Essays on the Social and Parental Determinants of Early Child Development
Pham, Van Ngoc Thu
Doctor of Philosophy
The thesis includes our works on measuring inequalities in parental investment captured by a variety of environmental conditions, such as home, language, and child care environments. Chapter one explains the importance of research in inequality in early human capital. The consequence of this disparity includes gaps in skills, education, employment, and earning. In addition, this chapter highlights how environmental contexts, such as home, language, and child care environments, can be used to measure parental investment and consequently influence child development. The second chapter estimates inequality in parental investments in children as measurements through the Home Observation for the Measurement of the Environment (HOME) and the language environment as measured by the Language Environment Analysis (LENA). We contribute to the literature in several ways. First, we introduce a new form to estimate the HOME Score based on Item Response Theory. Our results suggest that the HOME can identify low from not-low levels of investments, but does not have enough power to separate medium from high levels of investments. Second, we develop a model for the estimation of the quality of language environment that controls for measurement error and hour-by-hour fluctuation. The measures of quality of language environment are based on the Adult Word Counts (AWC) and Conversation Turn Counts (CTC) reported hour-by-hour by the LENA System. Armed with these three measures, we are able to investigate not only the correlations across measures, but also how to produce estimates of quality of adult-children interaction that minimizes measurement error. In the third chapter, I aim to evaluate the impact of maternal care and out-of-home day care on child development. Doing so requires the estimation of the causal impact of childcare type on developmental outcomes. The empirical challenge is to account for why mothers choose a specific type of child care arrangement. To correct this situation, I apply the Generalized Roy model by constructing maternal choice sets, modeling the choice explicitly, and estimating the child outcomes conditional on the selection. Using a sample of 7-9 months old infants from Philadelphia, I evaluate child development by using different developmental dimensions, Motor-Social Development (MSD) and temperament. I also provide information about actual average price and state-evaluated quality of the day care providers that are located within close proximity of the families from the sample. These cost and quality indicators play a crucial role in the mothers' decision and child outcome models. By applying the consistent estimators from the Generalized Roy model to produce the average treatment effect (ATE), the results suggest that infants who receive day care tend to have lower scores in MSD and higher scores in temperament than infants who are cared by their mothers. Furthermore, I construct a counterfactual exercise by setting the average full-time price at zero to detect any change in maternal decision and consequently infant's outcome scores. Initial results suggest that there is no change in MSD scores when the infants receive the nonmaternal care instead of maternal care. However, the results suggest a positive increase in temperament score among the infants after the price reduction. This result is consistent with the ATE estimate.
Labor Economics; Child Development