This analysis has been undertaken by the members of InfoSphere team of Centre for New Economics Studies, O.P. Jindal Global University. To review their work, visit their website here.Social identity plays a significant role in determining the type of employment a person is likely to be engaged in and their earnings across India, said a recent report by the Centre for Sustainable Employment at Azim Premji University.The report found that regular wage or salaried work tends to be the most remunerative, followed by self-employment and then casual wage work. Further, within each type of employment, significant heterogeneity exists in occupation, industry, and other job characteristics, all of which play a crucial role in determining labour earnings.Social disparities in labour earnings across social identities. Source: APU Report-InfosphereThe report found that significant earnings gaps exist in terms of gender, caste and religion among different social groups.The largest earnings gap is based on gender, with women reporting earnings at only 40% of those of men in self-employment, followed by 64% of men in casual wage jobs, and 76% of men in regular wage jobs.The gaps based on caste and religion are narrower but still significant. Muslims receive higher pay than Hindus in the category of casual wage labour but lower pay in terms of regular wage jobs and self-employment. There is a noticeable difference in earnings between SC/ST and other castes.Source: APU Report-InfosphereThe gender earnings gap is further broken down by deciles of the earnings distribution. As we move up the income distribution, the gender gap reduces for salaried or regular wage workers. One possible explanation for this trend could be that higher educated women have better bargaining positions, and discrimination is less prevalent in formal and regulated jobs. It also finds that the gender gap is much higher in self-employment.One of the reasons is that self-employed women are more likely to be engaged in lower-paying sectors. Another possibility is that self-employed women face more discrimination as compared to women doing regular wage work. This discrimination may be prevalent due to several factors, such as social norms that discourage women from starting their own businesses, or lack of access to credit and other resources.Findings from the report also suggest that the largest gender pay disparity is found among the lowest income group.This suggests that gender discrimination is most pronounced in the bottom quartile of the income distribution. Several factors may contribute to this, including that disadvantaged women are likely to work in informal, unregulated industries with weaker anti-discrimination laws.Source: APU Report-InfosphereThe report focused on earnings gaps based on caste, gender, and religion in regular wage work, and analysed them on a decile-by-decile basis.It showed that religion and gender-based pay gaps move in the opposite direction. Women at the upper end of the income distribution are less disadvantaged than men. However, Muslim workers at the upper end of the income distribution are more disadvantaged as compared to Hindu workers.At the top decile Muslim workers earn 75% of what Hindu workers earn. For the third and fourth decile, this number is much higher at around 94%, indicating that Muslim salaried workers earn almost as much as Hindu salaried workers at the lower end of the earnings distribution.Conversely, women at the bottom of the income distribution only earn 40% of what men earn but at the top decile the number is 93%.The increasing Hindu-Muslim gap for higher-paid workers is a matter of significant concern.Source: APU Report-InfosphereThe report shows that the gender earnings gap was highest in the first quartile and lowest in the fourth quartile in 2004 (see the graph above). This suggests that gender discrimination was most severe among the poorest earners. The gender earnings gap narrowed in all quartiles between 2004 and 2017. This is particularly evident in the second, third, and fourth quartiles.There are a number of possible explanations for the narrowing of the gender earnings gap, such as increased awareness of gender discrimination, anti-discrimination laws, and changes in social norms.Another possibility could be that women’s educational attainment has increased significantly in recent decades, making them more competitive in the labour market. There has also been an increase in the number of regular wage jobs available to women.Moreover, a structural transformation has caused a significant change in the female workforce composition. It is observed that older and less educated women are moving out of the workforce, while younger and more educated women are increasingly coming on board. The change in women’s earnings can be attributed to improvements in workplace discrimination against women as well as shifts in the characteristics of those who are employed.Thus, it can be argued that wage gaps based on gender can be bridged on two accounts: firstly, by reducing discrimination against women in the workplace, and second, giving them access to higher education and skilled qualifications, making women professionally more sought after.To understand the ‘unexplained differences’ causing the wage gap among two groups, the APU report’s authors employed the Blinder-Oaxaca decomposition.It has two main components – the “explained” component, which includes characteristics such as education, work experience, job tenure, etc., and the “unexplained” component, which includes factors such as discrimination, bias, or unobservable characteristics that affect wages.This method highlights how the “unexplained” portion of earnings is greater for women workers than SC/ST workers or Muslim workers in any type of work. Interestingly, the “unexplained” earnings gaps are much larger based on gender than caste. This is because women, especially in the private sector, face more discrimination based on gender than caste.Caste discrimination is more prevalent in the public sector vis à vis the private sector since the latter runs on