Sand, Benjamin M.Zhu, Shengyi2023-12-082023-12-082023-12-08https://hdl.handle.net/10315/41715This dissertation provides a comprehensive analysis of the Canadian gender wage gap over the past two decades, employing modern methodologies and tools. In the first chapter, selection bias and its impact on the entire earning distribution are examined. A selection-corrected quantile regression is utilized to provide a more accurate depiction of the gender wage gap distribution. The simulation of potential government child care benefits as an instrument helps address the selection bias issue. Findings reveal the persistent but inconsistent effects of selection bias across wage quantiles and time. The presence of negative selection for women entering the workforce is identified, and the absence of this bias would result in an even higher unexplained portion of the gender wage gap. Moving to the second chapter, a thorough investigation of the heterogeneity of the gender wage gap is conducted. High-dimensional models are employed to explore the diverse factors contributing to the wage gap. Advanced machine learning algorithms are utilized as robustness checks to address possible multicollinearity problems. The analysis reveals significant reductions in the gender wage gap attributed to age and several occupations, while penalties related to family structure persist. Finally, the third chapter explores the under-researched area of job-education mismatch and its impact on the gender wage gap. The study focuses on the differences in vertical and horizontal matching between women and men. Self-reported measurements of both vertical and horizontal mismatch, as well as an objective index of horizontal mismatch, are utilized. Results indicate that, unlike other countries, vertical mismatch does not contribute significantly to the gender wage gap. Furthermore, the role of horizontal mismatch is economically insignificant in relation to the overall gap. This dissertation enhances our understanding of the Canadian gender wage gap by addressing important aspects such as selection bias, heterogeneity, and job-education mismatch. The findings contribute to the existing literature on gender inequality, offering valuable insights for policy interventions and future research in this field.Author owns copyright, except where explicitly noted. Please contact the author directly with licensing requests.EconomicsElectronic Thesis or Dissertation2023-12-08Gender wage gapWage gapCanadaSelection biasQuantile regressionMismatchHorizontal mismatchVertical mismatchMachine learningHeterogeneityMatchingSelection correctionSelectionDecompositionEconomicsLASSORandom forestBoosted trees