The Hindu Business Line : Is there still evidence of caste discrimination in higher education?

While the debate on affirmative action in higher education remains highly emotive, it is often not based on the available empirical evidence. In this edition of Macroscan, C. P. Chandrasekhar and Jayati Ghosh examine the evidence of the latest large sample survey of the NSSO on the actual extent of inequality in educational attainment across major caste groups.

Discrimination and marginalisation in higher education is a more complex mosaic than is often presented.

The issue of affirmative action in higher education has direct implications for public policy as well as the administration and functioning of academic institutions and, of course, the fate of large numbers of students. Yet the highly charged discussion on this are often based on inadequate empirical investigation. The latest large sample survey of the NSSO relating to 2004-05 allows us to assess the actual extent of inequalities in educational attainment of different cast e categories in the population.

It should be noted that in all the subsequent discussion, the category “others” covers not only the “higher” caste Hindus, but also person from other religious categories. To the extent that those from other religions (such as Muslims) are also educationally deprived as a category, this would underestimate the actual differences. So the differential position of “caste Hindus” is likely to be even higher in relative terms.

Literacy scene

It is useful to begin with an assessment of the overall situation with respect to literacy. The continuing inability of the Indian state to ensure universal literacy and basic education must surely count among the most significant failures of the development project.

As is evident from Table 1, in addition to the overall failure in average terms, there are substantial differences across different categories of population, not only among rural and urban residents but also across social groups. While around one-third of India’s population is illiterate according to this survey, the literacy rate is clearly much higher in urban areas.

Gender gaps

Furthermore, there are very marked differences across gender and caste. The literacy rate was the highest among the category “others” (78 per cent), which includes both upper caste Hindus as well as those of other religions. This was followed by the OBCs with a gap of nearly 13 percentage points, then Scheduled Castes, and finally was the lowest among the STs (52 per cent).

But it is to be noted that gender gaps were very marked and typically even higher than differences across social categories, so that women among the socially deprived categories were the most highly discriminated even in terms of literacy. Thus, the gap between the group with the highest literacy rate — urban males of the “other” category — and the groups with the lowest literacy rate — rural females from Scheduled Tribes — was as high as 52.4 percentage points, or well more than double the lower rate!

The differences across caste groups — and therefore the persistence of discrimination in education at different levels — are confirmed by the distribution of population by educational categories as shown in Charts 1 and 2 for rural and urban areas separately. It is not just that illiteracy rates are high among the deprived social groups, but the spread of higher education is also extremely low.

Interestingly, the proportion of each rural social group that is literate and/or has received up to primary education is around the same — between 27 and 29 per cent. This may indicate the push for school enrolment in the age group 5-14 years through the Sarva Shiksha Abhiyan.

For those who have been excluded from this initiative, because they were already too old to be enrolled in school, the gaps across categories remain very large — in illiteracy as well as in the extent of further education. Also dropout rates from schooling are significantly higher for deprived social categories, and therefore educational distinctions become progressively more marked across different social groups.

Urban vs rural

In urban areas the distinction is even sharper between certain categories. However, here it should be noted that the ST group exhibits rather different characteristics relative to the rural areas. Illiteracy rates among urban STs are lower than for urban SCs, and the proportion that has been through higher secondary or graduate and above education is higher for urban STs than for urban SCs or OBCs.

Nevertheless, the gap between all three of these relatively deprived categories and “others” remains large, with more than one-fifth of the urban “other” population having been through graduate or higher education, compared to 9 per cent or less for all other categories.

Chart 2 shows that in rural areas, among the male population only 1.7 per cent of STs and 2.2 per cent of SCs have received graduate or higher education, compared to 3.4 per cent of OBCs and 6.5 per cent of “others”. Once again, gender gaps dominate over gaps across social categories, with females in the best off group of “others” faring worse than males among OBCs, and the worst off rural female category — STs — showing only 1.5 per cent having had graduate or higher education.

In urban areas, the gender gap is substantially reduced among the “others”, and urban females in this group show higher rates of higher education than all other categories even among the males. Gender gaps do remain in the other categories, but they are proportionately less significant than among the rural population.

