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

Caste and Class in Higher Education Enrolments

Challenges in Conceptualising Social Inequality

Amman Madan ( teaches at the Azim Premji University, Bengaluru.

The National Sample Survey 2014 data is used to explore the character of class inequality, over and beyond that of caste. The caste break-up of various social classes suggests that caste inequalities are greater amongst the more educated classes. Enrolments in higher education show greater social inequalities than in elementary education. The differences amongst various classes suggest that while caste is a strong factor in educational inequality, it is not a sufficient one. There is much less caste variation within the lower classes than the higher classes. Caste and class need to be seen as generative processes, and sub-jati networks are to be conceptualised and empirically examined to understand the actual roles of caste and class in educational and social inequality.

This paper was presented in seminars at the School of Education, Azim Premji University and School of Education, Tata Institute of Social ciences. Comments and questions received are gratefully acknowledged.

The way in which we theorise social life has played a key role in understanding and addressing the challenges to human dignity and freedom. The theorising of social inequality has made important contributions to the struggle to create an open education system by conceptually guiding our interventions, ranging from caste-based reservations to the development of inclusive forms of teaching, amongst others. This paper will examine two out of the several ways of conceptualising social inequality—caste and class—which are well known for their limiting and corrupting influences on meritocracy and the education system.

In the popular and public culture, we are seeing a broad shift from a time when class used to dominate discussions of inequality to a situation where class has mostly disappeared from the conceptual armoury. Gender and caste are the issues most prominent in our imagination these days, parti­cularly amongst anti-caste voices. This shift is, in part, an important corrective to the study of social inequality, which was earlier dominated by the discourse of class and redistribution. The importance of identity and culture in shaping social structures and the injustices inbuilt within them is now better recognised. On the other hand, there is the danger of ignoring social structures and processes that lie at a deeper level and are not part of the self-identity of mobilised or oppressed groups. There are many kinds of relations of exploitation and oppression between labour, political organisation and capital, which may or may not be part of the consciously articulated categories of the people involved.

Karl Marx had distinguished between class-in-itself and class-for-itself as structural and ideological entities. The very notion of ideology suggests that our knowledges and cultures are politically charged, and certain articulations may get suppressed at the cost of others. There may be structural inequalities that are present but do get clearly articulated, given the power relations of a historical moment. The search for social structures, which are not part of our contemporary theoretical apparatus but do lead to social inequality, is thus a struggle to overcome hegemonic ideologies. This makes an important contribution to the struggle for a more open and free society. Such structures must be deliberately searched for and analytically exhumed since our cultures and the ideologies we live in may not fully recognise them. Evidence must then be looked at with new lenses to see if it supports the reworked concepts.

There may be many structures that escape our current ideologies and which the terms of class, caste, and gender only inadequately represent. The term “intersectionality” has become a common way of flagging that there are many kinds of social inequality and they interact with each other in complicated ways. These may reinforce or sometimes even neutralise each other. We are increasingly aware that we need to conceptualise class, caste, gender, etc, not just individually but in terms of their interaction with each other. A system of social inequality may be said to be a unique and discernible system to the extent that its elements are integrated in a way that is different from other possible systems. The idea of intersectionality asks us to question that purportedly separate and watertight existence of systems of inequality. It emphasises that even when systems appear to be distinct, there may be elements that are shared across them, and the systems themselves may collide and collude in complicated ways.

While there is a good amount of work nowadays on the interweaving of caste and patriarchy (for example, Dube 1996; Devika et al 2013; Guru 1995; Rege 1998), this paper wants to draw attention to the ways in which class and caste are connected, and yet represent two distinct social processes. There is considerable scholarly work on this connectedness (for example, Beteille 2007; Ghurye 1957; Mukherjee 1999; Sharma 1984). This paper aims to spell out certain empirical dimensions that often go unnoticed and uncommented upon. Apart from their intrinsic value, they may also help us see in perspective recent moves for quotas for the poor amongst “upper” castes. Demands for reservations for economically weaker sections appear to eschew questions of evidence and understanding. This paper may help us to see better the complex and intersectional character of social inequality in India.

