India's Undertrial Crisis Why Bail Fails the Poor
- BerryBeat Team
- 2 days ago
- 11 min read
A jail system built for 4.53 lakh people held 5.11 lakh people at the end of 2024. That is the headline. The deeper wound is this: 73 percent of those inside Indian prisons had not been convicted of any crime.
They were undertrials. Accused, not proven guilty. Waiting for investigation, charges, hearings, witnesses, paperwork, lawyers, sureties, and trial dates. Waiting, often, because freedom had a price they could not pay.
The latest NCRB prison statistics do more than confirm Indian prison overcrowding. They expose a quieter failure at the heart of criminal justice: bail exists in law, but poverty decides who can use it. For thousands of undertrial prisoners in India, a bail order is not the same as release. It is only a piece of paper unless someone can produce a bond, a surety, an address proof, a guarantor, or the money to keep the process moving.
This article is for public information and discussion. It is not legal advice.
The presumption of innocence is collapsing inside overcrowded prisons
The Constitution does not treat jail as the default place for an accused person. The criminal process is meant to punish after conviction, not before trial. Bail is supposed to protect that principle.
Yet the numbers tell a harsher story.
At the end of 2024, India’s prisons were at 112.7 percent occupancy. Delhi’s jails were running at nearly double capacity. This is not just a question of cramped wards and overburdened staff, though those are serious problems. Overcrowding also slows everything that might help a person leave.
When prisons hold more people than they were built for, the legal system becomes harder to access. Prison interviews take longer. Medical production and court production suffer. Files get delayed. Lawyers struggle to meet clients. Families spend more money travelling and waiting. Every delay punishes the person who has the least power to absorb it.
The ethical shock is not only that prisons are crowded. It is that they are crowded mostly with people whom the State has not proved guilty.
An undertrial may eventually be acquitted. The prosecution may fail to produce evidence. Witnesses may turn hostile. Charges may be dropped. The case may be too weak to survive scrutiny. But the years spent inside do not return. For a daily wage worker, a young student, a migrant labourer, a single earning member, or a person already pushed to the edge by caste and poverty, pretrial detention can destroy a life before the trial even begins.
The NCRB data also shows who carries this burden. Around two-thirds of undertrials come from Scheduled Caste, Scheduled Tribe, or OBC communities. Sixty-three percent never studied beyond Class 10. Nearly half are under 30.
This is not an accident of criminality. It is a map of social power.
People with savings, family assets, stable addresses, and socially accepted guarantors can usually fight a criminal case from outside jail. People without those assets fight the same system from inside a cell. The law may use neutral words, but its daily operation is not neutral.
The phrase “innocent until proven guilty” sounds clean in a courtroom. In a prison register, it becomes conditional. Innocent, unless too poor to leave. Innocent, unless no one will stand surety. Innocent, unless the next hearing is months away.
A country cannot keep claiming fidelity to liberty while treating undertrial detention as routine. Overcrowding is not merely poor prison management. It is evidence that the front door of the criminal justice system is swallowing people faster than the courts can test the accusations against them.
Bail was meant to protect liberty but poverty turns it into punishment
Bail is often described as a balance. The court must ensure that the accused appears for trial, does not threaten witnesses, and does not obstruct justice. These are legitimate concerns. A fair criminal process needs safeguards.
The problem begins when those safeguards become financial filters.
For someone with money, bail conditions may be inconvenient. For someone without money, they can be impossible. A surety may need documents. A bond may require financial backing. The accused may need a local person with social standing to appear in court and accept responsibility. Families may have to pay for copies, travel, affidavits, lawyer visits, and repeated follow-up.
That is how bail poverty in India works. It does not always announce itself through a large cash amount. Sometimes it appears as a missing ration card, an employer who refuses to help, a family unable to travel from another district, or a landlord unwilling to vouch for a tenant.
A bail order without the means to satisfy bail conditions is freedom in theory and detention in practice.
In one Supreme Court filing, around 5,000 undertrials remained in jail even after courts had granted them bail because they could not furnish a surety or bond. That figure should have forced a national reckoning. It showed that the issue was not only judicial reluctance to grant bail. The issue was also the system’s failure to convert bail into release.
A person may hear the words “bail granted” and still remain in custody. Days become weeks. Weeks become months. Families borrow money. Some sell tools, jewellery, livestock, or small plots of land. Others simply fail.
