Chapter 14 Contemporary caste discrimination and affirmative action
In the spring of 2006, medical students at the elite All India Institute of Medical Sciences (AIIMS) in Delhi, dressed in their white coats, took up brooms and swept the streets. They were protesting against affirmative-action policies by suggesting that they would become untouchable “sweepers” if proposals to increase lower-caste student admissions were implemented. In 2007, the government-sponsored Thorat Committee report on the All India Institute of Medical Sciences documented continuing casteism, including denial of
promotions to low-caste faculty, segregation of low-caste students in distinct hostels (dorms), and blatant harassment of
low-caste students (Mitta 2007). In 2008, the Supreme Court of India upheld the expansion of affirmative action in higher education, noting the “duty of the State to promote
positive measures to remove barriers of inequality and enable diverse communities to enjoy the freedoms and share the benefits
guaranteed by the Constitution” (Ashoka Kumar Thakur v. Union of India and Others [2008], SCCL.COM 436).
Caste discrimination in India continues. Affirmative-action laws, known as “reservations” and based on caste and socioeconomic categories, are an important legal means to offset this discrimination. India’s affirmative-action policies are among the oldest and most far-reaching affirmative-action policies in the world, but the future
of these policies is the subject of ongoing societal and legal controversy and debate. This chapter will introduce the problem
of contemporary caste discrimination, the affirmative-action policies to combat it, and three contemporary developments: classification
controversies, international activism, and growing attention to other disadvantages, in addition to caste, in Indian affirmative
action.
Contemporary caste discrimination
Caste continues to be an organizing principle of social life in India, although the practices associated with “untouchability” are no longer legal. Caste is rooted in ancient religious laws and codified in texts written before, during, and after the colonial period.1 A four-part social hierarchy of var
as (classes) was canonized in the subcontinent beginning with 
g Veda 10.90 (c. 1200 BCE) and later elaborated in ritual manuals and the Dharma
stra literature. These varas were associated with different occupations, rules of behavior, and standards of ritual purity. Legal texts of the period
between 1550 and 1680 CE went into even more depth about the roles and duties of the lowest vara, the 
dras, than the frequently cited code of Manu, and the caste society that developed includes “axes of ritual status, occupation, endogamy, power.”2 “The Laws of Manu” became a contemporary “lightning rod” and has been burned by activists as a symbol of both caste and gender oppression (Olivelle 2004a: xvii–xviii, xxxv; Zelliot 1996).
as (classes) was canonized in the subcontinent beginning with g Veda 10.90 (c. 1200 BCE) and later elaborated in ritual manuals and the Dharmastra literature. These varas were associated with different occupations, rules of behavior, and standards of ritual purity. Legal texts of the period
between 1550 and 1680 CE went into even more depth about the roles and duties of the lowest vara, the dras, than the frequently cited code of Manu, and the caste society that developed includes “axes of ritual status, occupation, endogamy, power.”2 “The Laws of Manu” became a contemporary “lightning rod” and has been burned by activists as a symbol of both caste and gender oppression (Olivelle 2004a: xvii–xviii, xxxv; Zelliot 1996).
Such legal texts on caste preceded a colonial obsession with classification that has received much scholarly attention (Cohn 1987, 1996a; Metcalf 1995). The British appropriated the Portuguese term “casta” (referring to lineage or race), applying the word “caste” to South Asian social structures in the colonial era. The term
“caste” persists in contemporary legal and administrative jargon: The so-called “Untouchable” groups officially listed (or “scheduled”) as eligible for affirmative action are still known as the “Scheduled Castes.” Thus the word “caste” is still widely used by South Asians today. “Caste” is a rough translation of the
indigenous terms “jti” (referring to countless birth groups that vary depending on context and region) and “vara” (which refers to an idealized hierarchy of Brahmins, K
atriyas, Vaiyas, dras, and, below all of these, the avara – castes outside the vara system – sometimes referred to as “Untouchables”). “Dalit,” which means “oppressed” or “ground down,” is the label currently preferred by many in these lowest, or “Untouchable,” castes.
ti” (referring to countless birth groups that vary depending on context and region) and “vara” (which refers to an idealized hierarchy of Brahmins, Katriyas, Vaiyas, dras, and, below all of these, the avara – castes outside the vara system – sometimes referred to as “Untouchables”). “Dalit,” which means “oppressed” or “ground down,” is the label currently preferred by many in these lowest, or “Untouchable,” castes.
In practice, caste divisions are more ambiguous and regionally varied than any of the legal caste codifications (ancient, colonial, or contemporary) would suggest; nevertheless, caste continues to shape the life chances of many people in terms
of residence, social interaction, marriage, education, and occupation. Occupational and educational discrimination are the
forms most explicitly targeted via affirmative action.
