed by Schonberg (2007) and Pinkston (2009) to look for asymmetric learning in the model from Altonji Pierret (2001) augmented in Petre (2013b) to incorporate noncognitive skills in addition to cognitive skills. I find mixed evidence that learning done by a prior employer might not transfer completely to a new employer.

This paper establishes that a low dimensional vector of cognitive and noncognitive skills explains a variety of labor market and behavioral outcomes. For many dimensions of social performance cognitive and noncognitive skills are equally important. Our analysis addresses the problems of measurement error, imperfect proxies, and reverse causality that plague conventional studies of cognitive and noncognitive skills that regress earnings (and other outcomes) on proxies for skills. Noncognitive skills strongly influence schooling decisions, and also affect wages given schooling decisions. Schooling, employment, work experience and choice of occupation are affected by latent noncognitive and cognitive skills. We study a variety of correlated risky behaviors such as teenage pregnancy and marriage, smoking, marijuana use, and participation in illegal activities. The same low dimensional vector of abilities that explains schooling choices, wages, employment, work experience and choice of occupation explains these behavioral outcomes.

In my essays I focus on personality and its impact on the labor market outcomes. Using a randomized experiment, the first essay examines the impact of mother's human capital on the cognitive and non cognitive skills of her preschool children. The second paper examines the impact of the big five personality traits on the decision to be self employed and on the income of salaried vs. self employed people. We try to distinguish the impact of personality traits on labor market performance from the relationship between personality and preferences for entrepreneurship. In the third paper, we are trying to estimate the labor market wage premium for shift workers. We use an equilibrium sorting framework to model location decisions around the clock. Using the estimated model we try to disentangle the amenity value of daylight from social interaction effects.

Abstract: Results of research from the developed world suggests that noncognitive skills such as self-esteem or locus of control significantly affect key socioeconomic outcomes. More work is needed on whether studies of noncognitive skills in developing countries should use different measures of noncognitive skill, whether noncognitive skills have a different relationship with outcomes in developing countries, and whether programs can affect the formation of noncognitive skills. First, I briefly survey the measures which the developed and developing country strands of the literature use. I also draw on studies from psychology to argue that studies of noncognitive skill in developing countries do not require multidimensional measures of noncognitive skill, but that measures may need contextualization. Second, I look at whether a measure of noncognitive skill affects several youth labor force participation outcomes in India. Analyzing data from India using instrumental variables models with fixed effects (FE), I find evidence that noncognitive skills affect two of these outcomes. Finally, I look at whether participating in Head Start impacts child noncognitive skills. Using Children of the National Longitudinal Study of Youth (CNLSY) data and FE models, I find evidence of a negative impact.

"This thesis is comprised of three empirical essays on the theme of human capital. The essays use natural and laboratory experiments to study the determinants, returns and components of human capital. We first consider the determinants of human capital by studying the effects of maternal care as a determinant of children's human capital. Then we investigate the returns to human capital by studying the effects of immigration policies on immigrants' characteristics and labour market outcomes. Lastly, we examine specific components of human capital through an experiment on non-cognitive skills and preferences. The first essay estimates the causal impact of maternal care on the developmental outcomes of children aged 2-3 years using a parental leave reform implemented in Canada at the end of 2000 as an exogenous variation to maternal care. The reform increased the time mothers spend with their newborns by 3 months without affecting their income net of taxes, transfers and child care costs. Using the Canadian National Longitudinal Survey of Children and Youth, we employ a difference-in-differences methodology to compare children with a sibling born after the reform to those with a sibling born before the reform, relative to children of the same birth cohorts who did not have a younger sibling in the period surrounding the reform. We find that treated children enjoy a 16 percent increase in the time they spend with their mothers, with maternal care crowding out informal care. The increase in maternal care does not translate into better cognitive, non-cognitive or health outcomes in the short-run or the medium-run. The second essay uses a natural experiment to study the effects of a change in the point system, a system that selects immigrants based on specific observable characteristics, on immigrants' characteristics and labor market outcomes. Specifically, in 2001, Quebec changed its point system, by increasing the points for education and French language and decreasing the points for a subjective category "adaptability". The objective of the reform was to increase the number of French-speaking immigrants without deteriorating their labor market performance. Using a difference-in-differences and triple differences methodology, we show that, compared to immigrants to the Rest of Canada, immigrants to Quebec after the reform hold more bachelor's degrees and know more French than immigrants to Quebec before the reform. However, this does not translate into better labor market outcomes. This essay shows how point systems can be used to shape the immigrant workforce according to policy goals. Non-cognitive skills are a recently incorporated component of human capital in the economics literature. In the third essay, we contribute to this literature through a laboratory experiment on personality traits and risk and ambiguity preferences. We also study the effects of personality traits prevalence in a group on the decision making of each group member. In the experiment, subjects reveal their risk and ambiguity preferences through lottery choices. They then participate in an unstructured group chat. Afterwards, they are given the chance to revise their initial lottery choices. Results show that personality traits affect risk and ambiguity preferences before the chat. Specifically, conscientiousness is negatively related to risk and ambiguity aversion and agreeableness is negatively related to ambiguity aversion. We also show that the probability of changing decisions after the chat is affected by the individual's non-cognitive traits but not by the traits of the other group members." --