Secondary data available in the 71st round of the National Sample Survey (NSS), conducted in 2014, which had education as one of its specific focus areas, is examined so as to estimate the connections between educational inequality and contending ways of theorising systems of social inequality. Enrolments provide a simple, though thin, way of measuring educational inequality. The distribution of households and enrolments has been studied across various caste groupings on which data was gathered by the NSS. Due to limitations of space, the patterns to be seen in the expenditure made on education and enrolments in English-medium institutions will not be discussed here, but they show a similar pattern as that of overall enrolments.

The data has been examined through the statutory caste categories as defined by the Constitution and through class categories that have been operationalised in terms of monthly per capita consumption expenditure groups and as occupational groups, drawing from the work of Erik Olin Wright (1996) and John Goldthorpe (2000). Through such an analysis, a nuanced answer is sought to be given to the question of whether and how we need to think beyond caste when we try to understand and engage with educational inequality. This paper will argue eventually that we must take class inequality seriously, since caste is unable to explain a substantial part of the social inequality in society and education in India. It will also argue for combining the varna–jati model of caste with a model of caste as smaller kinship networks, which operate nested within a large system of secular economic, cultural and political inequality.

Caste Composition of Classes

We should acknowledge right at the outset the difficulties with data sets that use categories like Scheduled Castes (SCs), Scheduled Tribes (STs), Other Backward Classes (OBCs), and forward castes. These do not correspond closely with either the varna scheme or the pattern of dominant castes of a specific region. Within each category, there is a wide range of jatis, parti­cularly so in the OBCs and the forward castes. Yet, if we want to look at the large-scale data sets, this is the kind of data that is available in most of them. It is with caution that we must interpret what they tell us about the caste system. The problems with class information are a little less, but still considerable. Here, only broad-brush conclusions are being drawn from them and they appear to be valid for such kind of data.

The STs make up 9.5% of the households included in the survey, while the SCs are 19.0%. The OBCs are the largest single category with 42.7% of all the households. The forward castes are a little more than a quarter, that is, 28.8% of the surveyed households. The impact of caste identity as a process of social inequality can be seen from Table 1 (p 41), which depicts the over- or under-representation of particular cate­gories in different economic classes. That caste has an impact on social inequality is supported by this table. Perhaps, the most powerful occupational group as extracted from the data may be said to be Class I, the owners and chief executives of businesses, and Class II who are elected representatives and upper white-collar workers, including managers, professionals, academics, and so on.

The owners and chief executives make up a somewhat heterogeneous group, since the data does not permit easy separation of the owners of tiny one-person businesses from those of large businesses. However, even with this hetero­geneity, just 4.8% of the households reporting the ownership or control of businesses as their main occupation are from the STs, which is 0.5 times the number that should have been there if there were no systemic inequality (that is, instead of 4.8%, they should have been the same as the proportion of ST in the population, that is, 9.5%). The SCs are 10.4%, which is 0.5 times the number that should have been there without any systemic inequality. Meanwhile, the OBCs are 42.5% of the group, which is almost the same number that should be seen without systemic inequality. The forward castes are present in this class in a disproportionately large number. They make up 42.3% of this group, which means they are over-represented 1.5 times compared to their number in the population. Caste clearly continues to have an impact upon who is the most wealthy and powerful in this country, though it would not be accurate to say that it is the only determinant at work.

The poorest and the least powerful occupational group is arguably the Class VII (Table 1) households of landless agricultural labourers, forestry workers, fisherpeople, and so on. The STs are 13.4% of these households and are over-represented 1.4 times compared to what we would have expected if there were no systemic inequalities. The SCs are 32.7% of these households and are over-represented slightly more than even the STs, being 1.7 times of what we would have expected if there were no systemic inequality. The OBC interestingly are a little under-represented, being 41.1% of these households. The forward castes are 12.7% of these households and are strongly under-represented, being only 0.4 times what we would have expected if there were no systemic inequality. Caste seems to have a substantial impact even upon who is at the bottom of the class structure.

