Caste, Class, Gender in the Indian Economy

Caste, Class, Gender: Dynamism or Stasis?

Introduction

• The Indian economy has undergone rapid changes due to globalization and liberalization in the last three decades.
• Analyzing the changes in social divisions in response to the growing integration of the Indian economy with the global economy is complex.
• This chapter focuses on the economic dimensions of caste, class, and gender disparities.
• Other critical identities like religion and region also significantly shape economic outcomes and group antagonisms.
• The focus on caste and gender is due to affirmative action programs in India being designed along these lines.
• Due to space constraints, other social identities and in-depth assessments are precluded.

Impact of Liberalization on Inter-Group Disparities

• Argument for Reduction in Disparities:
  • Liberalization and market orientation may reduce inter-group disparities.
  • Labor market discrimination based on social identity is expected to decline.
  • New job creation will provide opportunities for marginalized groups to acquire new skills and move out of traditional occupations.
  • Market orientation prioritizes profit maximization and efficient resource allocation, making workers' social identity irrelevant.
  • Foreign agents are believed to be less concerned with local hierarchies and prejudices.
  • Globalization will boost certain sectors, creating demand for skilled labor.
• Counter-Argument:
  • Marginalized groups may be at a disadvantage due to historical exclusion from education.
  • Multinational corporations (MNCs) operate with local managers who share local social biases.
  • Foreign agents may not disrupt domestic norms unless essential for profit.
  • Economic theory suggests profit maximization and discrimination are not incompatible (Arrow 1971, Becker 1957, Akerlof 1984).
• Empirical Evidence:
  • Research indicates managers are aware of caste, class, religion, and gender, and believe merit is distributed along these lines (Deshpande and Newman 2007; Jodhka and Newman 2007).
  • This is true for both private domestic firms and MNCs.
  • Cross-national comparisons show no straightforward relationship between a country's prosperity and the severity of disparities.
  • Intergroup disparity is persistent across all levels of development (Darity and Deshpande 2003).
  • Higher economic growth rates do not always close economic gaps between racial and ethnic groups.
  • India's high growth rate post-liberalization may not be sufficient to close existing gaps.
  • Intergroup disparity is not systematically related to growth or lack thereof across different states in India (Deshpande 2017).

Contemporary Group Divisions: Caste, Class, and Gender

• Brief assessment of economic aspects of caste, class, and gender divisions.

Caste

• Debate over the continued relevance of caste in shaping economic outcomes.
• Question of whether public discourse on caste perpetuates its existence (Beteille 2012).
• Contrast between discussions on religion and caste: legitimacy of religion not questioned despite conflicts.
• All data sets collect data for individual religions but not for castes.
• Data on caste categories is dictated by the needs of affirmative action policy: Scheduled Castes (SCs), Scheduled Tribes (STs), Other Backward Classes (OBCs), and Others.
• Until 2011, the Indian census did not count OBCs separately, despite affirmative action policies for them since 1991.
• The 2011 census conducted a socio-economic and caste census, but the data are not yet available.

