Dissertation Defense: Mariel Cecilia Siravegna
Candidate Name: Mariel Cecilia Siravegna
Major: Economics
Advisor: James Albrecht, Ph.D.
Title: Essays on the Gender Wage Gap and Intergenerational Mobility in the Presence of Sample Selection
A random sample from the population of interest is not always available to the researcher. As a result, any analysis based solely on the non-random sample may yield biased estimates and lead to misguided conclusions. In my thesis, I study the gender wage gap and intergenerational mobility indicators in the presence of sample selection.
In the case of the gender wage gap, a sample selection arises since a subset of women is working and reporting wages. I address this problem by using a new copula-based methodology to account for female self-selection into employment and to analyze the gender wage gap between men and women across the distribution of wages in Chile. Then, I generate a counterfactual wage distribution and decompose the gender wage gap into two parts: gender differences in the rewards to workers’ characteristics (structural effect) and differences in the distributions of those characteristics (composition effect). This decomposition provides insights into potential policies that might reduce the gender wage gap.
As a by-product of my research, I (co-authored) programmed a command in the statistical software Stata to apply this new methodology and wrote a short paper demonstrating its use with an example of gender inequality in the UK. This community-contributed command allows the user to model selection in quantile regression using either a Gaussian or a one-dimensional Frank copula. Lastly, in the case of intergenerational mobility (IGM) in education, the researchers require linked information about children’s and parents’ educational attainment to calculate the IGM indicators. However, several economies do not offer better data alternatives to estimate these indicators than the use of coresident samples (i.e., samples with this link only available for individuals living with their parents). In this line, we study the importance of coresidence bias in estimating intergenerational mobility indicators and to what extent this data limitation allows comparison across countries and/or regions.
The main goal of my thesis is to generate rigorous quantitative evidence about inequality in Latin America from a novel perspective that can help to uncover, better understand and deliver a guide for future policy recommendations.