Another social class, which may be considered one of the most powerful in the country, is that of the upper educated workers. This includes managers, professionals, senior government employees, professors, and so on. It shows a distribution pattern similar to that of owners but is a little more skewed in favour of the forward castes. While amongst owners, the forward castes were over-represented by 1.5 times, amongst the upper educated workers, they are 1.8 times to their number in the population. A little over half of all households reporting this class (51.6%) come from the forward castes. In other words, caste skews who gets into the upper classes and being from an “upper” caste shows the greatest impact upon the membership of that particular class which is directly shaped by education.

Interestingly, caste seems to be showing the least impact (though it is discernible there as well) upon those who are only slightly above the bottom, namely the unskilled workers in occupations that are typically urban and do not require much schooling (Class IV). The ratio of over/under-representation for the STs/SCs/OBCs/forward castes is 1.0, 1.2, 0.9, 0.9, respectively in this class. A slightly greater impact of the caste system may be seen in the membership of skilled urban workers (Class VIII in Table 1).

So it is amongst the most powerful and the most powerless that caste seems to speak the loudest. Amongst classes that are formed by hereditary economic capital, like those of owners and farmers, it does not surprise us that caste inequality has a pronounced impact. After all, caste is about inheriting privilege. But education has often been pitched as a way of equalising society, as a way of creating human capital where economic capital cannot be redistributed. That caste inequalities amongst the educated classes appear to be even more than the propertied classes raises difficult questions.

Caste and Educational Enrolments

This connection between caste and the educated classes leads us to look at the relation between caste and educational enrolments. Unfortunately, the NSS data does not differentiate between enrolments in better or worse kind of institutions, where intensive and stimulative teaching takes place versus where there is only mechanical teaching and where learning occurs on paper alone. We may expect that the more advantaged will be present in larger numbers in the better kind of educational institutions. Be that as it may, with all their limitations, the NSS data tells us that the gap between different caste categories is rather low in elementary school enrolments (Table 2). The separation between the STs and the forward castes is just 6.4 percentage points. This is no doubt the result of concerted efforts by governments around India to improve school enrolments. However, we cannot say from this table whether the gap is the same in the better schools or is larger there. There is plenty of anecdotal evidence to suggest that the gap begins to widen as one move to schools with better teaching and learning.

When we look at progressively higher kinds of education, the gap begins to further widen. 39.5% of the forward castes in the age group of 18–21 are enrolled in some kind of tertiary education, while only 28% of the OBCs are enrolled. The enrolment rate drops even more drastically with the SCs (20.8%) and the STs (15.7%). Thus, disparity amongst the caste groups is much sharper in tertiary education than elementary. Again, we cannot say from this data what is happening in elite institutions; we do not know if the differences are greater there than in humbler colleges and polytechnics.

We should connect this observation of the sharpest caste disparities being in higher education with the previous observation that the inequalities of caste are expressed the most in non-agrarian occupations that recruit on the basis of higher education degrees. This is quite an important thing to note. Caste inequality is greater in higher education than in other sectors, in spite of many years of efforts to make education a site for the equalisation of castes. This frustrating observation reminds us of the importance of questions regarding the nature of the caste system: how is it practised, and why is it that it shows such a strong impact upon education in spite of all the years of reservations?

Caste and Inequality

The discussion so far has been in line with what is already well known, that, caste patterns substantially reinforce class inequality. However, this data can also be read to show that there is a considerable divergence between class and caste locations. To begin with, it should be pointed out that only a small number (a little over a third) of the forward castes are enrolled in tertiary education. It is not as if the forward castes are a group of castes that is completely saturated with higher education. The fact that almost two-thirds of the forward castes are not enrolling or are dropping out of tertiary education is striking. It suggests that while caste matters for enrolments, again there is something else that is at play, which is generating a pattern of dropouts and enrolments. The presence of a relatively small number of forward-caste enrolments does not fit very well into a “caste directly leads to educational inequality” kind of view.

