The Problem You Can Measure
Walk into any Indian corporate boardroom. Look at the names on the doors. Research by scholars cited in the journal Social Exclusion and Caste Discrimination in Public and Private Sectors (Institute for Social and Economic Change) shows that board positions in most Indian firms are occupied overwhelmingly by a small number of upper castes. India's economy is growing at 7 percent a year. But not everyone is growing with it.
India's caste system groups people by birth into a hierarchy. At the top are Brahmins (priests and scholars), then Kshatriyas (warriors), then Vaishyas (traders), then Shudras (laborers). Outside the system entirely are Dalits - historically called "untouchables" and assigned the most dangerous and degrading work. According to a Pew Research Center survey of nearly 30,000 Indian adults, 98 percent of Indians identify with a caste, regardless of their religion - including Christians, Muslims, and Sikhs.
That is not ancient history. It is today's labor market.
The Scale of the Problem
India's NITI Aayog published its Multidimensional Poverty Index in 2023. It found that 27.4 percent of Scheduled Castes and 32.6 percent of Scheduled Tribes are multidimensionally poor. The national average is 15 percent. The general category - upper castes - sits at 8.5 percent.
That gap is not explained by education alone. An audit study published by the Institute of Labor Economics (IZA) found that low-caste applicants need to send 20 percent more resumes than high-caste applicants just to get the same callback. A study by the Indian Institute of Dalit Studies found that the odds of a Dalit applicant being invited for an interview were about two-thirds of the odds of an equally qualified upper-caste Hindu applicant.
That means the discrimination happens at the name.
The National Sample Survey found that Scheduled Caste workers are employed for fewer days each year - an average annual gap of about 28 days. According to Economic and Political Weekly research, caste discrimination causes roughly 15 percent lower wages for Scheduled Caste and Scheduled Tribe workers compared to equally qualified peers. The discrimination is worse in the private sector than in the public sector.
The private sector is where India's growth is happening. Reservations only cover government jobs and public universities. They do not cover private firms, which now employ the majority of India's urban workforce.
The Business Cost Nobody Is Counting
Economist Sampreet Goraya's research, published in the Journal of Monetary Economics, documents how caste distorts business in India. His key finding: 75 percent of employees within a firm belong to the same caste as the owner. Owners hire people they know. They know people from their caste.
A firm owner from a lower caste starts at a disadvantage - not because their product is worse, but because credit is harder to get, suppliers are less willing to deal, and the customer base is caste-segmented. A separate paper found that Dalit exclusion from entrepreneurship reduces India's per capita GDP by an estimated 1.5 to 2 percent.
India's GDP is roughly $3.7 trillion. A 1.5 percent drag equals over $55 billion per year in output that does not exist - because of who owns what firm, who gets credit, and who gets hired.
Lower-caste groups own less than 15 percent of Indian firms, despite making up nearly 30 percent of the population. According to the All India Survey on Higher Education, Scheduled Castes make up only 14.3 percent of students in higher education institutions - and Scheduled Tribes just 5.8 percent - despite together accounting for over 25 percent of India's population.
Manual Scavenging: The Most Visible Failure
Manual scavenging means cleaning human waste by hand - from latrines, sewers, and septic tanks - without machinery and often without safety equipment. According to government data, 97 percent of manual scavengers in India are Dalits. The job is assigned by caste, not by choice.
India banned manual scavenging in 1993. It banned it again, with stricter rules, in 2013. The Safai Karamchari Andolan estimates that more than 770,000 people are still doing this work. The National Human Rights Commission said in 2021 that state government claims of "zero manual scavengers" are far from the truth.
People are dying in these sewers. The National Commission for Safai Karamcharis recorded 928 sewer deaths between 1993 and 2020. Research published in a peer-reviewed public health journal found that manual scavengers have a mean life expectancy of 58 years, compared to 68.3 years for the general population. That is a decade of life lost to a job assigned by birth.
Two laws. Thirty years. The practice continues.
What Has Already Been Tried
Reservations (Quotas) - From 1950 onward. Dr. B.R. Ambedkar, born into the Mahar caste and the architect of the Indian Constitution, secured the reservation system at independence. The constitution set aside seats in government jobs, public universities, and legislatures for Scheduled Castes and Scheduled Tribes. Today, 15 percent of government jobs and university seats are reserved for Scheduled Castes, 7.5 percent for Scheduled Tribes, and 27 percent for Other Backward Classes. The Supreme Court capped total reservations at 50 percent in 1992.
The results are mixed. According to research in the Journal of Economic Literature by economist Kaivan Munshi, there has been measurable convergence across caste groups in education, occupations, and income since independence. But the system does not cover the private sector - where most of India's job growth now happens. And benefits have been unevenly distributed: better-off sub-groups within OBCs often capture most of the reserved seats, while the poorest communities miss out.
The Mandal Commission - 1980, implemented 1990. Prime Minister V.P. Singh's government implemented a recommendation from the Mandal Commission to extend 27 percent reservations to OBCs. This triggered mass protests. The Supreme Court cleared the policy in 1992 but imposed the 50 percent cap. The commission based its recommendations on 1931 census data - the last time India counted all castes. That data has never been updated.