the principle of meritocracy as opposed to positive discrimination to empower groups. However, this unexplained gender disparity is more pronounced in rural areas as compared to urban areas. There is also a significant difference in the explained and unexplained reasons behind the gendered wage gaps, as can be seen in the table below.YearExplainedUnexplained19833664199341.458.6200442.357.7201131.568.5201716.583.5202118.981.1Source: APU Report-InfosphereOccupational segregation Occupational segregation occurs when one demographic group is over-represented or under-represented in a certain job category. Segregation in India operates both vertically – where dominant and marginalised groups clustered in two ends of the occupational ladder, and horizontally – where dominant and marginalised groups have different tasks and differential wages within an occupation category.The report uses a Duncan index which takes values between 0 and 1 where 0 represents no segregation and 1 represents complete segregation. The value of the index indicates the proportion of the workforce that needs to be reallocated between industries to achieve a state of no segregation.According to the Duncan Index, industrial segregation is highest along gender lines, followed by caste and religion. Source: APU Report-InfosphereAmong the different social identities examined in the analysis, the highest level of industrial segregation in the Indian workforce was observed between men and women.The segregation index between men and women in 1983 was 0.34 which increased to reach a value of 0.41 in 2021. Caste-based industrial segregation between SC/ST and others, including OBCs, remained almost the same, declining slightly between 2004 and 2021. Industrial segregation between Hindus and Muslim was even lower at around 0.17 to 0.21. This means that 40% of the workforce would need to be redistributed along gender lines across industries to achieve gender parity.The early post-reform era between 1993 (0.35) and 2004 (0.4) saw the majority of the growth in gender-based segregation across industries. However, throughout the course of the past four decades, women have consistently outnumbered males in a number of industries, such as in the production of tobacco products, domestic work (maids, etc.), health and social work, education, and textile manufacturing.Before experiencing a modest decline by 2021, the proportion of women in the tobacco sector reached its peak in 2017, surpassing five times their representation in the overall workforce.The over-representation of women in the waste management and sewerage sector has decreased, nevertheless, from an index value of 2.4 in 1983 to 1.2 in 2021, which is a positive development. Caste-based industrial segregation between SC/ST and others, including OBCs, remained constant at a Duncan index value of 0.23 in 1983 and 2021, with marginal changes in the years in between.The people from the Scheduled Caste community have remained consistently over-represented in waste-management and sewerage and manufacture of leather and leather products. These two industries have strong historical ties with their caste identities. However, the extent of over-representation has declined substantially. In the case of the leather industry, the representation index declined from 4.6 in 1983 to 1.4 in 2021.In case of waste management and sewerage, over-representation of SCs decreased from over five times their workforce share in 1983 to 1.6 times in 2011, with the largest drop occurring between 2004 and 2011, which was also the high growth period. However, in 2017, it increased to 2.1 and has remained above twice their workforce share in 2021.The industrial segregation between Muslims and Hindus is relatively less-pronounced than gender and caste-based segregation and it has also declined over time. The segregation index declined from a peak of 0.22 in 1993 to 0.17 in 2021. Among industries in which Muslims are consistently over-represented since 1983, their over-representation was the highest in the manufacture of tobacco, garments, and textiles in 2021.Entrepreneurship and casteIt is interesting to see how the APU study highlighted the disparity among different castes with respect to the ownership of enterprises.People belonging to the Scheduled Caste and Scheduled Tribe communities own very few enterprises as compared to their share in the population in urban and rural areas. A wage gap exists between people belonging to the general category and others much like the gap between men and women. This has also been seen in the case of MSMEs.Social GroupsShare in WorkforceShare in enterprise Ownership199820052013199820052013ST10.71010.14.44.15.4SC20.620.119.38.49.811.4OBC36.641.443.536.342.441Others32.228.527.150.943.742.1Source: APU Report-InfosphereIn the table above, a stark difference can be seen in the ownership of enterprises among the Scheduled Castes and Scheduled Tribes as compared to the OBCs and the general category members.This data is useful in highlighting the elite predominance that continues to remain across sectors, be it in terms of employment or ownership. However, the report falls short in distinguishing industries that are owned by different social groups and the kind of people employed by them. It still highlights the continued stronghold of elite groups in the entrepreneurial sector.In summation, the report brings to light the stark disparities between the privileged and underprivileged classes, on the lines of caste, gender, and religion.Wage gaps have a significant impact on people’s lives, leading to poverty, inequality, and social exclusion, and closing this gap requires comprehensive policy-based interventions.Amisha Singh and Aditi Desai are research analysts with CNES and team leads for its InfoSphere initiative.Deepanshu Mohan is professor of Economics and director at the Centre for New Economics Studies, O.P. Jindal Global University. He is an Honorary Research Fellow with Birkbeck, University of London (2023) and a Visiting Professor at the School of International Development and Global Studies, University of Ottawa.
