AFFIRMATIVE action comes in a variety of forms, but for present purposes I define it as positive discrimination – i.e., the provision of some amount of preference in processes of selection to desirable positions in a society to members of groups that are under-represented in those positions. The preference may be provided in various ways – e.g., reserved seats in separate competitions or preferential boosts in a single competition – but it always has the effect of increasing the number of members of an under-represented group selected to a desirable position.

Positive discrimination policies are most often oriented to ‘identity groups’ defined by characteristics that are physical or cultural – such as race, caste, tribe, ethnicity, and gender – because such groups are most likely to perceive themselves and to be perceived as meaningful communities. In this essay I will focus on positive discrimination (PD) in favour of under-represented groups in the sphere of admissions to higher educational institutions, but most of the points I will make apply also to PD in the sphere of employment.¹

Equity in access to educational opportunities is surely an important goal, and positive discrimination in higher educational admissions is often presented as a mechanism to promote this goal. In fact, however, PD policies in higher educational admissions cannot be expected to contribute much to the achievement of the goal. Because they target students who have already successfully completed their (higher) secondary education, such policies can do little to reduce the most important educational inequalities – those that characterize access to decent pre-university education.

 

Diminishing these huge inequalities will require major investments in primary and secondary education, and very likely in pre-school education as well, in order to put the poor on a less unequal footing with respect to access to a good education. Such investments are needed to help offset the all-important economic disadvantages suffered not only by most members of under-represented identity groups but also by many members of other groups. What is required is a well-funded and explicitly class-oriented policy, one designed to reduce class differences and to promote widespread social mobility. In India (as in most other countries), this would amount to revolutionary social change.²

Just as they cannot be expected to overcome unequal access to educational opportunities, PD policies in higher education cannot be expected to serve as a means of effecting reparations for past injustices inflicted by mainstream society on marginalized communities – because those most deserving of reparations are most unlikely to be among the beneficiaries. By the same token, such PD policies cannot be expected to serve as a significant means of aiding the poor in general, much less as a frontal assault on major socioeconomic inequalities.

What PD in admissions to higher educational institutions can be expected to do is to serve the goal of building a better society by integrating members of disadvantaged and marginalized communities into the upper echelons of the society, viz. into respected occupations and responsible decision-making positions where they have been significantly under-represented. Such integration serves the general interest because better representation for marginalized communities provides the political system with greater legitimacy, strengthens democratic institutions through wider participation, enables better performance of jobs involving familiarity with and understanding of disadvantaged communities, and motivates youths from such communities to work harder to improve their future prospects. In contributing to these objectives, PD policies are clearly superior to class-based programmes.

PD policies in higher educational admissions are thus best suited to reduce identity-based differentials in access to the upper reaches of the class structure. This is by no means a trivial accomplishment, but it amounts to major reform rather than revolutionary change.

 

A key question arising in any consideration of PD policies is clearly: which groups should be eligible for PD? This question is both crucial and difficult, and has given rise to controversies and problems wherever PD has been applied. In light of what PD policies can realistically be expected to accomplish, I believe that PD should be applied in favour of marginalized identity-based communities whose integration into the upper echelons will manifestly make a major contribution to the welfare of the society as a whole, and to whom the according of preferences in selection will not fairly be viewed as arbitrary favouritism. This suggests the following criteria:

a) A PD-eligible community must be very under-represented in society’s most desirable positions so that greater integration of its members into the societal elite will make a big difference. The degree of under-representation of any given community is bound to be highly correlated with the extent to which the average socioeconomic status of community members is below the societal average. But relatively low average socioeconomic status alone does not justify positive discrimination on the basis of a given community.

b) A PD-eligible community must have a history of major mistreatment at the hands of mainstream society, one that has been continuing right up to the present time (though it may be diminishing in its intensity). The mistreatment should be clearly linked to negative attitudes on the part of large numbers of people toward members of the community simply on account of their identity. A community that has been victimized by major identity-based mistreatment at the hands of mainstream society is more likely than a group defined by its low socioeconomic status to be recognized as deserving of some degree of preference in efforts to gain success in that society.

c) Members of a PD-eligible community must have a strong sense of common identity. To the extent that this is lacking, because the community is too heterogeneous to share much of a common identity, the benefits of integration are weakened. Some of these benefits have a ‘trickle-down’ character: for greater representation of a community among an elite to bring benefits to the community as a whole, or to large numbers of its members, it is essential that these members identify with those of their community who succeed in joining the elite (and vice versa). Otherwise successful PD beneficiaries will not really enhance the ability of community members to participate in the economic, social and political life of the society, nor will they be able to serve effectively as role models for less fortunate members of their group.

 

In order to achieve its integration objective, identity-based PD needs to be implemented in a way that promises good performance by its beneficiaries. A student who fails to complete the degree programme in which she is enrolled may derive some benefits from the experience; but unless she graduates, she is unlikely to be able to move into a high-status position. Moreover, such failures are costly for the institution, for the society, and for the legitimacy of the whole project of affirmative action.