Unequal opportunities in employment are due to both past discrimination and contemporary discrimination. The economist Narendra
Jadhav argues that past exclusion and discrimination have “impacted Dalit access to capital assets and employment opportunities. This has meant a greater incidence of poverty and deprivation among Dalits” (Jadhav 2007: 9). A recent study in the sphere of private-sector hiring, modeled on a US study of racial discrimination, vividly illustrates
the persistence of discrimination in India. A striking study of name-based racial discrimination in the United States (“Are
Emily and Greg more employable than Latisha and Jamal?”) was replicated in India (Bertrand and Mullainathan 2003). Because caste affiliations are sometimes identifiable by surnames, this type of study proved to be quite revealing within the Indian
context of discrimination. Researchers in India sent 4,808 applications from fictional, equally qualified male graduates to
548 private-sector employers in response to advertisements, using surnames associated with different communities: high-caste
Hindu names, Dalit (“Untouchable”) names, or Muslim names. A Dalit’s odds of progressing to the next stage in the hiring process were only two-thirds of that of a high-caste Hindu “applicant.”
A Muslim’s odds were one-third of that of a high-caste Hindu (Thorat and Attewell 2007). Such discrimination in the private sector has inspired increasing demands for affirmative action in this sector.
Dalit students continue to face discrimination in both school education and higher education. A recent human-rights report
described “discrimination, discouragement, exclusion, alienation, physical and psychological abuse, and even segregation,
from both their teachers and their fellow students” due to ongoing casteism in some Indian schools (Human Rights Watch and
Center for Human Rights and Global Justice 2007: 92). Discrimination at educational institutions persists even in urban areas and at the highest levels of higher education.
In 2006 Dalit medical students at India’s premier medical school, the AIIMS, filed a complaint about intimidation and discrimination,
accompanied by a memorandum from forty other students reporting similar incidents. One student wrote: “I have been subjected
to mental and physical torture from my very first day at this institute . . . I was abused on my caste and, in the last few days my room had been locked from outside because of which I was unable to attend classes” (Human
Rights Watch and Center for Human Rights and Global Justice 2007: 92, n. 429). A young Dalit activist described how caste discrimination persisted through postgraduate studies in Mumbai (Bombay), even when external faculty came to participate in the evaluation of students. One such faculty member routinely put
a hand across the backs of male students, a seemingly friendly gesture, to feel for the sacred thread worn by many Brahmin men (interview, October 2006).
Although many Dalit students have excelled in academic and professional spheres, caste discrimination in India continues to thwart many citizens, keeping them from achieving their full potential. Narendra Jadhav, former chief economist of the Reserve Bank of India and
vice chancellor of the University of Pune, is an inspiring example of a person who overcame caste, yet he himself recognizes
the continuing social and economic costs of caste: “When you meet a person in India today, even as you introduce yourself,
your caste will quickly be assessed from your family name. Consciously or subconsciously, Indians, whether in their own country
or abroad, still make judgments based on caste. Over the years, the caste system has taken on sophisticated dimensions; it
has become subtler, though no less pernicious” (Jadhav 2007: 3).
Caste-based affirmative action
The Constitution of India abolished untouchability and discrimination on the basis of caste (Article 15), but caste categories continue to be politically salient (Jaffrelot 2003, 2007) and legally recognized for purposes of affirmative action in the form of reservations, a system of quotas regulating access to government employment, university
admissions, and legislative seats (Galanter 1984). In fact, one must declare one’s caste and even prove it with an official caste certificate to receive the benefits of reservations.
Such affirmative action based on legally recognized caste categories offsets but does not solve the problem of past and current
discrimination. More fundamental changes, particularly the consistent enforcement of laws relating to education for all and
human rights, would augment low-caste advancement, but affirmative action remains an important legal means to increase opportunities
for lower castes.
Colonial-era reservations were most developed in the south, notably in the Madras Presidency and some princely states: “In 1902 the Kolhapur ruler adopted one of the earliest examples of an official
caste-based ‘reservations’ scheme, decreeing that 50 per cent of all administrative vacancies were to be reserved for those
of ‘non-Brahman’ birth. A similar measure had been enacted in Mysore state in 1895” (Bayly 1999: 242). Reserved legislative seats for non-Brahmins in Madras from 1919 resulted in a government that created a 48-percent quota for non-Brahmins in administrative posts (Jaffrelot 2005). Classifying the lowest of the groups within the larger non-Brahmin category, the British colonial government officially recognized the so-called “Untouchable” castes, previously known as “depressed classes,” by listing them in 1936 as the Scheduled Castes in order to implement the 1935 Government of India Act. This act gave special electoral representation to various minority groups.
Legal recognition and rights for the lowest castes were a key point of disagreement between nationalist leader Mohandas Gandhi and Dr. B. R. Ambedkar, a member of the depressed classes who returned to India from advanced study in Britain and the United States
to work for the rights of India’s lowest castes.3 Although both Dr. Ambedkar and Mohandas Gandhi fought for the rights of “Untouchables,” their divergent approaches in debates and negotiations over constitutional reforms in the last decades of
British rule reflected their different ideas about Hinduism, law, and the Indian nation. “Fundamentally, this debate was about the place that social difference would have in
the context of an emergent Indian identity” (Tejani 2007: 58).
