Dr BR Ambedkar's idea of graded inequality held that caste is not a single line dividing oppressor from oppressed but a ladder, every rung resented by the one above it and resenting the one below.
That ladder explains something the data below keeps confirming about an invisibilised segment of the Indian economy — Other Backward Class (OBC) and Most Backward Class (MBC) labour.
Data on Scheduled Castes (SC) and Scheduled Tribes (ST) are recorded and published by the notified caste and tribe name in the census. But OBCs and MBCs, one of the largest social blocs in India's workforce by any credible estimate, have never once been counted as a block in independent India's history.
The last full caste count was done in 193 under British rule. This gap now is what the 2027 census is trying, imperfectly, to close. The pattern that follows, across six parts of the economy, is consistent. Categories that the state built into its data systems get measured. Categories it did not, do not.
How big is the group? How big is the gap?
Estimates of the OBC population have ranged from 41% (National Sample Survey, 2006) to 52% (Mandal Commission, 1980) — a wide spread on a question that should have one answer.
Where states have conducted their own surveys, the real number has run higher than either national extrapolation. Bihar's 2023 caste survey found Extremely Backward Classes (EBC) at 36.01% of the state's population and OBCs at 27.13%. This is a combined 63% against a reservation quota for both groups capped at 27%.
Telangana's own caste survey, released in early 2025, put OBCs at 56.33%. Both figures came from state governments doing, at their own cost, what the Union government has declined to do nationally since 1931.
Poverty data confirms the scale, without the detail
A 2022 study in PLOS ONE by Pradhan, Kandapan, and Pradhan, using data from the National Family Health Survey, measured multidimensional poverty by caste.
It found 44.4% of STs, 29.2% of SCs and 24.5% of OBCs multidimensionally poor, against 14.9% among the general category.
OBCs are poorer than the general category by a double-digit margin, but the study, like nearly every large dataset in India, cannot break that 24.5% down further, because no national-level data exists for OBCs the way it does for SCs.
The wage and disaggregation gap
The State of Working India 2023 report found that OBC-owned enterprises sit between the two extremes of India's business landscape, better represented than SC- and ST-owned firms, still far behind general category ownership of large-scale enterprises.
The report also found that caste-based underrepresentation in manufacturing ownership cost SC, ST and OBC entrepreneurs an estimated Rs 42,000 crore in 2013 prices, close to a quarter of all private proprietary value added in the sector.
Its authors could disaggregate outcomes at the jati-level only for SCs, using Census data built for that purpose. No comparable national jati-level dataset exists for OBCs.
The Periodic Labour Force Survey (PLFS) reveals the same blind spot from a different angle and shows how much is actually at stake. Its 2023 to 2024 report records that the agriculture sector employs roughly 45% of India's workforce and construction another 12%. This is more than half the country's labour concentrated in just two sectors, independently documented as heavily OBC and MBC.
This is not a minority footnote to the Indian economy. It is close to its base, and PLFS publishes none of it by caste.
Who actually benefits inside the OBC list?
Ambedkar's ladder is visible here in exact numbers.
In 2001, Uttar Pradesh's Hukum Singh Committee counted the state's OBC population at 7.56 crore across 79 distinct castes, using rural family registers. That data, later reported by Tribune and other outlets, found Yadavs at 19.40% of the OBC population, Kurmis and Patels together at 7.4%, Nishads, Mallahs and Kevats at 4.3%, Lodhs at 4.8%, Jats at 3.6%, and Bhars and Rajbhars at 2.4%, with dozens of smaller castes making up the remainder.
Bihar's 2023 caste survey shows a similar concentration at the top. Yadavs alone are 14.26% of the state's entire population. A single OBC label holds together groups with wildly different demographic weights and wildly different access to what that label is supposed to guarantee.
Land loss in Kakinada and Polepally
Mohan Rao B., former principal of the Rajiv Gandhi Institute of Law, Kakinada, conducted a socio-economic survey of families affected by the Kakinada Special Economic Zone in Andhra Pradesh. The land was acquired by the state government in 2005 and handed to GMR, which holds a 51% stake in the project.
Farmers were compensated at Rs 3 lakh per acre, a figure independently confirmed by Land Conflict Watch's own record of the acquisition.
The survey found entire displaced hamlets predominantly inhabited by a single OBC caste. Samantha Agarwal's study of the Polepally SEZ in Telangana found that land loss triggered a sharp rise in household indebtedness. This was felt most severely by SC and ST families pushed toward dominant caste moneylenders, with OBC households drawn into the same pattern, albeit less violently.
Both cases show the same failure. Compensation rolls track who lost land, but not systematically by caste, so the caste dimension of India's SEZ-era displacement survives only in scattered, localised studies rather than in any national account.
The Vaddera stone cutters of Telangana
As a research fellow with WorkFREE, a Universal Basic Income pilot in Hyderabad, I worked closely with the Vaddera community, traditionally stone cutters, quarry labourers and makers of stone kitchen utensils.