PD policies in higher educational institutions result by definition in the admission of applicants who have weaker conventional qualifications than is required for general admission. Moreover, PD beneficiaries from marginalized communities are disproportionately likely to face financial obstacles to persisting with their studies; and they are also disproportionately likely to encounter a hostile environment in educational institutions dominated by members of more privileged communities. It is, therefore, a considerable challenge to implement PD policies in such a way as to assure that a significant number – and proportion – of the beneficiaries are successful. These considerations suggest that a good PD policy will be one that has the following characteristics:

a) The PD beneficiaries are relatively well-prepared to face the challenges of succeeding in a competitive educational environment. This implies that the qualifications of PD beneficiaries – as conventionally measured on the basis of past performance – should not be too far below those expected of non-beneficiaries. Since such qualifications tend to be correlated with family socioeconomic background, this also implies that one should not exclude relatively well-off members of marginalized communities from eligibility for PD preferences.

 

For a PD policy to achieve its objective, it is therefore necessary to set a limit on the size of the ‘preference magnitude’ – i.e., the difference in qualifications between the last admitted general applicant and the last admitted PD-eligible applicant. This can be done by limiting the number of reserved seats in a system of separate competitions; but it is most straightforward to do it by limiting the size of the preferential boost accorded to PD-eligible applicants in a single competition.

b) The selection process itself is well designed to identify beneficiaries with hidden potential. One would like as far as possible to select applicants from PD-eligible groups who are formally under-qualified, but who nonetheless have the potential to succeed in a higher educational institution, while avoiding the selection of those who would be incapable of overcoming their weaker formal qualifications.

PD policies can vary greatly with respect to the sensitivity of the process whereby applicants are preferentially selected. The most rigid, mechanical type of PD admissions process involves a quantitative procedure in which all applicants take some kind of standardized test and are ranked simply by scores on that test. A more sensitive admissions process would take account of a variety of criteria in ranking applicants, and involve considered judgment by admissions personnel rather than mechanically determined scores fed into a composite index.

 

A highly sensitive PD admissions process would not only include qualitative evaluation of the extent to which a beneficiary group applicant satisfies various relevant criteria, but also treat disadvantaged group status as a signal to look especially hard for evidence of additional applicant characteristics suggesting a strong potential for good performance. Sensitive selection processes are, to be sure, harder and more costly to administer than rigid ones, since they require that more information of various kinds be gathered from applicants and that more people be employed to implement the selection process.

c) PD beneficiaries – especially the more disadvantaged ones – are provided with substantial academic, social and financial support. It is not only a matter of providing additional academic opportunities, such as preparatory courses, workshops and tutoring, to help PD beneficiaries catch up with their better-prepared peers. Also needed are programmes and activities that help PD beneficiaries adjust to their new settings and function effectively in them. Furthermore, it is essential that an educational institution implementing PD in admissions take positive steps to assure that PD beneficiaries are not disparaged by their peers and that they are well mentored rather than patronized, much less mistreated, by their instructors. Just as it requires an investment of financial resources to improve the sensitivity of admissions processes, similarly it takes resources to provide developmental support for under-prepared beneficiaries.

 

Direct financial aid to needy beneficiaries is another kind of financial resource commitment that can make a significant contribution to the success of a PD policy. Many student beneficiaries may not be able to afford all of the expenses associated with attending an educational institution, so they may require financial aid just to enrol. Students from socioeconomically disadvantaged families are much more likely to have to drop out of an educational programme – temporarily or permanently – in order to help out their family by contributing their labour to work within the family or by earning additional income. Means-tested financial aid to PD beneficiaries (and indeed to all socioeconomically disadvantaged students) can clearly help to prevent such problems from compromising the success of PD admissions policies.

d) Since the primary purpose of PD in higher education is to integrate the societal elite, it is most desirable that PD policies be implemented in the most elite higher educational institutions. For one thing, PD is much less likely to be needed to assure adequate representation of marginalized communities in mediocre institutions. More importantly, it is precisely the elite institutions that provide access to society’s prestigious occupations and decision-making positions. Additionally, it is the elite institutions that are most likely to be able to possess the resources needed to invest in facilities and programmes that increase the prospects for students from marginalized communities to achieve success.

The oft-suggested alternative of pursuing a policy of class-based – rather than identity-based – PD suffers from a variety of weaknesses. First, it would generate fewer beneficiaries from marginalized communities, thus weakening its contribution to the integration objective. Second, it would substitute less-well-prepared for better-prepared beneficiaries, thereby reducing the proportion of beneficiaries likely to be successful – and/or requiring the investment of more resources in an effort to maintain a high success rate. This latter point applies equally to efforts to remove the ‘creamy layer’ of marginalized communities from PD eligibility or to assure that PD benefits go to disadvantaged groups within PD-eligible communities.

