ISSN (Print) - 0012-9976 | ISSN (Online) - 2349-8846

A+| A| A-

The Eternal Debate

Caste still remains an indicator of disadvantage as distribution of both income and wealth are skewed along caste lines. Though the data on OBCs is scanty, there exists a clear disparity between these castes and others in terms of educational attainment, occupational success and standard of living. The mechanisms for perpetuating inter-caste inequality are still strong and alive in contemporary India. Quotas, however, should not be seen as the beginning and the end of affirmative action.

The Eternal Debate

Caste still remains an indicator of disadvantage as distribution of both income and wealth are skewed along caste lines. Though the data on OBCs is scanty, there exists a clear disparity between these castes and others in terms of educational attainment, occupational success and standard of living. The mechanisms for perpetuating inter-caste inequality are still strong and alive in contemporary India. Quotas, however, should not be seen as the beginning and

the end of affirmative action.


he debate over reservations – the Indian version of affirmative action – is in the foreground once again. Since this time the debate, at least ostensibly, is about the other backward classes (OBCs), one of the first stumbling blocks in any informed intervention is the paucity of data. Most of what has appeared in newspapers and television commentaries are a set of pre-conceived notions intertwined with a virulent opposition to affirmative action. Prima facie, this could be attributed to the lack of data: reservations for OBCs could possibly be the first instance of affirmative action for a category that even the national census does not enumerate. The National Commission for Backward Classes only lists the jatis that comprise OBCs without any demographic data about these jatis. Also, there are no firm estimates of the distribution of OBCs in jobs and institutes of higher education.

It is entirely true that the government, whether in 1990 or now, ought to have preceded this major move with building a comprehensive case for this particular component of affirmative action, with adequate backing of data that could be released to the public. Also, while this move is supposed to be coupled with an increase in the non-reserved seats, very little thought seems to have been given to the logistics and practicality of increasing seats. Most departments and colleges already face huge constraints in the form of inadequate infrastructure and high student-teacher ratios and have very little scope for increasing student intake, without the concomitant structural changes that would be needed to sustain a higher intake. If the government is indeed serious about alleviating the condition of OBCs via increasing their share in public sector jobs and institutions of higher education, then this lack of preparation not only worsens its case but actually casts a doubt over the degree of its commitment to the reduction of caste-based barriers.

Having said that, it would be a mistake to believe that the opposition to this move is only or mainly because of lack of data and/or preparedness. The domination of an upper-caste mindset in the media, the academy and the wider public, is evident in most of the anti-reservation commentaries. Concerns about merit declining because “they”, rather than “us”, will swamp educational institutions; the completely mythical nightmare of “we” or “our children” being reduced to demeaning jobs like shining shoes and sweeping, whereas it is only fair that “they” continue to do so, etc, are only some of the indicators of a latent casteism that constitutes the reality of the supposedly caste-neutral contemporary urban India. Thus, the status quo, in which the share of the upper castes in both good jobs (in the public and the private sectors) and seats in higher educational institutions is far in excess of its share of population, is seen as essentially fair. While state-imposed reservations are seen as a murder of merit, the much more widespread “natural reservations” that come from birth in business (upper caste) families which ensure that no outsiders are ever considered for the top jobs in family-owned businesses; other manifestations of casteclass privilege being expressed via donations, capitation fees, etc, are never seen as threatening merit. Thus, the operation of caste barriers and a belief in caste hierarchies as natural underlies the dominant thinking so completely that, ironically, these beliefs are not seen as casteist, but their questioning is. Indeed, this mindset asserts that caste was virtually extinct before it was resurrected by V P Singh or Mandal or Arjun Singh. The possibility that the invisibility of caste from “our” lives is a privilege that “we” enjoy because “we” belong to a small minority of urban, upper-caste elite that has the luxury of leading a casteless existence, something that the majority is far from attaining, are issues that are virtually never discussed. However, distinct from this mainstream is the relatively small body of opinion that recognises the reality of caste-based barriers and disadvantages and has been engaged with

Economic and Political Weekly June 17, 2006


Chart: Caste Development Index, India, 1998-99Chart: Caste Development Index, India, 1998-99Chart: Caste Development Index, India, 1998-99Chart: Caste Development Index, India, 1998-99Chart: Caste Development Index, India, 1998-99








0.05 0

New DelhiHaryanaHimachal PradeshJammuPunjabRajasthanUttar PradeshMadhya PradeshBiharOrissaWest BengalArunachal PradeshAssamSikkimGoaGujaratMaharashtraAndhra PradeshKarnatakaKeralaTamil Nadu

OBCs as the new elite or as the new dominant castes that have presumably displaced the dominance of upper castes. As an economist, I am not qualified to comment on the nature of the political change; however, there seems to be no indication of a reversal in the relative economic position between upper castes and OBCs.