There is undoubtedly a dwindling proportion of STs, SCs, and OBCs within the upper levels of the Indian population, and these caste groups are present in increasing proportions as we look at the lower classes. But the story is more than this. The overall class structure and the presence of different caste groups within each class is better understood by looking at the data presented in terms of its proportions within the entire population and not just as percentages within a particular class. It nudges us to rethink the role of class and other social processes in the structure of social stratification and their roles in the challenge of building a meritocracy. Table 3 shows the relative numbers of different caste groups within each class, calculated as a weighted percentage of the entire sample (also in Figure 1). This is a better mapping and representation of different castes and classes in India. Again, we can see that caste undoubtedly seems to influence class inequality. But what also comes out clearly is that a substantial number of the poor cannot be said to be from the lower ranking castes.


A significant aspect of the Indian class structure is that the owners and the higher and lower educated workers, all together make up less than a fifth of the total population. As the preferred destination for upward mobility, this is a rather small platform onto which vast numbers are seeking to climb. The ladder to get here goes partially through education, ranging from elementary to tertiary levels. Those who wish to get into these classes find that the odds are stacked strongly against them. The window into this platform is rather small and getting through it calls for a lot of agility. As can be seen through the preceding discussion, the castes of one’s ancestors are clearly one factor that seems to aid or hinder that movement.

However, it is difficult to say that caste alone is at play. The proportions of various caste groups in the poorest classes, unskilled non-agrarian workers (Class IV) and agrarian, fishery, forestry workers (Class VII) tell an interesting story (Table 3). The numbers of the STs and SCs are disproportionately large, of course. But the numbers of the OBCs and even the forward castes are quite large amongst these poorest classes. The SCs in Classes III and VIII (the majority of uneducated manual workers) is 9% of all Indian households. This, of course, is a disproportionately high representation of them in these classes. On the other hand, the forward castes who are in these classes are not a tiny number either, being 6.7% of all Indian households. The number is so large that we cannot accept a simple equation of caste with class. A more complicated statement is called for.

The numbers of households of the forward castes from the upper and lower educated classes (Classes II and III) at 5.4% of the total households are actually lower than their numbers of households from the uneducated manual workers (Classes IV and VII), which are 6.7% of the total. In other words, there are more forward castes in the poorest classes than in the educated classes. If caste alone were responsible for the class location, we would not have seen such a pattern.


It is hardly surprising if the inequalities within a social structure also carry on to influence its education system. The reverse is also true; efforts within the education system to increase access and to create more inclusive curricula and pedagogies may have a balancing influence on the stratification system. There are limits to what can be achieved through education alone, of course. But it is important to acknowledge that the education system has a relative autonomy and deserves attention in itself, and not just as a mirror image of the rest of the social structure.

Enrolments provide a thin though still a useful way of examining stratification. I have examined the differences within enrolments in tertiary education institutions elsewhere using the 66th NSS survey and will not go into great detail of that again (Madan 2020). It will suffice to recall from it that there are strong caste as well as class patterns in enrolments into tertiary education. In certain streams like engineering, the ratio of forward castes to SC students is of the order of 11:1. In streams like medicine, commerce, and arts, it is closer to 3:1 (Madan 2020). Within each stream and caste group, there is a further influence of class. For instance, amongst SC engineering undergraduates, more than half are from the educated and owner classes.

In this section, I look not at who is enrolled, but at who is left out. Overall enrolments have improved tremendously in India. In 2014, the year of the 71st survey, 60.5% of all individuals between the ages of 18−21 were enrolled in some kind of school or college. Out of these, 28.5% were enrolled in tertiary education. A study of who dropped out or never enrolled, however, will give us a clearer picture of how social stratification operates to obstruct the functioning of a meritocracy. Table 4 shows the individuals between the age group of 15−17 who have been reported as having dropped out or never enrolled. Here, we will use “dropout” as a shorthand for dropout and never enrolled. When we look at the dropouts as a whole, there is a visible gradient between the forward castes, with a dropout rate of 19.63%, and the other caste groups. There is a substantial gap between the forward castes and the OBC, with the latter at a dropout rate of 28.33%. There is a slightly narrower gap between the OBCs and the SCs (33.3%) and STs (37.54%). The STs have a dropout rate of nearly twice as that of the forward castes. Caste clearly has a role to play in who drops out in secondary school.