Gandhi hoped to further lower-caste rights by reforming Hinduism and worried that distinct legal rights for lower castes or large-scale conversions from Hinduism by lower castes would splinter Indians during the anticolonial struggle. He even fasted to protest against Dr. Ambedkar’s proposals to give the “depressed classes” reserved seats and separate electorates. Reserved seats would mean
only candidates from the depressed classes could run for a certain percentage of seats. Separate electorates would mean only
voters from the depressed classes could vote in elections for those particular seats. Although Dr. Ambedkar gave up his proposal for separate electorates to end Gandhi’s fast, the depressed classes gained reserved legislative
seats as “Scheduled Castes” prior to Independence. Dr. Ambedkar insisted that such group rights for the depressed classes were necessary and criticized both Hinduism and Gandhi’s
reformist approach. Dr. Ambedkar eventually led a mass conversion to Buddhism in 1956 (Jenkins 2008).
After Independence in 1947, the Scheduled Caste list was reenacted with the Scheduled Caste Order of 1950, prepared for the
purpose of compensatory discrimination policies and other programs and protections for these groups. Dr. B. R. Ambedkar became independent India’s first law minister and the principal architect of India’s 1950 Constitution, which,
in amended form, is still in use today. India’s Constitution prohibits discrimination on the grounds of religion, race, caste, sex, or place of birth (Article 15). An early amendment in 1951 stated that this article should not prevent the government
from “making any special provision for the advancement of any socially and educationally backward classes of citizens or for the Scheduled Castes or Scheduled Tribes” (Article 15, Clause 4).
The Scheduled Tribes are listed by the government based on cultural and geographic distinctiveness and disadvantaged socioeconomic
status. The Scheduled Tribes, also known as Adivasis, are “those groups distinguished by ‘tribal characteristics’ and by their
spatial and cultural isolation from the bulk of the population” (Galanter 1984: 147). Like the Scheduled Castes, groups to be included in this category were initially listed as a protected minority in the 1935 Government of India Act and later recognized in the Indian Constitution for policy purposes, including affirmative action. The Constitution leaves the contents of the schedules of tribes and castes up to the president, in consultation with the governors; the Parliament can by law include or exclude groups from the
list (Constitution of India, Articles 341 and 342). Based on the constitutional commitment to these groups, Indian federal
law and many state laws have long included reservations of government jobs, public university admissions, and legislative
seats for Scheduled Castes and Tribes. Unlike members of the Scheduled Castes, who must be either Hindu, Sikh, or Buddhist
to maintain eligibility for affirmative action, Scheduled Tribe members may be of any religion and keep their ST designation.
Yet another affirmative-action beneficiary group, with a curious label drawn from the 1951 constitutional amendment cited above (Article
15, Clause 4), is the category officially known as the Socially and Educationally Backward Classes (SEBCs) or, more commonly, the “Other Backward Classes” (OBCs). The OBCs include low, but not “Untouchable,” castes and various similarly disadvantaged castes or groups within non-Hindu religions, including Islam and Christianity. Some groups in this category benefited from reservations as non-Brahmins in parts of India before Independence. After Independence Backward Classes Commissions chaired by K. Kalelkar (Backward Classes Commission 1955) and B. P. Mandal (Backward Classes Commission 1980) developed social and economic criteria for the Other Backward Classes. The Mandal Commission report, including a list of OBC communities, eventually became the basis for national-level reservations
for OBCs in government jobs, instituted in the early 1990s (Galanter 2002). The Indian Supreme Court upheld these politically controversial reservations, with some restrictions, in the landmark case Sawhney v. Union of India. In this case, the court capped the overall percentage of reserved seats and insisted that socioeconomically advanced individuals
from OBC communities (dubbed the “creamy layer”) be skimmed from the pool of people eligible for reservations (Indra Sawhney v. Union of India, 1992 SCALE 1: 68) The Other Backward Classes are now eligible for new reservations in higher education.
The affirmative-action system in India is a quota or reservation system, although in practice these quotas are not always filled,
especially in the higher ranks of the civil service or in the most competitive degree programs. Thus the difference, at a
practical level, between quotas and the looser “goals” used in other countries, such as the United States, may not be as great
as the terminology implies; a rigorously enforced system of goals may achieve more advancement for disadvantaged communities
than a less enthusiastically enforced quota system (interview with Delhi University sociology professor Satishe Deshpande, January 7, 2008). Nevertheless, Indian quotas do not face the legal
hurdles they would elsewhere, as long as the overall quotas remain below 50% (although even this limit is currently being
ignored in some Indian states). In central-government jobs, for example, the Scheduled Caste quota is 15%, and the Scheduled
Tribe quota is 7.5%, based roughly on their percentages in the population. The central-government job reservations for OBCs
are limited to 27%, so as not to exceed the 50% limit (Indra Sawhney v. Union of India 1992: 68).
In addition to being eligible for affirmative action, low-caste victims of discrimination, violence, and other crimes can bring claims under the Protection of Civil Rights Act of 1955, enforcing Article 17 of the Constitution, which abolished untouchability, or the Scheduled Castes and Scheduled Tribes (Prevention of Atrocities) Act of 1989. These far-reaching but unevenly implemented provisions have provided some recourse
for individuals confronting significant continuing problems of caste-based discrimination, ranging from socioeconomic barriers
to violent attacks (Narula 1999a: 39–41).