Caste Disparities and Discrimination

• Primary source of economic disparities between castes is the traditional link between caste and occupation.
• The caste system has undergone mutations, with the operative form being jati, which shares characteristics with the ancient varna system.
• Occupational distribution in the varna system indicates a rudimentary economy; the jati system is more complex.
• Occupational divisions have transformed with modernization, but mutations in jati divisions have not kept pace.
• Since Indian Independence in 1947, it is illegal to deny a job based on caste.
• Examination of the link between caste and occupation:
  • Are traditional caste-based jobs allocated randomly?
  • Is there overlap between the traditional caste hierarchy and the modern occupational spectrum?
• Data indicate persistent overlap: lower castes are disproportionately represented in lower-paying modern occupations, and vice versa (Deshpande 2017).
• Disparities persist in occupational attainment between SCs (or STs) and Others.
• Disparity based on this broad comparison actually understates the disparity between the top and bottom ends of the caste hierarchy.
• Given that income data are notoriously difficult to collect in India differences in education, occupation, consumption expenditure, and wages are examined as standard indicators of the material standard of living.
• Hnatkovska et al. (2012) find convergence between SCs and Others in educational and occupational attainment, consumption, and wages, driven by narrowing educational gaps.
• However, picture is complicated: educational attainment at lower levels is converging, but gaps from secondary education onwards are not converging.
• Deshpande and Ramachandran (2016) documented changes in caste disparities over time by splitting the data into age cohorts and comparing the average outcomes between cohorts, across castes, and time.
• Caste groups show convergence in literacy rates and primary and middle school completion rates.
• For education levels higher than middle school, divergence between caste groups is observed, reflected in divergence in access to prestigious white-collar jobs and consumption expenditure.
• A ‘caste development index’ (CDI) based on five indicators of standard of living (educational attainment, occupational attainment, land ownership, livestock ownership, and consumer durable ownership) showed a rough measure of inter-caste disparity (Deshpande 2001, 2011).
• Table 8.1 (Deshpande 2011: 90–1) illustrates the lack of a clear pattern in the relationship between state material prosperity, SC material well-being, and inter-caste disparity.
• Tamil Nadu ranks number 6 in per capita real SDP but had a disparity ranking that increased from 11 to 1.
• Uttar Pradesh, one of the poorest states, saw inter-caste disparity steadily increase.
• The relative prosperity of a state does not predict inter-caste disparity or SC absolute development level.
• Deshpande (2011: 93) reports correlations between the CDI-SC and the per capita real net SDP, which is positive and has become stronger over the three NFHS rounds, from $0.09$ to $0.72$, and between the CDI-SC and the rate of growth of the per capita real NSDP, which was very low but positive in the first round, but turned negative in the latter two rounds, with the strength of the negative correlation declining from $0.09$ to $-0.27$ to $-0.16$.
• The correlation between disparity and the per capita NSDP was positive in the first round ($0.54$), but declined to an insignificant $-0.03$ in the third round; the correlation between disparity and the rate of growth of the real per capita NSDP fluctuated in direction and magnitude and was ambiguous overall.
• Deshpande (2011) also discusses the relationship between overall inequality (as measured by the Gini coefficient calculated over the monthly per capita expenditure (MPCE) and both the CDI-SC and disparity.
• States with higher disparity not only have a higher Gini but a higher CDI-SC too.

Foreign Direct Investment and Intergroup Disparities

• Impact of liberalization and globalization on inter-group disparities examined through foreign direct investment (FDI) data.
• Disaggregated data available only for FDI approvals, not realized FDI.
• Examination of FDI data reveals spatial clustering.
• New jobs created are limited to geographical clusters, limiting access to the local population.
• The NCAER (2009: 18) suggests India’s attractiveness for investment rests on:
  • A large and growing market.
  • World-class scientific, technical, and managerial manpower.
  • Cost-effective and highly skilled labor.
  • A large English-speaking population.
• Sectoral distribution of FDI projects confirms labor demand for workers with specific education, skills, and English knowledge.
• Existing educational disparities by caste suggest a large section of Dalits and women would not be prime candidates for these jobs.
• Research suggests Dalits face discriminatory barriers in urban, formal sector labor markets even after acquiring higher education (Deshpande and Newman 2007).
• Growth in certain jobs could incentivize higher education and English skills over time.
• This could lead to further clustering of FDI unless the dynamic spreads to other areas.
• Deshpande (2018) finds that, based on district level data for FDI approvals between 1995 and 2010, a greater number of FDI approvals were associated with lower relative wages of Dalits versus Others.
• This result is obtained controlling for the relative proportions of Dalits versus Others with secondary education or more, and the relative proportion of those in regular wage/salaried jobs, as well as district-specific fixed effects.
• Earnings gaps are more likely to widen rather than diminish, controlling for other factors.

The Larger Impact of Globalization

• Creation of new job types through FDI is one channel affecting caste disparities.
• Globalization may empower Dalits by ‘universalizing access to aspirations.’
• Consumer goods are becoming plentiful and cheaply available to Dalits, breaking the exclusivity of the ‘predominantly Upper Caste Consumer Club.’
• Rise of Dalit entrepreneurs.
• Omvedt (2005) argues that globalization is inevitable and has many positive aspects, rather than taking a blind anti-globalization position and we need to think about a comprehensive and nuanced response of Dalits and Adivasis to these changing realities.
• Dalits are asked to view globalization as an opportunity rather than as a threat.
• Kapur et al. (2010, 39) argue that the shift from a state-led to a market-led path of development has changed patterns of interaction between castes, leading to a ‘rapid erosion in discriminatory practices that stigmatized Dalits’.
• Large shifts in the pattern of economic life, both within and away from the village.
• Change in occupational patterns, with Dalits moving away from traditional occupations.
• Changes in the nature of contracts between Dalits and upper castes, with practices such as bonded labor becoming obsolete.
• Extent to which the study of two blocks in one state can be generalized to the rest of the country is not immediately apparent.
• Significant changes in consumption patterns in contemporary India over the last two to three decades.
• Another channel through which liberalization will have implications for caste disparity is privatization.
• Affirmative action in India takes the form of quotas for the SCs, the STs, and the OBCs in public sector jobs and government-run educational institutions.
• As the economy privatizes and more jobs get created in the private sector, the scope of affirmative action in its present form is likely to shrink.
• Deshpande (2013) discusses the impact of the affirmative action program in detail.
• Overall, the impact of globalization on caste disparities is complex and contains many interwoven strands.
• The debate over whether globalization will be favorable to Dalits is not backed by hard empirical facts.
• More empirical work is needed to etch out the contours of the multidimensional impact of globalization on caste disparities.