Looking at the data by clustering it around the family’s monthly per capita consumption expenditure (MPCE), however, puts things in a somewhat different light. We now begin to see the relation of dropouts and non-enrolments with class as well as caste. Among 15−17-years-old, the forward castes in the poorest MPCE band of less than `1,000 have a dropout rate (42.59%), which is hardly distinguishable from the OBCs (41.39%) and the SCs (43.12%). Only the STs are significantly worse off-with a dropout rate of 48.3%. But, the overall difference amongst the poorest in terms of consumption is much less than the average overall difference across different MPCE bands. The different caste groups are much closer amongst the poor than amongst those slightly better off. The dropout rate declines sharply with an increase in the family’s ability to consume. Unfortunately, the figures reported in the survey are too small to draw any reliable conclusion on the comparison across castes in the higher consumption bands. But the gap does seem to reduce with decreasing family income. Overall, it can be safely said that low family incomes are a very strong contributor to dropouts and non-enrolments across each and every caste group.

The impact of caste amongst the poorest classes is seen to strengthen as we go up the ladder of education. The difference between the forward castes and the SCs was not tangible in the 15−17-years-old, who may be in and about the secondary education level, amongst those with an MPCE of less than `1,000, but is quite clear in dropouts and non-enrolments amongst 18−21-years-old (tertiary education levels) (Table 5).

However, when we go up the class ladder, there is a dramatic decline in dropouts amongst all caste groups. Being poor makes a big difference to tertiary education enrolment of not only the SCs and the STs, but also makes quite a large difference to the forward castes. Figure 2 shows the absolute numbers of dropouts and non-enrolments amongst 18−21-years-old in different MPCE bands. It gives a way of comparing the absolute numbers of dropouts from different caste groups and classes. It may be seen that the absolute numbers of forward-caste students who do not enter the tertiary education system are not insignificant. While the dropouts of the ST and SC are still disproportionately larger, the dropouts amongst the forward castes are also quite substantial. There is clear evidence of the influence of caste in educational inequality, but it is not possible to sustain that caste is the only factor at work.

Regional Variations

Another way of examining the relation between class and caste could be to take up a comparison of less and more developed regions and study the dropouts there. In this paper, the most and least developed regions of Karnataka are compared. The Bengaluru region is a place where rapid growth has taken place in recent years. We define this region as made up of Bengaluru urban, Bengaluru rural, and the adjacent Ramanagara districts. For comparison, the most economically backward region in Karnataka was chosen, which is widely considered to be Kalyana Karnataka, comprising of the districts of Bidar, Kalaburgi, Yadgir, Raichur, and Koppal.

A marked difference can be seen between the Bengaluru and Kalyana Karnataka regions. The dropout levels in Bengaluru vary sharply from the STs to the SCs, OBCs, and forward castes among 15–21 year old. The STs have drastically high dropouts here (70.64%), and the SCs (61.47%) too are dropping out in much larger numbers than the OBCs (38.58%) and forward castes (31.85%). Interestingly, in the Kalyana Karnataka region, quite a different picture emerges. While the differences between the different caste groups are comparable to the differences seen in Bengaluru, the forward castes have a much higher dropout rate in Kalyana Karnataka (55.37%) (Table 6).

The considerable difference in the dropout and non-enrolment patterns between Bengaluru and the Kalyana Karnataka regions again tells us that we need to complement the analysis of caste with some other factors too. The history of a region, its pattern of development, and the processes shaping the class structure there have a considerable impact upon how caste operates in education and society. Perhaps, the way the state and various institutions providing education operate also influences the patterns here. The way we understand social inequality and its relation to education has to face the challenge of explaining all these variations.

Alternative Models

A clear caste-based pattern can be seen of inequality in access to wealth and resources and also in educational enrolments, particularly those of tertiary education. However, there are vast disparities within the caste categories themselves and not just across them. This forces us to pay attention once again at class, gender, regional and other kinds of disparities as sources of educational inequality. We have focused here for reasons of space on class, since the patterns created by patriarchy and historically specific forms of development while interconnected with class need an elaborate discussion that may be too much to take on in this paper.