With economic liberalization, opportunities in the once dominant public sector are being overshadowed by the growing private
sector, so demands for affirmative action in private-sector jobs are escalating. Private-sector affirmative action is, so far, very limited and the
product of voluntary efforts or conditions placed on public/private initiatives, as in the state of Uttar Pradesh under Chief
Minister Mayawati, herself of Dalit origin. Indian industries have proposed their own affirmative-action plans to dissuade the government from enacting
mandatory reservations in the private sector. For example, the Confederation of Indian Industries has published a report detailing
their plans for “voluntary and self-regulated” affirmative action for the lowest castes and tribes (Confederation of Indian Industry 2007). Even if private-sector affirmative action became the law, the large informal economy in India would complicate implementation
because “a large and increasing percentage of workers are in the ‘unorganized sector’ in which reservations, let alone affirmative action, are not applicable because they are inoperable” (Wright 2007).
Recent controversies and debates over affirmative action
By simultaneously abolishing the worst forms of caste-based discrimination and recognizing official caste categories to achieve social justice, the government of India has sparked legal and societal battles over the future of affirmative-action policies. Recent developments related to caste-based affirmative action include (1) classification controversies,
(2) international activism on caste discrimination, and (3) growing attention to affirmative action rationales and policies
based on disadvantage rather than caste per se.
Classification controversies
Although sometimes assumed to be fixed, caste is complex and ever changing, leading to controversies over the classifications of castes and the appropriate uses of these classifications to achieve social justice. The bench of Supreme Court justices staying the implementation of the Central Educational Institution (Reservation in Admission) Act of 2006 lamented that “Nowhere else in the world do castes, classes or communities queue up for the sake of gaining backward status . . . Nowhere else in the world is there competition
to assert backwardness and then to claim we are more backward than you” (“SC Stays OBC Quota” 2007). Controversies over which groups should be eligible for which reservations and which individuals should be included in those
groups demonstrate the complexity of using caste as a legal or policy category.
In May 2007, fourteen people died in clashes with police when thousands of protestors from the Gujjar community demanded that
their entire community be recognized as a Scheduled Tribe rather than an “Other Backward Class,” their official category,
which is eligible for fewer benefits than the Scheduled Tribes (Reuters 2007). Their claim was referred to the Justice Jasraj Chopra Committee, which drew on their tours of 147 villages, 35,000 affidavits,
and nearly 200 video discs of Gujjar habitations, which were submitted to local collectors (Rajalakshmi 2008a). The committee suggested that rather than reclassify the Gujjars using the “obsolete and outdated” criteria used to define
Scheduled Tribes, the government should invest in infrastructure, education, and health to benefit the disadvantaged populations
within the Gujjar community (“Justice Chopra Committee Rejects Gujjars’ Demand for ST Status” 2007). The committee recommended revamping the Scheduled Tribe criteria, but not waiting for this revision to be completed before
infusing remote Gujjar areas with some resources to aid their development.
To assess whether Gujjars are a Scheduled Tribe, the committee was supposed to use criteria that are throwbacks to an era
of oversimplified and stereotypical views of tribal identity. These include the following anachronistic (and difficult to measure) criteria: “primitive traits, distinctive culture,
geographical isolation, shyness of contact and backwardness” (Rajalakshmi 2008a). In addition to critiquing these criteria, the Justice Jasraj Chopra Committee questioned the fundamental premise of organizing
affirmative action on the basis of communities such as castes or tribes. They found it “difficult to take caste as the basis for favoured treatment. It is the geography of caste that defines the sociology of deprivation. Replacement
of caste by an area-based strategy seems to be the only way out” (quoted in Rajalakshmi 2008a). Proposals such as this, to replace group classifications with other indicators of disadvantage, are part of a broader affirmative-action
debate that will be explored in more depth below.
Ultimately the state government of Rajasthan responded to the committee’s recommendations by denying support for the Gujjars’
claim to be a Scheduled Tribe, but also by proposing a special, 5% Backward Class reservation category for Gujjars, in addition
to the existing reservations in the state for Scheduled Castes, Scheduled Tribes, and Other Backward Classes. At the same time, the state government proposed a 14% reservation for economically backward higher
castes in the state (Rajalakshmi 2008b). By bringing the total reservations in Rajasthan above 50%, and by basing additional reservations on purely economic criteria,
this scheme has faced legal challenges based on the Supreme Court’s decision in Indra Sawhney v. Union of India (1992).
Such disputes over the proper designation for groups such as the Gujjars are only one form of classification controversy.
