A Comparative Study on Gender Discrimination and Its Indications: Principle and Facts from India |
Traditional analysis of gender wage gaps has largelyfocused on average gaps between men and women, and mean wage decompositionssuch as the Blinder-Oaxaca (1973) decomposition method. To answer the questionof whether there is a “glass ceiling” or a “sticky floor”, i.e. whether wagegaps are higher at the upper or lower ends of the wage distribution, this paperexamines the wage gaps across different quintiles of the wage distribution.These gender wage gaps are analyzed for regular wage workers in India using the66th round of the National Sample Survey’s Employment - Unemployment Schedule(2009-2010). The paper finds evidence of a “sticky floor”. In addition toestimating the standard OLS wage equations for men and women, quintileregressions are used to assess how different covariates such as education,union membership, and occupations, affect within and between group (gender)inequalities. Finally, the Machado-Mata-Melly (2006) decomposition method isused to decompose gender wage gaps at different quintiles to determine whetherit is the differences in characteristics (levels of covariates) or theunexplained (discrimination) component that drives the sticky floor effect. Thepaper concludes with a discussion on the possible reasons for observing asticky floor phenomenon in India. Gender inequality is an acute and persistent problem,especially in developing countries. This paper argues that genderdiscrimination is an inefficient practice. We model gender discrimination asthe complete exclusion of females from the labor market or as the exclusion offemales from managerial positions. The distortions in the allocation of talentbetween managerial and unskilled positions, and in human capital investment,are analyzed. It is found that both types of discrimination lower economicgrowth; and that the former also implies a reduction in per capita GDP, whilethe latter distorts the allocation of talent. Both types of discriminationimply lower female-to-male schooling ratios. We discuss the sustainability ofsocial norms or stigma that can generate discrimination in the form describedin this paper. We present evidence based on panel-data regressions acrossIndian states over 1961-1991 that is consistent with the model’s predictions.