The wide range of variation within each caste group makes one wonder if we are looking at the right kind of unit of caste. A widespread understanding of the caste system has been to see it as made up of four varnas plus the excluded antyaja. These are said to have been generated through the consolidation of different class strata: (i) priests, (ii) warriors, (iii) farmers and merchants, (iv) artisans and labourers, and (v) the marginalised and conquered people, who are also sometimes fused with artisans and labourers. This has been compared with the estates of medieval Europe, though there are also significant differences between the two. The divergence between this model and the reality of caste in recent years has been remarked upon by many scholars (for example, Cohn 1987; Dirks 2001; Ghurye 1957; Srinivas 1976). While the varna model may have corresponded to the reality of social inequality, in some parts of the South Asian region in ancient times, migrations and movements of people into different territory over the centuries, coupled with social and economic upheavals and conquest of groups by one another, have led to an emergence of a very different social structure. Scholars like S C Dube (1990), M N Srinivas (1976), and G S Ghurye (1957) amongst many others have argued that a smaller group, which they called the jati or the sub-caste, has become the functional unit of caste. The jati is defined by being the theoretical population within which marriage is permitted, or the endogamous unit of this social structure.

The patterns of wide class disparities within caste categories reinforce our rejection of the varna theory of caste. It cannot explain why, for instance, half of the forward castes are not getting into tertiary education. A jati theory of caste, which uses varna only to legitimise itself in a changing and dynamic way, is much more plausible. The patterns to be seen are consistent with the presence of a large number of jatis, which may or may not have a close connection with varna, but act to support their internal members in a competitive manner. Some of those jati collectives may be placed higher in the class and political structure and some placed lower. This is consistent with how the above-mentioned scholars and many others have documented the way caste works in different regions. The most powerful group is the dominant jati, and this may not conform to the varna scheme of hierarchy. To give just one illustration, in Punjab, the dominant jatis a hundred years ago were the Aroras and Khatris (erstwhile traders and merchants, technically of the Vaishya varna), and now are the Jats (erstwhile farmers, technically again of the Vaishya varna). The hold of the varna image of caste is partially because dominant groups everywhere find it a convenient ideology to claim a higher varna status for themselves. So, nowadays, voices insisting that the Khatris and the Jats are “actually” Kshatriyas are getting louder. Since there are hardly any Rajputs in Punjab, there is no strong resistance to this claim.

A third model of how caste operates is that of a smaller kinship network within the jati. This is also consistent with the NSS data, and may actually better explain the wide variations within each caste grouping than the jati model alone. The government’s classification of SCs, STs, OBCs, forward castes, etc, is done on the basis of jatis and not varnas. The marked differences to be seen within each of these categories suggest that the jati too is more heterogeneous than what was earlier thought. Unfortunately, the granularity of the data available does not allow us to examine kinship networks in greater detail. The kinship network model was mentioned by scholars like the prescient Ghurye (1957) and Bernard Cohn (1987). According to this, the functional unit of caste where support is offered, cultures are shared and reproduced, and marriages are actually contracted is smaller than the theoretical boundary of the jati. A jati, for instance, may have been spread across a large geographical area due to historical migrations. Marriages and kinship relations may be maintained only within a smaller circle. It is this which then shares the most immediate advantages and disadvantages of caste identity, not the jati or the varna.

It is such a network of kinship that Claude Meillasoux had in mind when he spoke of many countries of the world being run by a lineage-based mode of production (Meillasoux 1973, 1981). Such a model is quite consistent with the wide range of economic disparities that the NSS is reporting within each caste category. The variations of the fortunes of smaller networks seem actually more plausible than the variations in the fortunes of entire jatis altogether. There may also be a combination of the jati and the kinship-network model at work, with the balance tilting towards the latter as inequalities grow between regions. Conversely, as access to technology makes communication more possible, we may expect that larger units like jatis may become more cohesive. This may explain the relatively greater homogeneity within endogamous groups, like those amongst the Tamil Brahmins (Fuller and Narsimhan 2010). In contrast, where communication is less easy, the better economic and political conditions of some families will not translate easily into an improvement of others who share the jati name.