Even if the group in question is clearly on the official list of groups eligible for affirmative action, and satisfied with their current category, individuals’ identities can remain ambiguous. Caste classification
becomes particularly unclear for people who have shifted identities through religious conversion, intercaste marriage, or migration (Jenkins 2003). The Indian Supreme Court has, for example, grappled with a case in which a child of converts to Christianity reconverted to Hinduism and applied for a reserved job for Scheduled Castes (S. Swvigaradoss v. Zonal Manager, F.C.I. (1996), RD-SC 137, 1996 AIR 1,182, 1,196 SCC (3) 100 JT, 1996 (2) 182, 1996 SCALE (2) 11; Jenkins 2003: 35–8). A Presidential Order in 1950 restricted Scheduled Caste reservations to Hindu Scheduled Castes but was later amended
to include Sikh Scheduled Castes in 1956 and Buddhist Scheduled Castes in 1990. Christians and Muslims have argued that they
too face caste discrimination in India, but the Supreme Court has postponed considering whether Christians or Muslims of Scheduled
Caste origins should benefit from reservations. There is no such religious restriction on Scheduled Tribe or Other Backward
Class status.
Intercaste marriage also complicates individuals’ identities, resulting in legal challenges when, for example, the upper-caste
wife of a lower-caste man applied for a job reserved for OBCs (Valsamma Paul v. Cochin University [1996], 3 SCC 545; Jenkins 2003: 31–4). Migration also blurs legal identities because communities on a list of “backward classes” in one state may not be on the list in another. A case in 2008 involved a migrant in a local election,
competing for a seat reserved for OBC candidates. She claimed to be in a “carpenter” caste that had two different names in
two bordering states, “Badhai” in Madhya Pradesh and “Sutar” in Maharashtra. The movement of people and even the movement of state boundaries during the reorganization of Indian states became
issues in this case (Sau Kusum v. State of Maharashtra and Others, 2008 INDLAW SC 1,994).
A final form of classification controversy results from the use of caste certificates to prove individual identities in order to apply for reserved opportunities. The Supreme Court has pondered cases about what could be characterized as identity fraud, as in their decision that some “social-status certificates” verifying Scheduled Tribe identity, furnished
to a medical school as part of the admissions process, were “false.” In this case the students in question had their Scheduled
Tribe status revoked and were demoted (in terms of priority for reserved admissions) to the less “backward” category of “Other Backward Classes” (Kumari Madhuri Patil v. Additional Commissioner of Tribal Development [1994], 6 SCC 241, discussed in more detail in Jenkins 2004). Concerns about fake certificates continue, as in a recent exposé of a certificate racket in Amritsar (S. P. Jha 2008).
A 2008 Supreme Court Case dealt with several of these definitional challenges associated with deciding who is in a reserved category,
in a case involving a bogus certificate, the blurred line between “caste” and “tribe,” geographical complications, and disagreements between different arms of the government (Union of India and Others v. S. Krishnan and Another, 2008 INDLAW SC 156). A government employee, hired under a Scheduled Tribe quota, was found to have been hired using a bogus
caste certificate. He then claimed to be in another caste (Lambadi caste), which he claimed was a Scheduled Tribe in the state of Tamil Nadu. A letter from a Director of District Welfare stated that the Lambadi caste was considered a Scheduled Tribe in
most districts in the state, evidence that was given greater weight in the state High Court, yet the Supreme Court found that
the official state Scheduled Tribe list did not include the Lambadi group. The Supreme Court denied his Scheduled Tribe status.
Categorization controversies involving groups claiming a more “backward” status, individuals crossing social lines via conversion,
marriage, or migration, and the process of certifying identities will continue to pose legal and administrative challenges.
International activism
Activists are increasingly recognizing caste as a global phenomenon and participating in international legal debates, as in efforts to unite oppressed populations
from countries ranging from Japan to Nepal to Senegal in a call to declare casteism a form of racism at the 2001 UN World Conference against Racism. Such appeals
to international law have not convinced the government of India to equate caste and race or to submit meaningful reports to
international entities such as the UN Committee on the Elimination of Racial Discrimination. On the other hand, caste is getting
more recognition internationally; activists are gaining publicity and energy from international collaborations; and nongovernmental
organizations have created excellent “shadow reports” documenting ongoing problems related to discrimination and the implementation
of affirmative action.
The Convention on the Elimination of All Forms of Racial Discrimination (CERD) refers to discrimination based on “descent,”
making this convention the focus of activists challenging caste at the international level. Activism sparked by the United Nations World Conference against Racism, Racial Discrimination,
Xenophobia and Related Intolerance, held in Durban in 2001, drew new attention to “work and descent-based discrimination”
such as caste discrimination. The draft documents for this conference included caste discrimination, but the participating
governments did not adopt the caste paragraph in the official Declaration and Programme of Action.
The simultaneous Non-Governmental Organization (NGO) Forum Declaration and Programme of Action, however, did recognize caste-based discrimination as a form of racism and devoted several paragraphs to it. NGO interaction and activism at the forum
publicized and internationalized the struggle against caste discrimination as practiced not only in South Asia, including India, Pakistan, Sri Lanka, Nepal, and Bangladesh, but also in Japan and in several African countries, such as Nigeria, Senegal, and Mauritania (Bob 2009).
