Chin, Sayorn, authorZahran, Sammy, advisorMushinski, David, advisorMiller, Ray, committee memberBayham, Jude, committee member2023-08-282023-08-282023https://hdl.handle.net/10217/237002This dissertation explores several topics in welfare, well-being, and labor economics, with a focus on: (1) health, wealth, and racial and ethnic welfare inequality; (2) the natural environment and well-being; and (3) whether labor markets place a wage premia for jobs that require workers to consume disamenities. To achieve these goals, the study utilizes three distinct datasets and applies a range of machine learning and econometric techniques, including natural language processing algorithms, as well as dynamic panel data estimators, natural experiments, and microsimulations. In Chapter 1, titled "Beyond Income: Health, Wealth, and Racial/Ethnic Welfare Gaps Among Older Americans'', we estimate racial and ethnic disparities in well-being among the older U.S. population using an expected utility framework that incorporates differences in consumption, leisure, health, mortality, and wealth. We use longitudinal data from the Health and Retirement Study (HRS) supplemented with data from the Consumption and Activities Mail Survey (CAMS). Together, these provide a long and rich panel (1992-2016) for our analysis. Our measure broadly indicates that racial and ethnic inequality is larger than suggested by other welfare metrics such as income or consumption. We also find health, mortality, and wealth gaps are important in explaining the level of racial and ethnic welfare inequality among the older Americans in our sample, with leisure playing a comparatively minor role. Our decomposition exercises show that a majority of the estimated welfare gaps are determined by age sixty initial conditions as opposed to racial and ethnic differences in dynamic processes after age sixty. Our morbidity counterfactuals further suggest that eliminating common heath risk factors such as hypertension or diabetes in late-life only marginally closes overall welfare gaps. These simulations suggest that policies aimed at closing racial and ethnic gaps in late-life may be more successful and efficient if targeted earlier in the life-cycle. In other words, outside of direct wealth transfers, it may largely be too late to target such interventions directly at older populations. In Chapter 2, titled "The Morning Advantage: Differential Returns to Sunlight Exposure on Well-Being'', we estimate the effect of sunlight exposure on well-being by mimicking a natural experiment that utilizes the transition to daylight savings time as an external shock to the reallocation of sunlight between the morning and evening induced by differences in sunrise and sunset times across space, and time. We combine a collection of geolocated and timestamped tweets from Twitter with Natural Language Processing algorithms to create a comprehensive panel dataset of well-being (2014-2022) for the United States. Our findings show that the returns to sunlight on sentiment are stronger in the morning than in the evening. These results contribute significantly to the ongoing debate about whether to continue or abandon the practice of daylight savings. Specifically, the positive turn of sentiment in the morning highlights the underappreciated benefits to human well-being. Therefore, the potential shifting to darker mornings and brighter evenings following the proposed Sunshine Protection Act may do more harm than good. In Chapter 3, titled "The Compensation of Conscience: Evidence from the U.S. Labor Market'', we investigate compensating differentials in the U.S. labor market related to the degree of moral compromise required in different occupations. Specifically, we explore whether jobs that require workers to compromise their moral values offer higher compensation to compensate for the disamenities that contradict their moral beliefs. To conduct our analysis, we utilize data from the National Longitudinal Survey of Youth 1997 (NLSY97) and supplement it with data from the Occupational Information Network (O*NET) job descriptor, which allows us to develop a continuous measure of moral index across occupations. This data provides a rich and extensive panel spanning from 1997 to 2017 for our analysis. Our findings, obtained through the use of two-ways fixed-effects and first-difference models, indicate that jobs that require workers to compromise their moral principles are associated with higher compensation. This suggests that there is indeed a compensating differential for engaging in disamenities that conflict with a worker's moral values. Additionally, we observed that workers with a college education receive higher pay in jobs that require moral compromise, indicating that individuals with a college degree may have more employment opportunities and greater bargaining power, influencing their compensation preferences. Furthermore, we discovered evidence supporting an asymmetric relationship between changes in the occupational moral index and total hourly compensation. This relationship appears to be responsive to the intensity of moral compromise in the job.born digitaldoctoral dissertationsengCopyright and other restrictions may apply. User is responsible for compliance with all applicable laws. For information about copyright law, please see https://libguides.colostate.edu/copyright.inequalityNLPwell-beingmoralitycompensating differentialswelfareText