Enrolment ratios

However, since these estimates are for the entire population, they may not provide accurate estimates of social differentiation in higher education among the relevant age-group. Therefore, Table 2 provides data on the enrolment ratios among the age-cohort 20-24 years in both urban and rural areas. It can reasonably be supposed that this relates to higher education of some variety, whether in colleges, universities, institutes of technical training, etc.

The data refer to the “Usual Status” of activity, which allows for both principal and subsidiary involvement (in other words, enrolment in part-time and distance learning courses is also included).

This confirms some of the conclusions from the earlier charts. In rural areas, gender gaps in enrolment in higher education remain very high and dominate over social category gaps, although the latter gaps are also very large. Among rural males, the basic gap is between “others” and the rest; there appears to be relatively little distance between the other three categories.

Furthermore, enrolment among STs appears to have fallen slightly (though this may be due to statistical error) and enrolment among SCs and OBCs to have increased slightly in the first five years of this decade.

Among rural females, STs show higher enrolment in higher education than either SCs (who remain the lowest) and OBCs. There has been an improvement in the latter two, but from very low bases. Among rural women, the gaps across social categories do not appear to be very large, further emphasising the point that gender gaps dominate in enrolment in higher education in rural areas, and that young women are the most discriminated against regardless of social group.

In urban areas, the picture is somewhat different. First it should be noted that enrolment ratios appear to be reasonably high when compared to other developing countries at similar levels of income. The situation across social categories is more complicated. The “other” category has actually shown a slight decrease in enrolment for both urban males and females.

For urban males, the gap between STs and “others” is not very large, and indeed has been reversed in the latest period consequent upon significant enrolment increases among STs, such that male STs in the age group 20-24 showing a higher rate of enrolment in education than their “other” counterparts! However, there has been no increase in enrolment ratios for SC urban males, and only a marginal increase for OBCs.

Among urban females, also there was a significant increase in enrolment ratios of ST women and a slight increase in enrolment ratios of OBC women. However, enrolment ratios of both “other” and SC women actually declined. While these were small decreases and could reflect statistical error, it is important in the case of SCs because enrolment rates were already so low, especially when compared to the other categories.

Employment pattern

What explains these patterns of enrolment in what must be some form of higher education? There are obviously both demand and supply conditions, as well as social and economic factors affecting the ability of different categories to access higher education. In terms of the desire for higher education, many factors play a role, but it is fairly obvious that perceptions of improved employment prospects are important.

Therefore it is worth considering what the survey data have to tell us about the extent to which the probability of employment changes with higher education. Table 3 provides data on work participation rates and unemployment rates for those with at least a graduate degree.

Table 3 shows that ST women graduates show the highest open unemployment rate in rural areas, while for urban areas it is highest for SC women graduates. Indeed, for ST women graduates in rural areas, the probability of being unemployed is higher than of finding work!

Given these high open unemployment rates for women in particular, even after receiving higher education, it may not be so surprising that there is less enthusiasm for enrolment among these categories. (Note that these data relate to open unemployment, and therefore exclude those who have opted out of the labour force for whatever reason.)

affirmative action

What insights do these data provide for strategies of affirmative action? First of all, it is evident that very large differences in educational attainment and access continue to exist and therefore must be addressed through public action. Such action must necessarily include reservations, but there have to be other strategies in addition, to ensure wider and more democratic access. These can include more public provision of higher educational institutions in backward areas and for deprived groups, more scholarships and other incentives for deprived categories, etc.

Second, the data provided here have shown that the social reality of discrimination and marginalisation in higher education is a more complex mosaic than is often presented. Such complexity needs to be noted and addressed when designing public policies. In particular, some major gaps that are evident from these data need to be addressed. Most significant among them are the rural-urban gap and the gender gap, which cut across social categories especially in rural areas.

Third, it is also evident that higher education generates very different prospects of employment across social categories, and therefore strategies of affirmative action also have to incorporate actions designed to affect the labour market.