Class

• Rapid economic growth has led to expansion in the group of ‘high net worth individuals’ (HNIs, individuals with investible assets of USD 1 million or more).
• India’s HNI population grew at $20.8$ per cent to $1,53,000$ in 2010 compared with $1,26,700$ in 2009, according to the 2011 Asia-Pacific Wealth Report by Merrill Lynch Global Wealth Management and Capgemini.
• The cumulative wealth of Indian HNIs grew by $22$ per cent in 2009–10 to INR $28,60,000$ crore from a year ago.
• Globally connected individuals who enjoy lifestyles comparable to the elite in developed countries, and visibly shape patterns of conspicuous consumption in urban India, especially in metropolitan cities.
• The growth of this elite segment should be separated from the growth of the Indian middle class.
• Estimates of the Indian middle class vary from $25$ to $300$ million.
• Variations in estimates come from different databases and how the middle class is defined.
• The middle class, by definition, should be defined as having an income within some interval which includes the median.
• A common way to define the middle class in developed countries is those having incomes between $75$ and $125$ per cent of the median of that country.
• Ravallion (2009: 6) suggests a ‘developing world’s middle class’ defined as those who are not poor by the standards of developing countries but might be poor by the standards of developed countries.
• Lower bound of USD $2$ a day, which is double the official poverty line of India (and also China).
• Upper bound is USD $13$ a day, which is the US poverty line at the 2005 purchasing power parity (PPP).
• A person will have entered the ‘Western middle class’ if earning more than the US poverty line.
• Using this definition for India, based on the National Sample Survey (NSS) data, he finds only $3$ million people not poor by the US standards.
• Correcting for possible under-reporting of consumption expenditure in the NSS data by increasing all consumption expenditures by $50$ per cent, the count of Indians who are not poor by the US standards rises to only six million (Ravallion 2009: 9).
• Growth of the segment which he defines as the ‘developing country middle class’ (proportion living between USD 2 and USD 13 a day) has been substantial between 1990 and 2005: an extra 117 million people joined this group in India.
• This is linked to a reduction in poverty, as the bulk of the increase in the middle class came from an increase in the USD 2 to USD 6 segment, rather than from an increase in the upper middle class.

Inequality and Class Conflict: Growth of the Maoist Movement

• The other side of the Indian growth story is rising inequality reflected in the growing class discontent.
• This class discontent found expression through the rise of the Maoist movement that has spread to several states in India, including Bihar, Jharkhand, Chhattisgarh, Madhya Pradesh, Orissa, Maharashtra, and Karnataka.
• This movement, christened ‘left wing extremism’ (LWE) by the Indian state, poses a challenge significant enough for the former Prime Minister Manmohan Singh to express concern.
• The rapid spread of the movement over large parts of central and southern India reinforces the need for a rigorous assessment of both the causes and consequences of this conflict.
• Borooah (2008) finds that the probability of Naxalite activity in a district increased with a rise in its poverty rate, and decreased with a rise in its literacy rate.
• Districts with a smaller coverage of safe drinking water were more likely to have Naxalite activity.
• High poverty and lack of development are factors contributing to Maoist violence.
• Iyer (2009) finds a clear divergence between ‘leading’ and ‘lagging’ regions in the post-2001 period in South Asia.
• Incidents of violence increased sharply in the lagging regions, while they remained steady in the leading regions.
• Regions are defined as leading or lagging based on per capita income levels in 2004.
• According to this definition, the lagging regions in India were Assam, Bihar, Chhattisgarh, Jharkhand, Madhya Pradesh, Manipur, Mizoram, Meghalaya, Nagaland, Orissa, Rajasthan, Uttar Pradesh, Uttarakhand, and West Bengal.
• This trend further intensified after 2004, with lagging regions displaying a continued increase in conflict and leading regions showing a slight decline.
• Counter-terrorism activities proxied by the number of terrorists killed by security forces have been declining in leading regions after 2005, but violence continues to rise in lagging regions.
• Iyer finds that conflict is higher in poorer regions.
• An increase in poverty by $10$ percentage points is associated with $0.26$ incidents of conflict per district.
• This is quite high compared to an average of $0.77$ incidents per district.
• A $10$ percentage point increase in poverty is associated with $0.39$ more deaths for an average district (compared to the mean level of $0.60$ deaths per district).
• There is a significant positive relationship between intensity of conflict and the extent of mountain and forest cover.
• Areas with lower historical land inequality have a lower incidence of conflict, but the relationship is not statistically significant, and the presence of disadvantaged communities is not significantly associated with conflict intensity.
• Evidence suggests that poverty, illiteracy, or (more broadly) underdevelopment, combined with inequality in land ownership are the key factors responsible for the sustained growth of the Maoist movement.
• Forested terrains facilitate guerrilla activity and these are the regions where tribals form a substantial proportion of the population, but where there is disaffection due to lack of control over forest land and/or mineral wealth.
• Unanswered question of whether the movement is working as a Robin Hood state or more like an extortionist mafia.
• Debate about whether the Maoist strategy of individual annihilation will fundamentally alter the material conditions of the marginalized.
• Sufficient material deprivation exists to fuel further growth of Maoism.

