Contours of caste disadvantage

Traditional hierarchies are too deeply entrenched to be reversed through one single measure; they need a concerted push, backed by strong will from different segments of society, including, but not confined to, politicians

The rise of Other Backward Classes (OBC) and Dalit-Adivasi leaders in the political sphere is celebrated as India’s “silent revolution.” At the national level, this phenomenon has been especially marked since the early 1990s, leading to comments about the “Mandalisation” of the Indian polity. The political ascendancy of individuals from traditionally marginalised groups has been viewed as a large enough flux that it is believed to have either reversed, or certainly flattened, the centuries-old traditional caste hierarchies. The contemporary caste system is often represented as one where horizontally placed entities compete for government favours and for a space among the elites, often using the trump card of marginalisation, despite being powerful, rich and dominant groups within their local contexts. The pictures vary depending on the commentator: it is either one of competition between equals; or one where the upper castes are now the new marginalised and the so-called lower castes, especially the OBCs, the new elite. A running theme underlying these analyses is one that minimises the actual extent of disadvantage and discrimination, and celebrates the end of the caste system, or at least its dehumanising, hierarchical and exploitative avatar.

Change in political sphere

Indeed, the change in the political sphere is rather striking, especially in the Hindi belt. Until the early 1970s, upper caste Members of Parliament represented more than 50 per cent of North Indian MPs, compared to 5 per cent for the intermediate castes and at most 10 per cent for OBCs. In the South (and also in Maharashtra), due to a combination of reasons, including a long history of intense social reform movements, upper castes lost their predominant position in the political sphere rather early on. Even though the implementation of the Mandal Commission report is seen as “the” turning point, in actual fact, the share of intermediate castes and OBCs started rising in 1977. By 1989, upper caste MPs from Uttar Pradesh fell below 40 per cent for the first time, and the OBC share was 21 per cent. The same trend was getting reflected in the State Assemblies: e.g. in Uttar Pradesh, the share of upper caste MLAs decreased from 58 per cent in 1962 to 37.7 per cent in 1998.

Has this shift in the social composition of political leaders led to a corresponding change in material conditions of broad caste groups? The “OBCs-are-the-new-elite” theory certainly believes that it has: the personal enrichment of individual political leaders, coupled with some rags-to-riches stories are offered as evidence of this change. However, like several debates, this one has generated more heat than light, because these individual stories, dramatic as they are, do not give any sense of the broader contours of change in the relative ranking of the three broad social groups — Dalits-Adivasis, OBCs and “Others” (everyone else). In the absence of jati-specific data, “Others” are often taken as a loose proxy for upper castes, but it should be noted that the actual disparity between upper and lower castes would be larger than what is revealed by these broad data categories. Also, the OBC category in these data sets is the legal one, i.e. all jatis classified as OBCs, including dominant castes. Thus, the gaps between the truly “backward” OBCs and upper castes would be larger than what are revealed by aggregate data.

Material disparities

Along with Rajesh Ramachandran, I have tried to map the changing contours of caste disadvantage in India since independence using large-scale national data. Ideally, the best kind of data to examine changes over time would be longitudinal or panel data — one that tracks a large set of individuals over time. However, in the absence of that, we use National Sample Survey data and divide the population into birth cohorts, such that the oldest cohort in our data comprises individuals born between 1926 and 1935 and the youngest cohort are those born during 1976-85.

We examine gaps in several indicators (landholding, urbanisation, years of education, proportions of those with different levels of education, various employment indicators, wages and so forth) across caste groups, between each cohort, and see if the gaps are increasing or decreasing across cohorts. Suppose that upper castes have a higher proportion (than OBCs) of those who hold elite, white-collar jobs. That indicates disparity along traditional lines. But if this gap is lower for younger cohorts than it was for older cohorts, it would mean that OBCs are catching up with upper castes in terms of access to this tier of occupation. If, hypothetically, we were to find that among the younger cohorts, a greater proportion of OBCs had access to white-collar jobs than upper castes, it would suggest that the older hierarchy of access to such jobs had been overturned, and the OBCs are indeed the new elite in this dimension. We examined each of our indicators using this methodology, and here’s what we found.

The overarching picture shows that there are clear disparities in virtually all indicators of material well-being, with upper castes at the top, SC-STs at the bottom and OBCs in between. However, the picture is not uniform across indicators, and there are signs of change.

Education and occupation

We find evidence of convergence between the three broad groups in literacy and primary education over successive cohorts. However, in access to higher education, the groups are growing further apart. The fact that for education after the primary stage, and especially for higher education critical to achieving social mobility, traditional hierarchies have not only persisted, but widened over the last 50 years is significant. This suggests that policies targeted towards closing the gaps at the higher education levels are not entirely misplaced, as the gaps would probably have been larger in the absence of such policies.

In a three way division of all jobs into agricultural, blue collar and white collar, SC-STs record the highest proportion in agricultural jobs consistently for all cohorts, followed by OBCs and upper castes; whereas for white-collar jobs, upper castes record the highest proportions for all cohorts, followed by OBCs and then SC-STs. For blue-collar jobs, the picture is mixed, in that OBCs record the highest proportions, followed by upper castes and then SC-STs. Here, change can be seen in the fact that younger cohorts among OBCs seem to be closing the gap vis-à-vis upper castes in terms of access to prestigious white-collar jobs, whereas SC-STs continue to lag behind.

Jobs and discrimination

We find that SC-ST percentages with access to public sector jobs are consistently higher than those for OBCs, a trend at variance with access to white-collar jobs. We believe that the difference in the relative picture between SC-STs and OBCs reflects the longer operation of SC-ST quotas. Note, however, that upper castes as a group continue to have the highest percentage of public sector jobs across cohorts.

Here again, the OBCs are catching up, both with SC-STs and with upper castes. This is most strikingly true for the cohort born between 1956-1965, individuals who would have been between 35 and 25 years old in 1990 and hence eligible to take advantage of the new quotas. This catch up continues onwards to younger cohorts. We see a similar convergence between SC-ST and upper castes, which is in contrast to the picture of divergence between SC-ST and upper castes in access to white-collar jobs.

A standard method of estimating discrimination in the labour market is via decomposition of the wage gap between two groups. A part of the wage gap between any two groups can be explained by the fact that members of one group are likely to be better educated, more skilled or have other attributes that command higher wages. However, if the entire wage gap cannot be explained by these factors, the residual gap is taken as a proxy for labour market discrimination. Comparing the oldest cohort born after independence to the youngest cohort, we find that average wage gaps between OBCs and upper castes have been narrowing over the years, but notably, the discriminatory component is rising. The gaps between SC-STs and upper castes are larger than those for OBCs, and the trend is again of rising discrimination for younger cohorts.

Traditional hierarchies persist

Overall, despite significant gaps in all indicators (consumption expenditure, wages, educational attainment, occupational attainment and so forth), we find substantial evidence of catch up between OBCs and upper castes among younger cohorts (especially in literacy, primary education, access to white-collar jobs, wages), but we find continued divergence in all education categories after the middle school level. The convergence for SC-STs is very limited, confined only to literacy and primary education. However, our disaggregated regional examination reveals that the association of these changes with political representation is weak, at best. That should not be taken to mean that political representation has not worked or that it is not vital to achieve an inclusive, and broad-based structure of governance and decision-making. What the evidence indicates is that traditional hierarchies are too deeply entrenched to be reversed through one single measure; they need a concerted push, backed by strong will from different segments of society, including, but not confined to, politicians.

(Ashwini Deshpande is Professor, Delhi School of Economics. E-mail: ashwini@econdse.org )