Mainstream WeeklyThe Indian Constitution reserves seats in legislatures for Scheduled Castes and Scheduled Tribes. It does not reserve seats in Cabinet, in the higher judiciary, in the Indian Administrative Service, in university vice-chancellorships, in newspaper editorial boards, in corporate boardrooms, or in the commanding heights of civil society. These are the spaces where consequential decisions are made — where budgets are written, where laws are interpreted, where knowledge is certified, where narratives about society are produced and legitimised. They are also, by any honest accounting, among the most caste-homogeneous spaces in contemporary India. The Telangana SEEEPC Survey 2024 named itself, in full, the Socio-Economic, Educational, Employment, Political and Caste Survey. The "P" — political — was placed at the centre of the survey’s declared scope. In the two volumes of the Independent Expert Working Group (IEWG) report that followed, it generated almost no analysis at all. This essay argues that this is not a minor oversight but a foundational gap — that political representation is not one dimension of caste deprivation among several, but the dimension that determines whether all other dimensions can be addressed, and that a survey which cannot account for the distribution of power cannot fully account for the reproduction of disadvantage.
What Political Representation Means — and Does Not Mean
The IEWG report identifies three analytical frameworks it considers necessary for a comprehensive assessment of caste-based deprivation: share of population, share of participation in social and economic life, and share of representation in power. It acknowledges, with admirable candour, that the SEEEPC dataset covers only the first two. Share of representation — what the report calls the third lens — is noted as absent and recommended for future study.
This framing, however, already contains a narrowing. "Political representation" in common usage is taken to mean electoral representation: the share of MLAs, MPs, and municipal councillors from each caste group. This is the form of representation most legible to survey instruments, most amenable to counting, and most frequently discussed in public debate. But electoral representation, important as it is, is the most visible and the most formally regulated form of political power — and therefore, paradoxically, the least revealing about the actual distribution of caste power in a society.
Consider what is not captured by counting legislators. A state may have SC representation in its assembly at population-proportional levels through reserved constituencies, while simultaneously having no SC person in its Cabinet with a significant portfolio, no SC person as Principal Secretary in its Finance or Home departments, no SC editor of a major Telugu newspaper, no SC vice-chancellor in any of its central universities, and no SC promoter of a business with revenues above a threshold. Reserved constituency representation produces presence in the legislature without necessarily producing presence in the rooms where legislative decisions are actually made. The gap between formal representation and substantive power is precisely what a political dimension of a caste survey should measure — and precisely what the SEEEPC exercise does not.
The Three Tiers of Political Invisibility
Political representation, understood broadly, operates across at least three tiers in contemporary Telangana, and the SEEEPC survey’s silence extends across all three.
The first tier is electoral. How many MLAs, MPs, Zilla Parishad members, Mandal Parishad presidents, and Gram Panchayat sarpanches come from which caste groups, and how does this compare to their population share and their CBI backwardness scores? This is the easiest tier to measure — the data exists in Election Commission records — and the IEWG report does not use it. A survey that ranks 242 castes by deprivation and does not cross-reference that ranking with their electoral representation misses one of the most direct tests of whether formal political inclusion has translated into substantive policy influence for the most backward communities. The SC Dakkal community, with a CBI of 116 — the most backward community in the state — how many of its members have held elected office at any level? The question is not asked.
The second tier is bureaucratic. The Indian Administrative Service and the Indian Police Service are the administrative spine of the Indian state. At the district level, the Collector and the Superintendent of Police are effectively the state. Who they are, which communities they come from, and how caste shapes their administrative dispositions and priorities is a question of immediate policy relevance in a state that is using a caste survey to redesign welfare targeting. The IEWG report notes, briefly, that 22.9% of IAS/IPS officers in the dataset identify as "No Caste." It does not analyse the caste composition of the bureaucracy systematically, does not compare it to the state’s caste distribution, and does not examine whether there is a correlation between caste representation in district administration and the performance of welfare schemes in those districts. This is a significant missed opportunity. Research in other contexts has consistently shown that bureaucratic representation affects both the targeting and the quality of public service delivery — a finding directly relevant to the report’s own welfare misallocation analysis.
The third tier is epistemic. Who produces knowledge about Telangana society — who edits its newspapers, who chairs its universities, who sits on its planning bodies, who constitutes its expert committees — is a form of political power that is rarely counted and almost never named. The IEWG itself is a small data point here: its nine members include a retired Supreme Court judge, a prominent intellectual-activist, a Congress party data strategist, two economists, and several social scientists. Their caste composition is not disclosed in the report. For a document whose central argument is that caste shapes life outcomes in ways that are invisible until measured, this is a conspicuous silence.
Why the Survey Did Not Measure What It Named
The absence of political representation data from a survey that includes "Political" in its title invites explanation. Three explanations are plausible and probably operate in combination.
The first is methodological. A household survey instrument — even one covering 97% of households — is poorly designed to capture political representation data. Asking a household whether any of its members holds elected office, or is employed as an IAS officer, will generate accurate data only for the rarest outcomes. The denominator (total households) is so large relative to the numerator (elected officials, senior bureaucrats) that the statistical signal is drowned in noise. Political representation data is better sourced from institutional records — Election Commission data, civil service employment rolls, university appointment registers — than from household surveys. The survey design was not equipped to collect it, and the IEWG did not supplement the survey data with institutional data.
The second explanation is political. A finding that dominant castes hold political power grossly disproportionate to their population share — which is almost certainly what the data would show — is a more explosive political claim than a finding that they are educationally or economically better off. The latter can be addressed through welfare redesign. The former implies a restructuring of power itself. In a state where the survey was commissioned by a government composed substantially of members of dominant caste groups, the political incentives to leave the representation question unasked are not difficult to identify.
