Abolition on Paper Why Sewer Deaths in India Keep Rising
India’s official story on manual scavenging has a strange split. Ask how many people died while doing manual scavenging, and Parliament has been told the number is zero. Ask how many sanitation workers died while cleaning sewers and septic tanks, and the number changes sharply.
In August 2026, the Social Justice Ministry told the Lok Sabha that 498 sanitation workers died while cleaning sewers and septic tanks between January 2019 and June 2026. The difference is not a clerical error. It is a legal and administrative boundary.
The Manual Scavengers Act 2013 defines manual scavenging around the handling of human excreta from insanitary latrines, open drains, and similar sites. Sewer and septic tank deaths are counted separately as deaths during hazardous cleaning. On paper, this distinction allows the government to say that manual scavenging deaths are nil, while still reporting hundreds of deaths in sanitation work.
That distinction may make sense inside a file. It is harder to defend inside a manhole.
The public record is fragmented across labels such as “manual scavenging India”, “sewer deaths India 2026”, “NAMASTE scheme”, “sanitation workers deaths”, and “Manual Scavengers Act 2013”. They look like separate categories. On the ground, they often describe the same chain of vulnerability: caste-bound labour, informal contracting, poor enforcement, weak mechanisation, and families left to fight for compensation after preventable deaths.
This article is policy analysis, not legal advice.
The number changes when the definition changes
The first question is not only how many people died. It is who gets counted, and under which legal category.
The 2013 law, formally the Prohibition of Employment as Manual Scavengers and their Rehabilitation Act, sought to prohibit manual scavenging and rehabilitate workers who had been pushed into it. Its core definition focused on the manual handling of human excreta from insanitary latrines, open drains, railway tracks, and pits before the waste fully decomposes.
Sewer and septic tank cleaning sit in a related but distinct space. The law also prohibits hazardous cleaning of sewers and septic tanks without protective gear and safety precautions. Yet the official reporting often separates “manual scavenging” from “hazardous cleaning”.
That separation produces two realities:
Question asked | Official answer described in Parliament | What it leaves unresolved |
Deaths due to manual scavenging | None reported | The definition excludes many sewer and septic tank deaths |
Deaths while cleaning sewers and septic tanks | 498 between January 2019 and June 2026 | These deaths still arise from caste-linked, unsafe sanitation work |
This is why the debate cannot stop at the word “abolished”. If manual scavenging is defined narrowly enough, abolition can look successful while workers continue to die doing adjacent forms of the same degrading labour.
The government’s classification has legal consequences. If a death is treated as “hazardous cleaning” rather than manual scavenging, it moves through a different administrative channel. It may still attract compensation. It may still trigger an inquiry. But it avoids the direct admission that a prohibited practice survived.
That matters because law is not only a punishment tool. It is a truth-telling instrument. A category that hides a pattern weakens public accountability.
The official count is smaller than civil society counts
The state is not the only source of data. Civil society organisations, especially those working directly with sanitation workers and bereaved families, often report higher numbers.
Official records show 43 sewer and septic tank deaths in 2025. The Safai Karamchari Andolan, a movement against manual scavenging, counted 121 deaths in 2025, and 101 more by early July 2026 across 16 states.
The gap is large. It could reflect differences in method, timing, state reporting, verification, or classification. It could also reflect under-reporting by local authorities, especially where contract chains are unclear and families face pressure to settle.
The precise number matters. But even the lower number should be treated as a policy failure. Forty-three deaths in a year from entering sewers and septic tanks is not a residual problem. It is a recurring failure of urban governance.
Three points are clear from the data available:
The deaths are not isolated accidents
They repeat across states, cities, and years.
The victims are usually workers with little bargaining power
Many are Dalit workers, often hired through contractors or informal intermediaries.
The work continues despite legal safeguards
The law bars hazardous cleaning without protection, yet workers still enter confined spaces that can contain toxic gases and low oxygen.
A sewer death is often described as an accident. That word can hide the chain of decisions that came before it. Someone issued the work order. Someone failed to send a machine. Someone failed to test the gas. Someone failed to supply protective equipment. Someone allowed a contractor to deploy workers without training.
When those decisions repeat, the result is not accidental in any meaningful policy sense.
Caste and contracting keep the risk hidden
The social profile of the workforce matters because this is not a random occupational hazard.