In the following year, the United Nations Committee on the Elimination of Racial Discrimination “expanded the meaning of the
term ‘descent’ in Article 1 of the International Convention on the Elimination of All Forms of Racial Discrimination (ICERD),
to include discrimination based on caste” (Sengupta 2008). The Indian government had earlier ratified this convention but persistently argues that caste is not race. The Indian government
submitted a report in 2006 to the Committee on the Elimination of Racial Discrimination, arguing that the Indian Constitution
lists caste and race as distinct categories, in Article 15, which prohibits discrimination on the basis of either. The Committee
on the Elimination of Racial Discrimination, in its “observations,” challenged the Indian government’s arguments: “Discrimination
based on ‘descent’ includes discrimination against members of communities based on forms of social stratification such as
caste and analogous systems of inherited status which nullify or impair their equal enjoyment of human rights” (quoted in
Sengupta 2008).
The UN Committee on the Elimination of Racial Discrimination receives and offers such “observations” about the predictably
self-serving and political reports from national governments. This process lacks “teeth,” although it arguably publicizes
problems of discrimination. Another recent attempt by the UN to publicize “Untouchables” (as well as to laud private-sector clinics for low castes and to draw attention to the International Year of Sanitation) brought several Dalit women, whose previous employment was manually cleaning septic systems, to the UN in New York, where they walked down
a runway in UN-blue saris, accompanied by professional models in chiffon (Hughes 2008). This kind of publicity is not likely to solve problems of caste discrimination.
On the other hand, bringing the issue of caste discrimination to the attention of the UN has inspired many other organizations to contribute substantive, documented
“shadow reports” to the Committee on the Elimination of Racial Discrimination, reports that can be much more revealing than
government reports or special events in New York City. One such report from Human Rights Watch and the Centre for Human Rights
and Global Justice, Hidden Apartheid (2007) is a comprehensive overview documenting ongoing discrimination on the basis of caste. Such documentation is a major benefit
of the internationalization of caste struggles.
Moreover, despite the refusal of official Indian UN delegations to recognize casteism as racism, Prime Minister Manmohan Singh,
at a conference in Delhi in 2006, drew parallels between discrimination against Dalits and apartheid (Rahman 2006). In addition, Indian justices in affirmative-action decisions often take a more global view of the parallels between caste and race. The Supreme Court of India routinely refers to American case law on affirmative action. Their most recent decision on affirmative
action in higher education includes the following argument: “Of the classifications on which there is case law, the one that
most closely resembles caste is race. This is because both are immutable traits. They are used by the powerful, or those seeking
power, to justify oppression. Racism and casteism have long haunted both nations” (Ashoka Kumar Thakur v. Union of India and Others, writ petition [civil], paragraph 192, decided on April 10, 2008). This decision goes on to cite two US Supreme Court decisions
from 2003 on affirmative action at the University of Michigan (Grutter v. Bollinger [2003], 539 US 244 and Gratz v. Bollinger [2003], 539 US 306), as well as several older US cases, and even a (US-based) Cato Institute study entitled “The Affirmative
Action Myth” and a New York Times article entitled “The New Affirmative Action.” The Indian Supreme Court notes that US cases are “not binding” but “have great
persuasive value and they may provide broad guidelines as to how we should tackle our prevailing condition” (Ashoka Kumar Thakur v. Union of India and Others, paragraph 183). Likewise, caste-based affirmative action in India has gained some international attention as a potential
model for similar policies elsewhere (Ginsburg and Merritt 1999; Cunningham, Loury, and Skrentny 2002).
Caste versus disadvantage
In India, longstanding policies of caste-based quotas are increasingly shifting toward policies and proposals targeting beneficiaries based on more multifaceted
indicators of disadvantage. Blending caste and other criteria is not new in India but these other criteria are an increasingly
important aspect of reservation policies. The shift toward recognizing “multiple discrimination” in affirmative action may primarily be inspired by legal developments, but it follows longstanding feminist and sociological
theorizing on the “intersectionality” of identities and the importance of tracing the effects of multiple axes of discrimination
based on race, caste, class, gender, religion, and so on. At a practical level, however, a move toward policies based on disadvantage or diversity could
dilute affirmative action for specific caste and racial minorities that still face discrimination. Given ongoing discrimination
on the basis of categories such as caste and race, a shift away from policies that recognize race and caste as legitimate
and primary indicators of disadvantage could make affirmative action a much weaker tool to combat discrimination against low castes or racial minorities.
The Other Backward Classes category is based on a complicated blend of caste and other socioeconomic criteria. The Mandal Commission on Backward Classes developed criteria for OBC status based on social, educational and economic indicators
of “backwardness” (Backward Classes Commission 1980). As discussed above, Sawhney v. Union of India upheld central-government job reservations that were largely based on the Mandal Commission’s recommendations, but recalibrated the definition of beneficiaries. The Supreme Court insisted that governmental classifications of Other Backward Classes must take into account both caste and economic status, which necessitates attention to both group and individual criteria (Jenkins 2001, 2003). This means that purely economic quotas, for poor individuals within upper castes, are not allowed, and that socioeconomically advantaged individuals from an Other Backward Class, officially known
as the “creamy layer,” cannot benefit from OBC reservations, despite the “backwardness” of their caste as a whole.
What is the “creamy layer”? People are in the creamy layer when they or their parents have reached specified levels in high
government posts, the civil service, the armed forces, or public-sector undertakings, or have surpassed an income/wealth ceiling.