Theories of Inequality

The way we are able to theorise social inequality guides the way we may struggle against it. Quite clearly, we need to take class inequality very seriously, since without theorising and identifying its dynamics, we will not be able to work upon it. At the same time, closed kinship networks of the advantaged, which sometimes escape the way we commonly label caste, are a very real problem for meritocracy. Understanding how they work and trying to get closed systems to open up is an important challenge.

It may be better to look at caste and class as generative mechanisms rather than to visualise them as frozen categories and social morphologies. Such generative mechanisms or codes make up the environment within which individuals and groups struggle and negotiate through discursive practices. The approach of looking for generative codes has been developed upon by several social theorists and philosophers of science, including Pierre Bourdieu (1990), Mary Douglas (2001), and Roy Bhaskar (1998). These can be thought of as basic underlying principles that unfold into the many surface patterns of social inequality. While inquiring into the generative codes, one must also try to understand social inequality in its unique everyday character, and pay attention to the narratives and meanings constructed by actors through which they make sense of and combat and negotiate with social structure.

People are not puppets in the hands of generative codes; they also reflect and create ideas and practices that may challenge or reinforce them. To understand the challenges inherent in trying to create a meritocracy, we need to look at both structure and agency together, not just one of the two. This is not a positivistic search for timeless universal laws. Instead, it is a search for structural processes that help us understand how historical trends move. This is the sense in which some people have interpreted Marxist and Weberian traditions and have sought to use concepts like capitalism and rationalisation as historical generative processes to understand growing patterns and contradictions of our times.

The work of Meillasoux (1973, 1981) and Anand Chakravarti (2001) can be seen as standing within this approach of seeing caste and class as generative processes. They argue that caste operates as a form of social reproduction that can be alive within larger systems, like those of capitalism. Anthropologists like Maurice Godelier (1986) argue that kinship systems are a way of organising a society. As Meillasoux put it, they may form a lineage mode of production that is organised around lines of descent and marriage. The networks of descent and marriage operate as a core organisational component of production. Caste networks can thus be seen as one way of reproducing the organisational bases of production. In many parts of India, today educational strategies and institutions play a key role in this social reproduction and also contribute to changing the ways in which the larger system seeks to reproduce itself. They therefore become an important site for struggles by various social groups to enhance their power and prestige.

Caste and sub-caste networks operate to gather status, power, and wealth, within historically changing and specific environments. The character of these environments has, in recent years, been shaped by the Indian state, by capitalism and universalising processes of rationalisation and differentiation. Endogamous networks give advantages for economic and educational activities, but those are no longer the only way to gain advantages for individuals and families. Caste identities have become only one out of several resources within the larger system being constituted by specific state-capitalist formations. It is important to try and understand the broader character of social inequality. It is contributed to but not completely shaped by caste. It should also be kept in mind that there is a great deal of regional specificity in social inequality, and we need to see the way kinship networks play out in different situations.

Caste should be seen as a status group operating within larger political and economic systems. The salience and strength of a local status group will depend upon how it negotiates with local conditions. The more powerful caste groupings within the Bengaluru region are able to achieve a greater disparity in education from the rest than the disparities to be seen in education in Kalyana Karnataka. The variations across different regions suggest that the character of the larger systems of production and politics play a crucial role in shaping educational inequality, and it is not sufficient to explain it by referring to the reproduction of caste alone. The salience or otherwise of a local status group is influenced by the processes of class, politics, and culture (including kinship). Caste does not work as a nationwide varna but more as a regional jati and as a smaller set of kinship networks too that vary from region to region. The identities they give themselves may or may not correspond exactly with the actual functioning of the networks.

It is necessary, therefore, to try and understand the changing political economic conditions which lead different caste networks to benefit or not benefit from them. Those do not follow the logic of caste but have a logic shaped by political democracy and by the capitalism-state alliance. It is beyond the scope of this paper to try and outline what those generative mechanisms are. The point being made here is a relatively smaller one: if we wish to understand what obstructs merito­cracy and what are the ways of creating a society where people are indeed able to flourish without being held back by social structure, we need to understand not just one but several different kinds of generative mechanisms. In schools and universities, we need to better understand how social networks operate and how they negotiate with the larger institutional, economic, political, social, and cultural environment. It is then that we will be able to better intervene so as to create a more level-playing ground for everyone.