Second, these results are based on the OBCs as a whole without any distinction between the creamy layer and the more disadvantaged OBCs. It follows then, that if we exclude the “creamy layer”, then the disparities would be even sharper. The National Commission for Backward Classes ( has several criteria (economic, social and educational) to define the creamy layer among OBCs that ought to be excluded from the proposed reservations.

Third, there is a view that the preparation of the OBC list, despite very detailed

Selected States



the question of the combination of instruments that would be needed to break this.

Is Caste Past or Present?Is Caste Past or Present?Is Caste Past or Present?Is Caste Past or Present?Is Caste Past or Present?

One of the key issues in this debate is whether caste is an indicator of disadvantage at all, in the present. There is now a growing volume of literature that establishes empirically the reality that all indicators of advantage/disadvantage (monthly per capita expenditure, landholding, occupational attainment, educational attainment, ownership of consumer assets, type of housing, poverty ratios and so forth) are very clearly stratified by caste. Thus, while reasonably accurate income data are hard to come by, all these indicators suggest the distribution of both income and wealth are skewed along caste lines. This is not a trivial finding as it challenges the notion that disadvantage or advantage is randomly distributed between castes and, thus, establishes a case for focusing on groups, rather than individuals, as targets for policy-making. Of course, this does not deny intra-group inequality, but any group-based programme can be fine-tuned to take that into account, if necessary.

Since the current debate is about OBCs in particular, it is true that the census does not count OBCs as a separate category. However, there are some other data sources that give a fairly instructive glimpse into



the state of the OBCs. I am familiar with two large all-India data sets: the National Sample Survey (NSS) and the National Family Health Survey (NFHS) that have added the OBC category to their surveys after the late 1990s. Both of these indicate a clear disparity between OBCs and “others” (the non-SC-ST-OBC residual: an approximation for upper castes). According to the second round of the NFHS, caste disparities in occupational attainment are clear. Sixty-eight per cent of SCs and 75 per cent of STs are agricultural and manual labourers, with the proportions among OBCs being 61 per cent and the upper castes, 47 per cent. At the other end of the spectrum, among the professional, managerial and high-level administrative jobs, the proportion of OBCs is 8 per cent with that of upper castes over 13 per cent.

As a part of my research, using the NFHS data, I have created and calculated a “Caste Development Index” (CDI) that is based on five indicators: landholding, occupational attainment, educational attainment, consumer durable ownership and asset ownership. It is only the second round of the NFHS data set that allows for the OBC category and it turns out that both in rural and in urban India, in 1998-99, the CDI for the OBCs was less than that for “others” in virtually all states of India, as the chart shows:

These results are striking for several reasons. First, they challenge the notion of objective indicators of backwardness listed in the Mandal Commission report, has been influenced by lobbying such that fairly prosperous and dominant jatis have managed to get themselves labelled as “backward”. Again, I am not competent to comment on the validity of these claims. However, assuming such errors have occurred, these would bias the standard of living indicators for the OBCs upwards. It follows then that the gap between the truly disadvantaged OBCs and the upper castes would only be larger than what our estimates indicate.

Finally, my limited experience with field work in rural UP indicates that not only is the dividing line between purity and pollution critical, but that the distinction between “dwija” and “non-dwija” jatis matters too. Recently, at a small workshop in Delhi, Yogendra Yadav circulated tables based on his election survey data that provided broader empirical evidence in support of this.

Of course, this still begs the question – whether these disparities are a hangover from the past or whether they are being perpetuated in contemporary India. The literature on dalits indicates that mechanisms for perpetuating inter-caste inequality are strong and alive in contemporary India. Indeed, comparing changes in monthly per capita expenditure (MPCE) between SCs and “others” over the last two decades does not indicate a closing of the gap; there is evidence of wide regional variation, with disparities increasing in several states, particularly since the late

Economic and Political Weekly June 17, 2006 1980s. A similar analysis with OBCs would require more data points.

Discrimination inDiscrimination inDiscrimination inDiscrimination inDiscrimination in
Labour MarketsLabour MarketsLabour MarketsLabour MarketsLabour Markets

In addition to disparities that are groupbased, most labour markets in the world display evidence of discrimination, viz, that average wages of groups differ more than what a skill or merit differential would warrant. This discrimination is over and above what economists call “pre-market” discrimination that individuals are subjected to before they enter the job market. So, while pre-market discrimination would lead to differences in access and quality of education between, say dalits and uppercastes, job market discrimination would lead them towards different kinds of jobs (job discrimination) or different wages for similar jobs (wage discrimination), after controlling for merit.

Indeed, India is no exception to this nearly universal tendency for labour markets to discriminate on the basis of group identity. Thus, a common misconception is that the only reason dalits fare badly in labour markets is because they are less educated or are less meritorious. This argument assumes an absence of discrimination or caste prejudice: if dalits managed to acquire better education and were similarly qualified, then this argument assumes that their occupational profile would be similar to that of high castes. A corollary of this argument is that, therefore, all policy efforts should be directed towards equalising educational opportunity and educational attainment and this would eventually equalise labour market outcomes.