This ceiling was originally Rs. 100,000 rupees per year, but it was raised to Rs. 250,000 per year in 2004 and subsequently
in 2008 to Rs. 450,000 per year. People with “professional class” jobs, property owners, or those in trade, business, or industry can also be excluded based on property ownership and/or income or wealth (National Commission for Backward Classes 2008). For caste certificates to be provided by OBC candidates for reserved opportunities, local officials must not only certify that
the candidate is a member of an Other Backward Class but also that he or she is not in the creamy layer.
A subsequent policy development in Other Backward Class reservations renewed debate over the appropriate beneficiaries of
affirmative action and sparked protests by medical students, described above, and others: The Central Educational Institution (Reservation in Admission) Act of 2006 increased quotas to 49.5 percent of seats in national
public universities, including the renowned and fiercely competitive All India Institute of Medical Sciences and the Indian
Institutes of Technology (IITs). In response to protestors, the government agreed to increase the overall number of seats
in these institutions, so that the number of unreserved seats would remain constant. In the face of numerous legal challenges,
this expansion was approved by the Supreme Court in April 2008. The 49.5 percent quota in central universities remains controversial because it increases the quota and broadens the eligible groups
beyond the most disadvantaged castes and tribes to include the less disadvantaged and more numerous Other Backward Classes.
The expansion of OBC reservations to higher education precipitated the resignation, in 2006, of prominent scholars Andre Beteille
(sociology) and Pratap Bhanu Mehta (law and politics) from the government’s National Knowledge Commission, which was trying
to rethink and improve the Indian education system. In his letter of resignation, Mehta argued for more nuanced affirmative action: “What we needed, Honorable Prime Minister, was space to design more effective mechanisms of targeting
groups that need to be targeted for affirmative action.” More effective targeting of affirmative action combined with more
comprehensive reforms reaching everyone would be better than quotas, he argued: “As a society we focus on reservations largely
because it is a way of avoiding doing the things that really create access. Increasing the supply of good quality institutions
at all levels (not to be confused with numerical increases), more robust scholarship and support programmes will go much further
than numerically mandated quotas” (P. B. Mehta 2006). Mehta has also critiqued the Indian Supreme Court’s political “balancing acts” in its decisions on reservations and other issues (P. B. Mehta 2007: 112–13).
The Supreme Court decision on Other Backward Class reservations in higher education includes approving references to the United
States Supreme Court decisions in the University of Michigan cases, which have moved colleges and universities in the US to
retool admissions policies away from race (particularly racial point systems) and toward a variety of indicators of socioeconomic
status or disadvantages. In his judgment in the Indian case, Justice Dalveer Bhandari notes that “Justice Sandra Day O’Connor
opined that there may be a time-limit to promoting diversity via preferential treatment for certain races” (Ashoka Kumar Thakur v. Union of India and Others, paragraph 248). Although recognizing that the nine-judge holding in Sawhney necessitates that a combination of caste and economic criteria continue to define Other Backward Castes, Bhandari urges the government to move toward a more
economically based system of affirmative action: He laments that Sawhney “rejects purely economic criteria (occupation / income / property holdings / or similar measures of economic power) with respect to classification . . . It also precludes us from
forcing the government to wean itself off caste-based reservations by a certain date in order to achieve a casteless and classless
society,” but the judge advocates that “after a lapse of ten years, special preference or reservation should be granted only on the basis of economic criteria as long as grave disparity and inequality persist” (Ashoka Kumar Thakur v. Union of India and Others, paragraph 248).
As the court pushes for less focus on caste and more attention to economic indicators of disadvantage, the current government has proposed the creation of an Equal
Opportunity Commission. As political scientist Suhas Palshikar envisages, “the proposed commission of equal opportunity permits
the government to comprehend the complex networks of diversity, disadvantage and discrimination instead of remaining entangled
in the avoidable path of attending to only one community or focusing on only one axis of this complex phenomenon” (Palshikar
2008: 84). Several voluntary affirmative-action programs in Indian higher education are possible models for policies based on complex calculations of disadvantage.
Examples include the admissions policy at Jawaharlal Nehru University (a public, postgraduate institution) and the selection process for the Ford Foundation’s international
fellowship program in India.
At Jawaharlal Nehru University (JNU) in Delhi, the admissions system included “deprivation points” to give a boost to potential students who are disadvantaged but
not eligible for current government mandated reservations for Scheduled Castes or Tribes or the physically handicapped. This sort of voluntary affirmative action scheme has been in place at JNU on and off since the 1970s. Under the JNU policy, applicants from backward
districts got an additional three to five marks (“regional deprivation marks”). JNU’s own geography and regional-development
professors designated districts, giving them three to five marks depending on their degree of “backwardness.” The JNU geography
professor in charge of the “index of regional backwardness” described three aspects of the district-level index: rural parameters
using agricultural productivity data from the Ministry of Agriculture, educational parameters using census literacy data (male
and female), and workforce participation parameters (male) also from the census. They divided districts into four quartiles,
which determine the number of points applicants got, if any (interview with Professor Sachidanand Sinha, January 16, 2008).