The burden falls hardest on people outside formal systems. Migrant workers often lack local sureties. Homeless or semi-housed people struggle to prove residence. Young people from poor families may have no one with the documents a court expects. Members of marginalised caste communities may face the added barrier of social exclusion, where even finding a person willing to stand for them becomes difficult.
This is why bail cannot be understood only as a courtroom event. It is also a question of social capital.
A wealthy accused person can hire lawyers, prepare documents, arrange sureties, and challenge harsh conditions quickly. A poor accused person may not even understand the order passed in court. If legal aid is missing or irregular, the family becomes the clerk, messenger, and financer of release. Many do not know where to go. Many give up.
The result is a two-tier system.
One India experiences arrest as a legal crisis that can be managed. Another experiences arrest as a social and economic collapse. The first can return home pending trial. The second waits in prison and watches the case shape every part of family life.
Children leave school because an earning parent is gone. Families lose rented rooms. Employers move on. Debt grows. Mental health suffers. The accused person becomes more isolated with every missed court date and every failed attempt to arrange surety.
This is punishment before proof. It is not called a sentence, but it behaves like one.
India has known the solution for decades
The undertrial crisis is not new. India has recognised it again and again, in judgments, statutes, committees, circulars, and schemes. The tragedy is not that the law has no answer. The tragedy is that answers exist and still do not reach the people they were written for.
In 1979, the Supreme Court’s decision in Hussainara Khatoon forced the country to confront the horror of people languishing in jail for periods longer than the maximum punishment for their alleged offence. The Court linked speedy trial to the right to life and personal liberty. It ordered release in cases where detention had become legally indefensible.
That judgment remains one of the defining moments in Indian legal history. It gave constitutional language to a simple moral claim: the State cannot accuse a person and then forget them in jail.
Parliament later introduced Section 436A of the Code of Criminal Procedure in 2005. It said that an undertrial who had spent half of the maximum possible sentence in custody should be released on personal bond, with or without sureties, except in the gravest categories of offences. The 2023 criminal code rewrite retained this principle and added a lower one-third threshold for first-time offenders.
On paper, this should have reduced needless detention. In practice, many eligible prisoners remain inside because the system does not identify, process, and release them in time.
The failure has many layers.
Prison records may not be updated properly. Courts may not receive timely lists. Legal aid lawyers may be overburdened. Families may not know that the accused has crossed the statutory custody threshold. Prisoners may not understand that they can seek release. Applications may not be filed. Orders may not be implemented. Surety conditions may still block release.
A right that depends on the prisoner discovering it is a weak right. A safeguard that works only when someone outside pushes for it is not a safeguard for the poor.
The same gap appears in the government scheme announced in 2023 to support poor prisoners who cannot afford bail or fines. The idea is sound. If poverty keeps a person in jail after bail, public funds should help remove that barrier. But the scheme has reportedly gone largely unused by the states responsible for running it.
That pattern is familiar. India often creates legal welfare measures with good language and weak delivery. The scheme exists. Committees exist. Forms exist. The money may even be available. Yet the person in the prison ward remains unaware, unassisted, or unable to pass through the administrative maze.
This is where reform loses its force. A right written in Delhi must survive the police station, the district court, the prison office, the legal aid clinic, and the family’s economic reality. If it fails at any point, the person stays inside.
There is also a deeper cultural problem. Bail is still too often treated as an exception rather than the normal rule in cases where custody is not necessary. Courts worry, often rightly, about absconding, intimidation, and public safety. But those concerns should be addressed through specific reasons, not broad suspicion.
The Supreme Court has repeatedly said that bail should not be mechanical and detention should not be endless. Yet lower courts operate under heavy pressure, large caseloads, police narratives, public anger in serious cases, and fear of being blamed if something goes wrong. In such a climate, refusing bail can feel administratively safer than granting it.
That caution has a human cost. When courts overuse custody, prisons become holding zones for unresolved cases. When prisons fill with undertrials, trials move slower. When trials move slower, more people become eligible for statutory release. When statutory release is not implemented, the prison population keeps growing. The system creates the very crisis it later claims it lacks capacity to solve.
Wrongful prosecution remains a private catastrophe
A person who spends years in jail and is later acquitted does not automatically receive restoration. The legal system may declare them not guilty, but it rarely repairs what custody took.
India still has no dedicated law for compensation after wrongful prosecution. Courts can award compensation in some cases, including through constitutional remedies, but the system is thin, discretionary, and uneven. There is no clear national statute that defines wrongful prosecution, sets standards, creates a process, and ensures timely payment.
That absence matters.