Without doubt educational opportunities and attainment need to be equalised with urgency, given huge disparities in quality of and access to basic education. However, studies of the formal sector have shown that there are differences in average wages of SCs and non-SCs that cannot be accounted for by education differences alone. In other words, controlling for skill or human capital characteristics (indeed, for all possible measurable characteristics that explain wages), a part of the wage gap between SCs and non-SCs can be attributed to labour market discrimination.

Admittedly, similar studies for OBCs have not been undertaken, partly because data availability has been a problem. Also, as a field of research, while this is now very common in the developed world, particularly North America, it is still in its infancy in India. It would be instructive to see what a similar exercise for OBCs would reveal.

Additionally, while the role of caste in the allocation of more traditional rural jobs is documented and accepted, it needs to be noted that the job search process in urban labour markets is marked with a complete lack of transparency: most jobs are never advertised and networks/connections of families and friends play a very important role in how jobs are allocated. This puts the whole question of “merit” in a different perspective: to quote from a study on the US job markets, often “who you know” is more important than “what you know”. One of the studies that I am involved with suggests that even for highly qualified university graduates, there are big differences between dalits and others in the job search process, eventual placements and salaries, time taken to find jobs, and the gap between ideal jobs and the jobs they eventually find.

Growth or Affirmative Action?Growth or Affirmative Action?Growth or Affirmative Action?Growth or Affirmative Action?Growth or Affirmative Action?
For Whom?For Whom?For Whom?For Whom?For Whom?

My calculations with the NFHS and the NSS indicate that it is not possible to draw a neat relationship between the per capita State Domestic Product (SDP), the rate of growth of SDP on the one hand, and the absolute level of development of, say SCs, and the SC-others disparity, on the other. In other words, one sees all kinds of patterns: rich states with low disparity, rich states with high disparity, rich states with higher development of the SCs and rich states with relatively lower development of the SCs (and similarly for poor states). There is a similar confounding of patterns when we look at rates of growth (faster growing states and slower growing states). This suggests that while rate of growth and levels of SDP are matters of concern, these do not provide an easy solution to tackle caste disparities that need to be targeted independently. Indeed, this confirms the international experience where both highly developed and poorly developed countries find themselves grappling with the problem of inter-group disparity. One is, thus, led to conclude that caste may not wither away with development but in fact, some form of affirmative action might be imperative to tackle caste-based disparity and discrimination.

Of course, affirmative action need not be synonymous with quotas and different sections warrant different kinds of affirmative action. Dealing with the complexity of issues inherent in the method and form of affirmative action would require a separate article; suffice it to say that one of the results of the Indian quota system has been social mobility, evident in the growth of a dalit middle class, conspicuously missing from the neighbouring Nepal that has a similar social structure but virtually no social mobility among the dalits. However, it is equally true that the programme is mechanically implemented, in that the announcement of quotas is seen as the beginning and the end of affirmative action and typically, no other form of support or remedial teaching is ever provided to help the “quota” students cope with the pressures of higher education, leading to large dropouts. In recruitment and promotions, a lack of will ensures that efforts are made to undermine quotas. Institutions can get away with the minimal or sometimes no regard to the quota system, since there is never any effort to monitor and assess the functioning of the quotas and no penalties for exploiting loopholes in the system. More importantly, since the debate is only centred around quotas, a variety of other powerful instruments, particularly relevant to improving the lives of the mass of rural dalits, are never discussed.

One is, therefore, wary of a similar mechanical approach to the problems of the OBCs. Also, the greater heterogeneity among the OBCs and the lack of a stigma associated with untouchability, suggest that not only should the “creamy layer” be excluded from a blanket provision of reservations but that alternative proposals being suggested by scholars, such as a combination of academic qualifications with points to redress various forms of disadvantage, must be considered. Also, we might go much further in bridging the gap in education if equal and immediate attention is given to the dropouts at the primary and middle school levels and to improving both the quality and access to education in rural areas. This need not be an “either-or” choice: both affirmative action and attention to the larger structural problems can complement each other in a virtuous cycle. This would necessarily require thinking outside the box. What the government urgently needs to do is to set concrete goals and monitor outcomes very strictly, rather than paying lip service to only one possible instrument with no regard to the outcomes.



Economic and Political Weekly June 17, 2006

To read the full text Login

Get instant access

New 3 Month Subscription
to Digital Archives at

₹826for India

$50for overseas users


(-) Hide

EPW looks forward to your comments. Please note that comments are moderated as per our comments policy. They may take some time to appear. A comment, if suitable, may be selected for publication in the Letters pages of EPW.

Back to Top