Official Response to the LWE

• Official references regard LWE as an extremist menace, with frequent references to ‘Naxalite-infested’ areas.
• The government views the movement as causing deprivation in order to sustain its ideology.
• Mainstream official view is towards the elimination of the movement through force.
• Increased Maoist activity has been met by increased police repression, resulting in heavy casualties on both sides.
• Since 2005, the Chhattisgarh state government has sponsored a counter-insurgency operation called ‘Salwa Judum’ which has killed several suspected Maoists and displaced thousands of people.
• There have been several reports of the terror unleashed by Salwa Judum, including sexual attacks on women.
• State-sponsored counter-violence, bypassing the due process of law and justice, is now a matter of serious concern.
• Special counter-insurgency forces targeted towards the Maoists have been created.
• In Andhra Pradesh, an elite security force called the Greyhounds was launched in 1989.
• The Greyhounds operate very much like the Naxalites’ own squads and are ‘bound by no law, including the constitution of India’ (Harris 2010: 12).
• The approach is one of suppression by sheer force, reflecting the state’s attitude that regards Maoism as a menace or, at best, a ‘law and order’ problem.
• Attempts have been made through the appointment of commissions, most notably through the formation of an expert group set up by the Planning Commission in 2008, to examine the ‘Development Challenges in Extremist Affected Areas’ (Planning Commission 2008).
• The report of this group comprehensively outlines the multi-faceted sources of discontent in areas of Maoist activity, with a detailed outline of the multiple class–caste–gender disadvantages of Dalits, Adivasis and women.
• It highlights the sources of tribal alienation and a deep sense of exclusion, since a majority of the Scheduled Tribes live in conditions of serious deprivation and poverty despite several special and targeted policies.
• The conflict between the Adivasis and the Indian state is seen not only in the Maoist areas but in the North-Eastern states as well.
• Poverty and illiteracy, absence of self-governance, forest policy, excise policy, land-related issues, and political marginalization, are the causes of tribal discontent.
• Failure to implement protective regulations in Scheduled Areas, absence of credit mechanism leading to dependence on money lenders and consequent loss of land, and violence by the State are also rampant.
• The right to forest land is a central part of the tension between the Adivasis and the State.
• Large-scale displacement of tribals due to large development projects.
• The expert group points out that unless tribals are resettled in scheduled areas, they will lose their special rights and are likely to be further marginalized.
• The Indian state has access to a perspective highlighting the multiple causes of discontent and suggesting that the Maoist ‘menace’ is a political problem, not only a law-and-order one.
• The mainstream approach continues to be a security-based one.
• The danger with this approach is that the discontent can actually multiply manifold as, in addition to the already existing dissatisfaction against material deprivation and the resulting alienation, the local populations will now have to deal with the consequences of repression and violence by the state too.
• Historical episodes during the Naxalite movement that have seen political solutions.
• At the time of the first Naxalbari uprising, the United Front government was in power in West Bengal.
• Acquisition of ceiling-surplus land of zamindars and big landlords by the government and its redistribution among the landless peasants.
• This was an important part of the anti-Naxalite strategy that proved effective in weaning a section of the peasantry away from the Naxalites, accompanied by conventional law-and-order measures.