JNU also voluntarily used the existing government OBC category. For OBC applicants, if the applicant was not from the “creamy
layer” (at JNU this simply meant parents had enough income to have to pay income tax), female OBCs got ten points and male
OBCs five, but there was a cap at ten total deprivation points, so a female OBC from a backward district could not get fifteen
points (interview with JNU Registrar/Director of Admissions Professor K. C. Upadhyaya, January 4, 2008). JNU became subject
to the new government policy to provide quotas to OBCs in central universities. Geography Professor Sinha argues that JNU’s existing
policy was reaching a large percentage of OBCs while targeting the most disadvantaged and creating less controversy due to
its nonreliance on caste alone (interview with Professor Sinha, January 16, 2008).
In a related approach using points based on various indicators of disadvantage, Delhi-based social scientists Yogendra Yadav (Centre for the Study of Developing Societies) and Satishe Deshpande (Sociology
Department, Delhi School of Economics) helped design the Ford Foundation’s international fellowship program selection procedures
and published their ideas as an alternative to the government’s proposed expansion of OBC reservations in higher education
(Deshpande and Yadav 2006). Professor Deshpande opposes purely income-based affirmative action in favor of their multifaceted approach that retains caste/community as a key target. “We need a variety of policies rather than a single policy . . . in higher education economic
reservations could be against the very goal of reservations . . . If income alone [is used], you could be excluding the only
segment of a group that can get into higher education at all” (interview with Professor Deshpande, January 7, 2008).
Professor Yadav argues that reservation policies “should target” and should “be seen to be targeting” the “most disadvantaged.”
The former is good policy, he notes, and the latter is good politics (interview with Professor Yadav, January 12, 2008). Complicating
the already complicated OBC category, Yadav would like to design policies that reflect this “varied group.” The “best scheme”
in his view “has exactly the map of all these gradations [of disadvantage] and reverses that.” The best (and most “non-fudgeable”)
of the many indicators of educational disadvantage in their scheme is the “school from which you did your matriculation.”
He states, “You simply do not go to a government school in a village if you can afford not to.” From a social-science perspective,
indicators such as mother and father’s literacy are very telling, he said, but from a practical-policy perspective, you cannot
prove parental illiteracy. “Crude but robust indicators” he has worked into his proposal still feature caste and community, but also gender and type of school (based on the type of town, region, type of school, medium of instruction, etc.) (interview with
Professor Yadav, January 12, 2008). One small-scale implementation of their vision, the selection process for the Ford Foundation
International Fellows from India, uses, in the words of the head of that program, a “rather detailed matrix” including literacy,
rural location, primary and secondary education levels in locality, number of family members and number of earners in family,
whether the applicant went to government school, and even health standards (all indicators included in the application form itself) (interview
with Vivek Mansukhani, Ford Foundation, India, January 17, 2008).
By allowing Other Backward Class reservations and embracing the concept of the “creamy layer,” the Supreme Court has reinforced the need for complex designations of disadvantage for affirmative action. Jawaharlal Nehru University, by implementing affirmative action based on both government categories and geographical and gender indices of their own design, offers one model for the future of affirmative action, although their innovative approach
may be homogenized by national-level OBC reservations in higher education. Complicating rather than abandoning caste categories, Professor Yadav and Professor Deshpande would include other cross-cutting indicators of social, economic,
and, especially, educational disadvantages, while retaining the centrality of caste. Because both Yadav and Deshpande were
on the government’s committee to design a new Equal Opportunity Commission, their ideas should continue to have an impact
on the future shape of affirmative action.
Directions for future research
Opportunities for future research abound for those interested in the study of contemporary connections between caste and the law. Research could fruitfully center on the most recent permutations of caste and tribal classifications, on
international legal strategies of low-caste groups, or on the potential globalization of legal rationales for continuing affirmative action based on disadvantage or diversity rather than group categories such as caste or race. More broadly, the
linkages between caste, religion, and the law shed light on the varied meanings and philosophies of secularism in different
legal and cultural contexts (Needham and Rajan 2007). The impact of religious and legal categories on the lives of women has inspired rich critiques and commentaries capturing
the intersections of religion, caste, gender, and the law in women’s lives (Ray and Basu 1999).4 The potential expansion of affirmative action in the private sector in India, and the challenges of affirmative action or
anti-discrimination policies in unorganized or informal economic sectors, are timely topics for further study.
Laws based on caste or religious identities highlight the complex, contingent nature of social and legal groups, as individual and group
identities change over time through conversion (Robinson and Clarke 2003; Viswanathan 1998) or through social mobility up the caste hierarchy (Srinivas 1962).5 Assigned and asserted identities interact, as legal definitions of who is a Scheduled Caste, a Scheduled Tribe, or an Other
Backward Class shape, but do not determine, the identities claimed by individuals or groups. Indeed some groups claim identity categories explicitly rejected by the government, such as the Dalit activists asserting that they face “racial” discrimination. Caste as a basis for both discrimination and affirmative action has taken on new and unexpected forms in contemporary India and continues to be a salient social, political,
and legal category.