The Prevention of Atrocities Act - 1989, amended 2015 and 2018. This law criminalized violence and discrimination against Scheduled Castes and Scheduled Tribes and set up special courts to hear cases faster. The National Crime Records Bureau shows that 96 percent of caste atrocity cases against Scheduled Castes were still pending trial at the end of 2021. The conviction rate sits at 36 percent for Scheduled Caste cases and 28.1 percent for Scheduled Tribe cases. Crimes registered under the act rose 13.1 percent between 2021 and 2022, reaching 57,582 cases. The law exists. Enforcement does not.
The Manual Scavenging Bans - 1993 and 2013. The government auditor found in 2003 that the 1993 Act had "failed to achieve its objectives even after 10 years of implementation." The 2013 law added stricter penalties and rehabilitation support. The National Crime Records Bureau stopped publishing data on cases filed under the 2013 law after 2016. Enforcement effectively went dark.
The Socio-Economic and Caste Census (SECC) - 2011. The government conducted a national caste survey in 2011 at a cost of roughly Rs 4,900 crore. The caste data was never released. India is now making policy for hundreds of millions of people using caste population estimates from a 1931 British colonial census.
How Other Countries Fixed This
Japan's Burakumin Program - The 33-Year Infrastructure Push
Japan has its own descent-based excluded group: the Burakumin. They are not ethnically different from other Japanese. They are descended from people who did "impure" work in the feudal era - butchers, leather workers, executioners. Like Dalits, they were assigned their status by birth and confined to specific neighborhoods and occupations.
In 1969, Japan passed the Law on Special Measures for Dowa Projects. The law funded a 33-year program of targeted investment in Burakumin communities - repairing housing stock, building roads and sewage infrastructure, improving schools, and improving teacher-student ratios in Burakumin districts. The program also required private companies with more than 100 employees to run workplace training to combat discrimination.
The result, documented in the Asia-Pacific Journal, was measurable improvement in the physical living conditions of Burakumin communities. School attendance and educational attainment rose significantly. The infrastructure gap closed substantially over 30 years.
Discrimination in marriage, employment, and housing continued even after the law ended in 2002. Japan shows what targeted, long-term, funded investment in excluded communities can do for their physical and educational conditions - and where it falls short when attitude change is not also measured and required.
India's reservation system focuses on quotas at the top. Japan's Dowa program focused on building the foundation at the bottom. Both are needed. India has mostly done one of them.
Who Is Accountable
The Ministry of Social Justice and Empowerment is responsible for Dalit welfare, manual scavenging rehabilitation, and implementation of the Prevention of Atrocities Act. The National Commission for Safai Karamcharis - a statutory body under this ministry - monitors sewer deaths and sanitation worker conditions. It has repeatedly documented gaps between what states report and what is actually happening. The Ministry of Home Affairs oversees the National Crime Records Bureau, which stopped publishing enforcement data on the manual scavenging ban after 2016.
State governments hold primary responsibility for law enforcement. The five states with the highest caste atrocity cases - Uttar Pradesh, Rajasthan, Madhya Pradesh, Bihar, and Odisha - together account for over 70 percent of all cases. Nobody in these state home ministries is being fired for it.
The SECC 2011 caste data is held by the Registrar General and Census Commissioner of India and has not been released. The government has announced that the next census will include caste data. But the 2011 data - already collected and paid for - should be released and audited publicly.
What Would It Cost
A 1.5 to 2 percent GDP drag from Dalit exclusion alone equals roughly $55 to $74 billion per year at India's current economic size. That is money that does not get invested, businesses that do not get started, and workers who never reach their productive potential.
The infrastructure to deliver change is there: post offices, banks, Aadhaar identity cards, and direct benefit transfer systems already reach most of India. What is missing is measurement and enforcement.
The SECC 2011 cost Rs 4,900 crore to conduct. Releasing and acting on that data costs nothing extra. It has already been collected.
What Needs to Happen
First: Release the caste data and keep collecting it. The SECC 2011 data should be released with its methodology clearly disclosed. Every future five-year economic survey should include caste-disaggregated income, employment, and education data. Without this, every policy is a guess.
Second: Extend anti-discrimination enforcement to private employers. Requiring large private employers to report the caste composition of their workforces - as the US requires race data reporting - would make the problem visible and create accountability. South Africa, Brazil, and the US all collect demographic employment data. India currently does not require private firms to report this at all.
Third: Enforce what is already law. India does not need new laws on manual scavenging. Mechanized cleaning robots exist and have been deployed in some cities. Every municipal body above a certain size should be required to procure this equipment within a fixed timeline, with the Ministry of Social Justice tracking compliance publicly. The National Crime Records Bureau should resume publishing enforcement data under the 2013 manual scavenging prohibition law.
Fourth: Tie community investment to outcomes. Outcome-linked funding - where a portion of each district's development allocation is contingent on closing documented gaps in Scheduled Caste school enrollment, graduation rates, and access to formal credit - would create the right incentives for state governments.
Fifth: Fix credit access for lower-caste entrepreneurs. Goraya's research shows that removing financial constraints for lower-caste business owners would expand productive firms, raise wages, and create jobs. Public sector banks should be required to report caste-disaggregated loan approval rates. Banks with documented gaps should face corrective review from the Reserve Bank of India.
None of these recommendations require dismantling the reservation system. They require adding what the reservation system cannot reach: the private sector, the data layer, and the enforcement that turns law into practice.