 

Indeed, any effort to shift the composition of PD beneficiaries in the direction of a higher proportion of less well-off members of under-represented groups, and a lower proportion of better-off members of such groups, will be detrimental to the realistic objectives of a PD policy. Either it will result in fewer successful outcomes, or require the investment of considerably more resources in selection processes, performance support, and general financial support – so that members of under-represented groups have a better chance of entering and persisting in the challenging academic environments with which they will be confronted. It follows that proposals to combine class and community as criteria for determining PD eligibility, such as the ingenious one advocated by Deshpande and Yadav,³ miss the point: PD policies can best address the under-representation in high-status positions of members of marginalized communities, not the general class inequalities of a society.

 

I believe that the extension of reservations from SC and ST categories to a large number of OBCs – first at the state level and more recently at the national level – has on the whole been counter-productive for several reasons. First, OBCs have a significantly weaker case for inclusion in PD eligibility as marginalized communities because they are much more heterogeneous, are generally not as under-represented in the societal elite, and do not have a history of mistreatment at the hands of mainstream society as starkly oppressive and debilitating as in the case of Dalits and Adivasis.

Second, including OBCs along with SCs and STs as beneficiaries of PD has rendered – in most cases – virtually one half or more of the population eligible for PD preferences. Even when the actual proportion of PD beneficiaries falls somewhat short of such high levels, it exacerbates concerns about the displacement of members of non-eligible groups by members of PD-eligible groups, and dilutes the resources that can be used to provide much-needed support for PD beneficiaries. The more groups and people who benefit from PD, the less any given group or person can benefit.

If there is a case for extending eligibility for PD benefits beyond Dalits and Adivasis, it is on behalf of groups with a much stronger case than OBCs (as currently defined). Any other group aspiring to reservations should be able to demonstrate a degree of homogeneity, under-representation and marginalization at the hands of mainstream Indian society broadly comparable to that of Dalits and Adivasis. India’s Muslims may have the strongest case to join these two groups as beneficiaries of affirmative action. But the addition of Muslims (or any other marginalized community) to the ranks of PD-eligible groups would be hard to justify unless it is accompanied by a reduction in the current overall proportion of PD beneficiaries.

 

In questioning the need for PD in favour of OBCs, I do not wish to suggest that OBC members deserve no assistance. Most of them are indeed among India’s less privileged citizens, and there is a strong case for governmental action seeking to reduce the huge socioeconomic disadvantages they face. My point is that positive discrimination is only one of many possible policy tools that could be enlisted in the struggle to bring about greater equality of opportunity in unequal societies. PD policies should, therefore, be applied only when and where they are most likely to generate significant net benefits. To deal with most of the inequalities that still plague Indian society, policy-makers should look to different kinds of social and economic policies that can be deployed alongside positive discrimination.

With regard to the structuring of PD policies, quotas of reserved seats in proportion to shares of population make sense only in the sphere of political representation – not in the spheres of employment and education.⁴ In the latter spheres, the disadvantages in prior education and training characteristic of members of PD-eligible groups make it difficult to fill quotas for desirable positions without according such a high a magnitude of preference that many beneficiaries are unlikely to be successful. Quotas accompanied by reasonably demanding minimum qualifications are better than unconstrained quotas; but it may still be difficult to resist pressures to fill the full quota by lowering the minimum qualifications.

It would, therefore, be advisable not to define PD admissions policies in terms of quotas proportional to population, but instead to structure such policies around preferential boosts involving more modest preference magnitudes than those required to fill population-proportionate quotas. This will of course result in a smaller number of PD beneficiaries, but will also result in a significantly higher rate of successful performance by beneficiaries. Choosing the optimal magnitude of preference is, to be sure, a difficult task, but to focus on doing so will lead to greater net benefits than a focus on fulfilling a population-proportionate quota.

 

In implementing PD policies in India it would also be desirable to devote more resources to careful selection and to (several kinds of) support for the selected beneficiaries. The process of selection of PD beneficiaries should best be sensitive enough to be capable – at least to some extent – of identifying those formally under-qualified applicants who are most likely to succeed if given the opportunity. And PD beneficiaries should be afforded stronger human and financial support after being preferentially selected. The greater the extent to which a PD policy is structured to meet the above desiderata, the more successful it will be in integrating members of India’s disadvantaged and under-represented ethnic communities into the societal elite and thereby enhancing the well-being of the nation.

 

Footnotes:

1. I have drawn here extensively on my book, Affirmative Action in the United States and India: A Comparative Perspective (Routledge: London, 2004; printed and distributed in South Asia by Foundation Books).

2. In the case of India, Hasan and Mehta, ‘Under-representation of Disadvantaged Classes in College: What do the Data Tell Us?’, Economic and Political Weekly 41(35), 2 September 2006, pp. 3791-3796 have shown that economic differentials are more important than caste differentials in explaining unequal access to higher education in urban areas, while both are significant in rural areas. As they argue, the first-best solution to unequal representation in higher education of the most disadvantaged identity groups – Dalits, Adivasis and Muslims – is improved access to (and quality of) primary and secondary education for these groups.

3. See S. Deshpande and Y. Yadav, ‘Redesigning Affirmative Action’, Economic and Political Weekly 41(24), 17 June 2006, pp. 2419-24.

4. See Andre Beteille’s contributions to Seminar issues 268 (1981) and 375 (1990) for a clear and cogent articulation of this point.