EDUCATIONAL OUTCOME AND EARNINGS DIFFERENTIALS IN INDIA: DYNAMICS OF CASTE

The study uses nationally representative unit level secondary data from Indian Human Development Survey (IHDS) collected first in 2004-05 and again in 2011-12. To analyze caste based discrimination in earnings and job opportunities and effect of increasing education and its return for different occupations, the study estimated first, linear regression analysis, followed by logit regression and finally decomposition techniques. The earning function clearly support the existing evidences of significant positive coefficients for education and marginal wage effects are increasing with the level of education for every social group, but return to education for scheduled caste was very low compared to forward caste. There are clear evidences of gender and location bias in earnings too. Our decomposition results show that endowment component is much larger than the discrimination component, because of existence of pre-market discrimination in India. Discrimination explains 10.9 per cent of the lower wages of scheduled caste individuals as compared to forwards in 2011. Over the two periods, the discrimination component has been decreased significantly for rural areas but for urban areas it has been dramatically changed to a negative unexplained component showing in urban areas no such evidences of discrimination in wage differentials is being found. Discrimination component was negative (-14.0 per cent and -18.4 per cent) for professional/administration and clerical/sales and services activities respectively and no such caste discrimination was found against scheduled castes (due to negative discrimination component). But discrimination component is highest at 59.8 per cent for agricultural activities; followed by production (19.9 per cent).