Gender

• Gender disparities in educational and occupational attainment have been a persistent feature of Indian society.
• Considerable concern and attention are also centered on the adverse and declining sex ratio, the phenomenon of sex-selective foeticides and infanticides, and an overall neglect and malnutrition of the girl child.
• One specific aspect is gender wage gaps to comment on the nature of economic discrimination faced by women workers.
• A standard way of estimating discrimination in labor markets is by decomposing the average wage gap into an ‘explained’ (by wage-earning characteristics) part and the ‘residual’, which is the measure of discrimination in the labor market.
• A newer technique of quantile regressions allows us to assess how returns to wage earning characteristics vary over different quantiles or over different parts of the wage distribution.
• Using quantile regressions, we can see if wage gaps are higher at the top or bottom ends of the wage distribution or do women face a ‘glass ceiling’ or a ‘sticky floor’.
• Estimating quantile regressions separately for men and women also allows for separate decompositions for each quantile.
• Do high earning women face greater discrimination (greater residual or a greater unexplained part) as compared to low wage earning women?
• Deshpande et al. (2018) use nationally representative data from the employment–unemployment schedule (EUS) of two large rounds of the NSS, for 1999–2000 and 2009–10, to explore gender wage gaps among the regular wage/salaried (RWS) workers.
• RWS workers segment of the workforce jobs presumed to be allocated on meritocratic lines.
• Underestimation is likely to be minimal among the RWS workers.
• Examine wage gaps at the mean as well as along the entire distribution.
• Decompose the gaps into an ‘explained component’ (due to gender differences in wage earning characteristics) and an ‘unexplained component’ (due to gender differences in the labor market returns to characteristics).
• Perform the standard mean decomposition as well as quantile decompositions.
• Evaluate changes in each of these over the ten-year time period.
• Finding:
  • Significant gender gaps among the RWS workers who constitute about $17$ per cent of the Indian labor force.
  • The raw (unconditional) gender wage gap at the mean changed from $55$ per cent to $49$ per cent between 1999–2000 and 2009–10, but this change is not statistically significant.
  • In both the years, even after accounting for differences in observable characteristics, the average female wages were less than those for males.
  • The Blinder-Oaxaca (BO) decompositions indicate that the bulk of the gender wage gap at the mean is unexplained, that is, possibly discriminatory.
  • The discriminatory component of the wage gap also increased.
  • In 2009–10, had women been paid like men, they would have earned more than men on account of their superior characteristics.
  • Male wages are still higher than female wages across the entire wage distribution.
  • The gender wage gaps are higher at lower deciles and decline thereafter.
  • In 2009–10, the gap is the highest at the first decile at $105$ per cent and declines to about $10$ per cent at the ninth decile, indicating the existence of the ‘sticky floor’, in that gender wage gaps are higher at lower ends of the distribution and steadily decline over the distribution.
  • This is true for all the RWS workers, as well as for rural and urban workers separately.
  • The sticky floor became ‘stickier’ for the RWS women over the decade.
  • The quantile decompositions also reveal that the bulk of the gender wage gaps are discriminatory and that the discriminatory component is higher at lower ends of the distribution.
  • High growth, new job openings, greater integration with the global economy, and an increasing domestic privatization in India.
  • Seguino finds that gender inequality, which lowers women’s wages relative to men’s, is actually a stimulus to growth in export-oriented economies.
  • Greater inequality is ‘less likely to produce social conflict if the burden is borne by women, a group traditionally socialized to accept gender inequality as a socially acceptable outcome’ (p. 1212).
  • High growth has not been accompanied by an increase in female labor force participation rates (LFPRs).
  • In 2009–10, only about 10 per cent of women in the labor force were in RWS jobs (as opposed to 16 per cent for men), and an overwhelming share of the RWS jobs were held by men (83 per cent).
  • Women face adverse returns to their characteristics.
  • In 2009–10, women were recorded to earn less than men throughout the wage distribution due to labor market discrimination, even though they had better characteristics than men.
  • At the lower end of the wage distribution, for the bottom $10$ per cent where women face higher discrimination, the wage gaps have increased.
  • 42 per cent of women in the regular wage/salaried category earn less than the statutory minimum wage.
  • Strict implementation of minimum wages.

Conclusion

• The Indian economy has undergone momentous changes over the last two decades.
• The translation and impact of these changes on inter-group disparities has been uneven.
• Caste inequality shows very strong inter-state variations and some convergence, but no clear relationship between growth and convergence.
• Gender wage gaps are substantial and greater for the lower part of the wage distribution.
• Class inequality has increased sharply, particularly in certain parts of the country, leading to protracted armed insurgency.
• The state has to step in decisively to act on a strategy of redistribution in order to quell discontent and to ensure that the gains from growth are distributed more equitably across the diversity of India’s population.
• The state needs to immediately improve the provision of public goods and make the policy framework as well as governance structures accountable, transparent, and inclusive.