The third explanation is conceptual. The IEWG report’s intellectual framework is fundamentally a framework of deprivation — it asks who lacks what. Political representation is not straightforwardly a deprivation in the same sense as illiteracy or open defecation. It is an absence of power, which is a different analytical category. The CBI, designed to measure backwardness through material and social indicators, has no natural slot for political power. The equal-weighting methodology that gives the CBI its transparency would be severely strained by the inclusion of a parameter like "share of legislative seats" — which is already formally regulated through reservation — alongside parameters like "proportion of open defecation." The conceptual architecture of the survey, in other words, may have pre-empted the political question before it was asked.
What Political Representation Analysis Would Have Shown
While the SEEEPC survey did not collect or analyse political representation data, existing evidence from other sources allows a provisional sketch of what such analysis would likely have found in Telangana — and why it matters.
Studies of Andhra Pradesh and Telangana legislative assemblies consistently show that Kamma, Reddy, and Velama communities — all classified as Other Castes (OC) in the SEEEPC survey and among the least backward groups in the CBI ranking — hold legislative representation far in excess of their population share, particularly in Cabinet positions and Speaker roles. SC communities, despite reserved constituency quotas, are almost entirely absent from Cabinet portfolios covering Finance, Home, Revenue, and Irrigation — the departments that control the largest budgets and the most consequential administrative decisions. BC communities, constituting 46% of the population by the SEEEPC survey’s own count, have produced only a handful of Chief Ministers in the combined history of Andhra Pradesh and Telangana. The pattern in the higher bureaucracy is similarly skewed, a finding corroborated by studies of IAS officer caste composition at the national level (Banerjee and Somanathan, 2007; Krishnaswamy, 2019).
This matters for the SEEEPC exercise in a specific and direct way. The report’s welfare misallocation finding — that 30% of ₹54,521 crore in annual welfare spending goes to communities less backward than the state average — is presented as a consequence of scheme design: agricultural support schemes linked to land ownership exclude landless SC communities. This is true. But it is an incomplete explanation. Welfare scheme design is not a technical accident. It is a political outcome — the result of decisions made by legislators, senior bureaucrats, and planning officials about whose interests welfare architecture should serve. If those decision-makers are drawn overwhelmingly from communities that own land and benefit from land-linked schemes, then the misallocation the CBI reveals is not merely a design flaw. It is a representation failure. The survey measures the outcome. It does not name the mechanism. Political representation analysis would have named it.
The Representation-Deprivation Relationship
There is a deeper theoretical point that the absence of political representation data forces us to confront directly. The IEWG report’s paradigm — measuring deprivation to redesign welfare — implicitly assumes that the state is a neutral instrument that can be redirected toward the most deprived once their deprivation is accurately measured. This is a reformist assumption, and not an unreasonable one for a policy document. But it is an assumption that the report’s own welfare misallocation finding puts under pressure.
If welfare has been misallocated for decades toward less-backward communities, and if that misallocation is partly a consequence of those communities’ disproportionate representation in the bodies that design welfare, then the solution is not only better measurement. It is also different representation. The CBI can tell the state who needs more. It cannot, by itself, create the political conditions under which those who need more are able to demand it and receive it. That requires representation — not merely in the formal sense of reserved constituencies, but in the substantive sense of presence in the rooms where decisions are made.
Ambedkar understood this. His insistence on separate electorates — overridden by Gandhi’s fast and the Poona Pact of 1932 — was grounded precisely in the recognition that formal inclusion in a common electorate would not produce substantive political power for untouchable communities, because their political leverage would always be diluted by their dependence on dominant caste political patrons. The reservation system that replaced separate electorates has produced SC presence in legislatures without producing SC power in governance. The SEEEPC survey, by naming the "P" and then not analysing it, has inadvertently reproduced this gap — measuring the visible while leaving the structural untouched.
A caste survey that does not measure political representation tells us, with great precision and methodological sophistication, how unequal the outcomes of caste are. It does not tell us how power is distributed among those who have the authority to change those outcomes. These are not the same question — and answering the first without the second produces a map of deprivation without a map of its reproduction. The SEEEPC exercise has produced a remarkably detailed map of the former. The latter remains, as it has always been in Indian social science, the harder and more dangerous question — dangerous precisely because it names not just who suffers but who benefits, and not just from history, but from the present arrangements of power that a survey commissioned by a government, analysed by an expert group, and published by a state apparatus must, in the end, find a way to survive.
(Author: Dr Santhosh Juvvaka, Independent Scholar, Hyderabad, Telangana)
References
- Ambedkar, B.R. (1945). What Congress and Gandhi Have Done to the Untouchables. Thacker and Co.
- Banerjee, Abhijit and Rohini Somanathan (2007). "The Political Economy of Public Goods: Some Evidence from India." Journal of Development Economics, 82(2), 287–314.
- Guru, Gopal and Sundar Sarukkai (2012). The Cracked Mirror: An Indian Debate on Experience and Theory. Oxford University Press.
- Jaffrelot, Christophe (2003). India’s Silent Revolution: The Rise of the Lower Castes in North India. Columbia University Press.
- Krishnaswamy, Sudhir (2019). "Caste and the Constitution." In Oxford Handbook of the Indian Constitution, eds. Sujit Choudhry, Madhav Khosla, and Pratap Bhanu Mehta. Oxford University Press.
- Pai, Sudha (2013). Dalit Assertions. Oxford University Press.
- Philips, Anne (1995). The Politics of Presence. Oxford University Press.
- Government of Telangana (2025). SEEEPC Independent Expert Working Group Report, Volumes I and II. Hyderabad.