Under the NAMASTE scheme, launched in 2023, the workers profiled so far show a strong caste pattern. Sixty-seven per cent of the sewer and septic tank workers profiled belonged to Scheduled Castes, and about 90,000 workers have been validated for support.
Those figures confirm what sanitation workers’ movements have argued for decades. Sewer and septic tank cleaning is not merely dirty work that cities have failed to modernise. It is labour assigned through caste, normalised through poverty, and outsourced through contractors.
Contracting changes the visibility of responsibility. A municipal body may say it did not directly employ the worker. A contractor may say the worker ignored instructions. A subcontractor may disappear after the incident. Police records may name lower-level actors while leaving senior officials untouched.
This structure matters because many sanitation workers do not enter sewers as regular public employees with stable benefits, training, unions, and protective systems. They enter through casual arrangements, daily wage work, emergency calls, or contractor-led teams.
That creates predictable risks:
Workers may not know the legal rights linked to hazardous cleaning.
They may not receive safety gear, gas testing, harnesses, oxygen support, or emergency backup.
They may hesitate to refuse unsafe work because refusal can mean losing income.
Their families may lack the resources to pursue criminal cases.
Local authorities may frame the death as a contractor’s fault while continuing the same contracting model.
The caste dimension sharpens the injustice. If the same communities are repeatedly pushed into lethal work, the policy problem is not only lack of equipment. It is the persistence of a labour system that treats some lives as cheaper than others.
Urban India depends on sanitation work every day. Drains, sewers, septic tanks, pumping stations, and treatment systems form the hidden infrastructure of public health. Yet the workers who keep that system running often remain invisible until death makes them briefly visible.
Mechanisation is growing but does not yet control the sewer
Mechanisation is the most practical route to ending hazardous entry. If machines can clean sewers and septic tanks, workers should not have to enter them except in tightly regulated emergency conditions.
India has taken steps in that direction. The NAMASTE scheme aims to identify sanitation workers, support their safety, provide training, link them to assistance, and promote mechanised cleaning. The validation of about 90,000 workers is significant because invisible work cannot be regulated well.
Yet identification is only the start. The central question is whether mechanisation reaches the point of work.
A city can claim it has machines, but the machine may be too few in number, unavailable during emergencies, poorly maintained, too large for narrow lanes, or operated only by select teams. Smaller towns may rely on manual methods long after larger cities acquire equipment. Septic tank cleaning in peri-urban and rural areas can remain outside regular municipal systems.
There is also a difference between buying machines and changing work practice. A municipality can purchase suction-cum-jetting machines, robotic devices, safety kits, and gas detectors. If contracts still reward speed over safety, workers may still be sent down. If ward-level staff are not trained, machines may sit unused. If households and local businesses call private operators outside municipal systems, the same unsafe work returns through another door.
Mechanisation has to answer four operational questions:
Policy question | Why it matters |
Who owns the machines? | Ownership affects availability, maintenance, and accountability |
Who operates them? | Operators need training, wages, and protection from retaliation |
Who pays for use? | If machine use is costly or slow, contractors may cut corners |
Who records each job? | Work logs can show whether manual entry was avoided or hidden |
The technology debate can also become too narrow. Machines reduce the need for human entry, but they do not replace the need for enforcement. Workers still need oxygen monitoring, rescue protocols, supervision, insurance, medical checks, and a clear right to refuse unsafe work.
The most important measure of mechanisation is not the number of machines purchased. It is the number of sewer and septic tank entries prevented.
Compensation is improving while punishment remains rare
Compensation has improved in important cases. The Calcutta High Court ordered ₹30 lakh for each of four workers who died in a 2021 Kolkata sewer incident. Official data also says most families receive the ₹10 lakh minimum compensation.
These developments matter. Compensation can reduce immediate hardship for a worker’s family. It can help pay debts, support children’s education, cover medical costs, and replace some lost income. Courts have played a role in pushing authorities to respond.
But compensation is not justice by itself.
A payment made after death does not prevent the next worker from entering a toxic sewer. It does not identify who gave the order. It does not fix the contract. It does not make a municipal engineer, contractor, or supervisor fear real consequences for breaking safety rules.
The accountability record remains extremely weak. The Centre told a parliamentary committee that out of 616 cases against contractors for failing to provide safety gear, only one ended in conviction.