Mechanisation and changing consumption patterns, financed largely by other castes, made this trade economically unviable within a generation. Many Vaddera households in Hyderabad's municipal wards have shifted into garbage collection under the Greater Hyderabad Municipal Corporation (GHMC). Framed in policy language, this looks like a shift into urban employment.
But on the ground, it is not mobility but displacement — one caste-marked form of manual labour for another. Residents routinely assume that these workers are regular GHMC employees, drawing a government salary with its usual protections. They are not.
They work without safety equipment or social security, often travelling long distances to worksites that offer neither. A separate 2022 study of the same community by the Centre for Economic and Social Studies (CESS) in Hyderabad corroborates this independently, documenting young Vaddera school dropouts turning to garbage collection and casual labour once the family trade collapsed.
What the 2020 lockdown showed
The Union government told the Supreme Court in April 2020 that roughly 600,000 migrant workers had already walked back to their villages. A 2022 panel study by Allard and colleagues, tracking migrants displaced from Bihar and Chhattisgarh, found more than 85% of the sample belonged to SC, ST, OBC or EBC groups. In comparison, only 7% of male migrants and 3% of female migrants were from the general category.
A separate 2020 study by development researcher Nitya Rao and colleagues followed men from Bihar's Chakai block migrating to Kerala, Gujarat, Uttar Pradesh and Maharashtra. Most of the men were from ST and OBC communities. This study too found the same pattern: stranded, jobless overnight without food or shelter, and in a category the state was not tracking closely enough to send help to.
Reporting in The India Forum found labour contractors in brick kilns across Andhra Pradesh and Telangana and on Delhi-NCR construction sites routinely controlling workers' wages and remittances through debt. These arrangements disproportionately fall on SC, ST, OBC, and EBC workers.
No data on the gig economy
A 2024 ethnographic study by political scientist Dyotana Banerjee, based on fieldwork in Ahmedabad between 2016 and 2020, found food delivery gig workers are overwhelmingly drawn from SC and OBC communities in Gujarat. The study describes gig work as a route through which caste-based solidarity networks, not an escape from caste, dictate who finds work.
Despite the gig economy’s scale, no comprehensive national study of its caste composition exists.
Political representation hasn’t closed the economic gap
Christophe Jaffrelot termed the rise of OBC political representation since the 1990s India's silent revolution.
In a study, economists Ashwini Deshpande and Rajesh Ramachandran examined whether that revolution reached OBC workers economically. They found substantial gains and catch-up between OBCs and the general category among younger cohorts in literacy, primary education, and access to white-collar jobs, alongside continued gaps in wages and higher education.
But their regional analysis found that OBC economic outcomes were not closely related to the strength of OBC political representation in state assemblies.
Economists Sukhadeo Thorat and S Madheswaran found that discrimination accounts for roughly 28.5% of the wage gap between SC communities and dominant castes. It was 19.4% in the public sector and 31.7% in the private sector, operating substantially through occupational segregation against both SCs and OBCs.
Representation in a legislature does not, on its own, touch this mechanism, because it operates at the point of hiring, not the point of voting.
What the data adds up to
Every dataset above measures a different domain, and every one runs into the same wall. SC and ST outcomes can be disaggregated because the state built the categories to allow it. OBC and MBC outcomes cannot be, because it never did.
This is a population that grows a large share of India's food, builds much of its cities, walks its highways when the cities shut down, and now delivers its meals.
It is, in the plainest sense, a substantial and invisibilised segment of the
Indian economy. It is indispensable to how the country runs and absent from how the country counts itself.
Ninety years after the last full caste count, India allocates a 27% quota to a population its own state surveys put at 56% to 63%, and still cannot tell a Yadav household's outcomes from a Kahar household's on any national dataset.
The 2027 census is finally asking the question, but imperfectly: OBC and MBC responses need to be coded at the point of collection, the same way SC and ST data have been since 195. They should not be gathered freely and classified afterward as was the case with the 2011 caste census. Ambedkar warned that a graded system survives precisely by keeping each rung too fragmented and too busy resenting the rung below to demand the same accounting as the rungs above it.
The fix does not require new theories.
Statistician PC Mahalanobis, architect of India’s data system, argued that statistics must serve two purposes: advancing science and promoting human welfare. By that measure, data on OBC and MBC labour fails twice over. It is rarely collected at the caste level, and where it is, as in a handful of state surveys, it remains disconnected from the wages, land records, and welfare schemes it should inform.
That failure has consequences. A backbone becomes visible only when it is named, a census column that records caste, something India has not attempted at scale since 1931.
It becomes real only when that name carries into labour law, wage slips, and gig‑economy contracts, wherever OBC and MBC labour is present yet unaccounted for. This is not a niche demand; it is the demand of half of India, and it deserves to be answered as one.
Sai Ganesh Akarapu is a researcher whose work engages political and sociological theory to examine caste, labour, and representation, with a particular focus on the structural exclusion of OBC and MBC communities.
His research is shaped by his OBC background and grounded in enduring concerns of justice, dignity, and recognition.
Opinions expressed are the author’s own.