Without a dedicated law, wrongful incarceration compensation in India depends too much on the facts of the case, the court approached, the quality of legal representation, and judicial discretion. Many people who deserve compensation never file claims. Many do not know they can. Many lack the money to continue litigation after acquittal. Many simply return home and try to rebuild quietly because survival leaves no space for another legal battle.
The harm is not abstract.
Years in custody can mean lost wages, broken education, untreated illness, damaged family ties, stigma, and debt. A young person acquitted after four years does not receive those years back. A parent released after trial cannot rewind childhood milestones missed from a prison barrack. A worker who loses a job because of arrest may find that acquittal does not erase suspicion.
The State has enormous power to arrest, prosecute, and confine. When it uses that power wrongly, or when its case collapses after years of delay, accountability cannot be left to sympathy. A rights-based democracy needs a statutory route to repair.
Compensation will not solve the undertrial crisis by itself. But it would change incentives. If the State must pay for wrongful or excessive incarceration, delay becomes costly. Poor investigation becomes costly. Casual prosecution becomes costly. Detention without urgency becomes costly.
At present, the cost is carried mainly by the accused person and their family. That is why the system can continue.
The missing reform is enforcement with a human face
India does not need to choose between public safety and liberty. A serious justice system can protect witnesses, ensure appearance in court, and avoid unnecessary detention. The question is whether it has the courage to treat poor accused persons as rights-bearing citizens rather than administrative burdens.
The reforms already known to the system must become ordinary practice.
Personal bonds should become the default in appropriate cases. Courts should use surety and cash conditions only when they are truly needed. If the accused has roots in the community, no record of absconding, and faces a less serious charge, poverty should not decide release.
Bail orders should be tracked until release. A court’s work should not end when it grants bail. District systems should record whether the person actually left prison. If someone remains inside because of surety or bond failure, the case should come back automatically for modification.
Custody thresholds should trigger automatic review. No undertrial should have to discover Section 436A or its newer equivalent through luck. Prisons and courts should run regular audits for people who have crossed the half-sentence mark, and for first-time offenders who cross the one-third mark under the new framework.
Legal aid must move inside prisons with real capacity. Legal aid cannot be a name on a board. Undertrials need regular counselling, case status updates, help with bail modification, family contact, and support in gathering documents. This requires trained lawyers, paralegal workers, interpreters, and accountability for missed visits.
The 2023 support scheme should publish state-wise use. A scheme for poor prisoners cannot remain invisible. States should publish how many people received assistance, how much money was used, how long release took, and why applications failed. Public money for liberty must be publicly tracked.
Compensation needs a statute. A dedicated law should define wrongful prosecution and wrongful incarceration, create a simple claims process, set timelines, and include rehabilitation support where needed. Acquittal after years in custody should not leave a person with only a certified copy of judgment and a ruined life.
None of this is utopian. These are practical steps. They require money, staff, data, and discipline. They also require a shift in imagination.
The poor do not need charity from the criminal justice system. They need the system to obey its own law.
When a person is eligible for release, release should happen. When bail is granted, prison should not continue because a surety is missing. When detention exceeds statutory limits, review should not depend on family pressure. When prosecution fails after years, compensation should not depend on extraordinary luck.
The undertrial crisis is often described as a prison problem. That lets too many institutions escape scrutiny. Police arrest practices, prosecutorial decisions, court delays, bail conditions, legal aid failure, prison administration, and state budgets all feed the same machine.
A solution must be just as connected.
Every district should know how many undertrials are inside, how long they have been there, how many have bail but remain unreleased, how many are eligible for statutory release, and how many lack legal representation. These numbers should not sit in annual reports alone. They should guide weekly decisions.
Liberty should have a dashboard, but more than that, it should have owners. A named judge, a prison officer, a legal services authority, and a state official should all be responsible for ensuring that no one remains in jail only because they are poor.
That is the heart of the matter. India has built many doors in law. Bail. Personal bonds. Statutory release. Legal aid. Welfare schemes. Constitutional remedies. But a door that the poor cannot open is not a door. It is decoration.
The measure of reform is not the number of provisions passed. It is the number of people who sleep at home because those provisions worked.
India’s undertrial crisis asks a direct question of the republic: when the State has not proved guilt, how long may it take a person’s freedom merely because they lack money, documents, or standing?
The answer should not require another landmark case, another committee, or another prison report. It should be visible in district courts, prison registers, legal aid rooms, and homes where families no longer have to sell their future to secure a freedom the law already promised.