Beteille, Andre (2007): “Classes and Communities,” Economic & Political Weekly, Vol 42, No 11, pp 945–52.

Bhaskar, Roy (1998): The Possibility of Naturalism: A Philosophical Critique of Contemporary Human Sciences, London and New York: Routledge.

Bourdieu, Pierre (1990): The Logic of Practice, Stanford: Stanford University Press.

Chakravarti, Anand (2001): “Caste and Agrarian Class: A View from Bihar,” Economic & Political Weekly, Vol 36, No 17, pp 1449–62.

Cohn, Bernard S (1987): “Notes on the History of the Study of Indian Society and Culture,” An Anthropologist among the Historians and Other Essays, Delhi and New York: Oxford University Press, pp 136–37.

Dirks, Nicholas B (2001): Castes of Mind: Colonialism and the Making of Modern India, Princeton, NJ: Princeton University Press.

Douglas, Mary (2001): Purity and Danger: An Analysis of the Concepts of Pollution and Taboo, 1966, London and New York: Routledge.

Dube, Leela (1996): “Caste and Women,” Caste: Its Twentieth Century Avatar, M N Srinivas (ed), New Delhi: Viking, Penguin India, pp 1–27.

Devika, J, Kalpana Kannabiran, Mary E John, Padmini Swaminathan, Samita Sen and Sharmila Rege (2013): “Intersections of Gender and Caste,” Economic & Political Weekly, Vol 48, No 18, pp 35–36.

Dube, S C (1990): Indian Society, New Delhi: National Book Trust.

Fuller, C J and Haripriya Narasimhan (2010): “Traditional Vocations and Modern Professions among Tamil Brahmans in Colonial and Post-colonial South India,” Indian Economic & Social History Review, Vol 47, No 4, pp 473–96.

Godelier, Maurice (1986): The Mental and the Material: Thought, Economy and Society, London: Verso.

Ghurye, G S (1957): Caste and Class in India, 1950, 2nd ed, Bombay: Popular Book Depot.

Goldthorpe, John H (2000): “Social Class and the Differentiation of Employment Contracts,” On Sociology: Numbers, Narratives and the Integration of Research and Theory, Oxford: Oxford University Press, pp 207–29.

Guru, Gopal (1995): “Dalit Women Talk Differently,” Economic & Political Weekly, Vol 30, Nos 41 and 42, pp 2548–50.

Madan, Amman (2020): “Unpacking Intersectionality: Conceptual Explorations of Convergences and Divergences of Class, Caste and Gender in Education,” Knowledge, Politics and Education: Proceedings of International Conferences of Philosophy of Education (2013−2017) Volume II, Prakash Iyer (ed), Goa: CinnamonTeal, pp 136−73.

Meillasoux, Claude (1973): “Are There Castes in India?” Economy & Society, Vol 2, No 1, pp 89–111.

— (1981): Maidens, Meal and Money: Capitalism and the Domestic Community, 1975, Cambridge: Cambridge University Press.

Mukherjee, Ramakrishna (1999): “Caste in Itself, Caste and Class, or Caste in Class,” Economic & Political Weekly, Vol 34, No 27, pp 1759–61.

Rege, Sharmila (1998): “Dalit Women Talk Differently: A Critique of ‘Difference’ and Toward a Dalit Feminist Standpoint Position,” Economic & Political Weekly, Vol 33, No 44, pp WS39–46.

Srinivas, M N (1976): Remembered Village, New Delhi: Oxford University Press.

Sharma, K L (1984): “Caste and Class in India: Some Conceptual Problems,” Sociological Bulletin, Vol 33, Nos 1/2, pp 1–28.

Wright, Erik Olin (1996): Class Counts: Comparative Studies in Class Analysis, Cambridge: Cambridge University Press.


Updated On : 28th Jul, 2020


(-) 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