That figure is perhaps the clearest evidence of why deaths continue. If hundreds of cases produce almost no convictions, then the legal deterrent is minimal. Contractors and officials can treat compensation as a cost that follows death, rather than punishment as a consequence that follows criminal negligence.
Reports by Scroll have also found that many cases end in out-of-court compromises, with families pressed to settle. This is not hard to understand. A bereaved family may need money quickly. Police cases can take years. Contractors and local power brokers may pressure families. Witnesses may be workers who fear losing future employment. Legal aid may be weak or absent.
The result is a predictable pattern:
A worker dies during sewer or septic tank cleaning.
Local attention rises for a few days.
Compensation is announced or contested.
A case may be filed against a contractor or lower-level person.
The family faces pressure to settle.
The larger system continues.
This pattern turns accountability into paperwork. It allows the state to say it responded while avoiding the harder question: who should lose a licence, a contract, a job, or liberty for sending someone into a manhole without protection?
What abolition has to mean below the street
Abolition cannot mean only that the state no longer recognises manual scavenging as officially existing. It has to mean that no worker is forced by caste, poverty, contract pressure, or municipal negligence to handle human waste or enter a deadly confined space.
That calls for a wider test of success.
The narrow test asks whether manual scavengers are officially identified. The wider test asks whether the sanitation system still depends on manual entry, informal labour, and weak punishment.
A serious abolition framework would track:
Every sewer and septic tank cleaning operation
Cities should know who did the work, which equipment was used, whether gas testing happened, and whether manual entry took place.
Every contractor and subcontractor
Public bodies should not be able to hide behind outsourced arrangements. Contractor licences should depend on safety compliance.
Every death and injury
Fatality data must be public, timely, and comparable across states. Injuries and near-fatal incidents also need tracking.
Every compensation claim
Families should not have to negotiate for a legal entitlement after a preventable death.
Every prosecution
Cases should be followed from filing to conviction, acquittal, settlement, or closure.
Every machine deployed
Procurement data should be linked to actual usage, not only purchase.
This is not only a welfare issue. It is an urban governance issue. Sewer deaths reveal whether cities can manage the most basic promises of public infrastructure: sanitation, health, safety, and dignity.
The administrative solution must also include workers’ voices. Profiling workers under schemes like NAMASTE can help, but only if workers can report unsafe practices without losing work. Worker cooperatives, unions, and movements need a formal role in monitoring contracts and safety compliance. Families of deceased workers need legal support, not only compensation forms.
A credible enforcement system would make three changes visible.
First, no manual entry should occur without written authorisation, recorded reasons, safety equipment, gas testing, and rescue arrangements. If a city cannot produce the record, the presumption should favour the worker.
Second, contractors who violate safety rules should face swift blacklisting and criminal prosecution. A contractor who sends a worker into a manhole without protection should not be eligible for future public work while the case disappears into delay.
Third, public officials should also face consequences. Outsourcing should not become an accountability shield. If a municipal body awards, supervises, or benefits from the work, it must carry responsibility for the conditions under which the work happens.
India has enough law to name the wrong. It needs enough enforcement to stop it.
The legal fiction ends only when punishment becomes real
The central contradiction is now plain. India can report zero deaths from manual scavenging because of how the term is defined. It can also report 498 deaths from sewer and septic tank cleaning between January 2019 and June 2026. The law can separate the categories. The families cannot.
The data points in one direction. Deaths continue. Civil society counts are higher than official counts. Scheduled Castes form a large share of the identified sewer and septic tank workforce. Mechanisation is advancing, but not fast enough to end manual entry. Compensation has become more visible, yet convictions remain almost absent.
Abolition on paper is not abolition in practice. A city has not ended manual scavenging if it still sends Dalit workers into sewers through contractors. A ministry has not solved the problem if deaths move from one column to another. A compensation cheque cannot stand in for a conviction.
The test is simple. When a worker dies in a sewer, the system should not ask only how much the family will receive. It should ask who ordered the work, why a machine was not used, where the safety gear was, which official supervised the contract, and why the responsible person should not go to prison.
Until that happens, the sewer will keep exposing what the paperwork conceals. The job may be abolished in law, but the manhole still knows who is being sent down.
