Gender, Parenthood, and Parental Leave: Labor Market Inequality in Kazakhstan from Observational and Experimental Evidence

Abstract

Kazakhstan presents a striking paradox. Women outnumber men in tertiary education and participate in the labor force at high rates. However, the official gender wage gap exceeds 20% and female representation in senior positions remains low. This dissertation investigates gender inequality in Kazakhstan's labor market through three interconnected empirical studies, combining observational wage analysis with experimental investigation of hiring discrimination. Paper 1 uses three waves of the Kazakhstan Labor Force Survey (2013-2015, N = 280,411) and applies Mincerian wage regressions, Oaxaca-Blinder decomposition, and quantile regression to examine the gender wage gap across the income distribution and across sectors. Women earn approximately 11% less than observationally equivalent men, with roughly 70% of this gap unexplained by observable characteristics such as education or work experience. Fathers of young children receive a wage premium while mothers face an additional penalty. Sectoral heterogeneity is substantial. For instance, women in mining face gaps five to six times larger than women in education. Occupational ranking analysis extends analysis to 2001-2023 and shows persistent vertical segregation across the full period. Paper 2 and Paper 3 present the first vignette experiments on hiring discrimination in Central Asia. Paper 2 tests discrimination based on candidate gender and parental status using a 2×2 between-subjects design (N = 171). Paper 3 shifts the question to maternity leave by varying leave duration at 6 months, 1.5 years, and 3 years (same N = 171). Both studies use the multidimensional model of agency provided by Ma et al. (2022) to measure how candidates are perceived on agentic traits such as competence, ambition, diligence, dominance, independence and self-assurance. The standard Western motherhood penalty does not replicate. Female candidates are not rated lower on competence or commitment and receive higher ratings on independence and commitment. However, these trait advantages do not translate into higher hiring recommendations. Extended leave of three years produces significantly lower hiring recommendations, mediated through reduced competence perceptions and reduced work priority perceptions. Male evaluators who are parents drive this penalty; female non-parent evaluators show no significant effect. These results suggest that Western discrimination theories have contextual boundaries and should not be universally applied. They also indicate that extending parental leave entitlements without addressing how prolonged absence is evaluated by employers may have unintended consequences and undermine the very goals such policies are designed to achieve.