This page is informational and is not legal advice. It describes rules, orders and rulings as published or reported on the dates shown; read the primary text before relying on any of it.
Doorstep cash collection has one structural weakness, and it is not the honesty of the person doing the collecting. It is the gap between the moment cash leaves your hand and the moment it appears in the institution's ledger. Every failure mode described below lives inside that gap.
Pigmy collection has run in India since 1928 and serves millions of depositors who have no realistic alternative. Nothing here should be read as a claim that the product is generally fraudulent. What follows is the opposite of that claim: what can go wrong, the cases we could verify, and what each side can do about it.
The four ways a collection goes wrong
| Failure mode | What the depositor sees | What the institution's record shows |
|---|---|---|
| Under-recording | A passbook or receipt entry smaller than the cash handed over, or no entry at all | The smaller figure, or nothing |
| A receipt that never reaches the ledger | A correct, valid-looking receipt in hand | No credit posted against the account |
| Delayed crediting | The credit does appear, later than it should | A value date after the date of collection |
| Absconding with collections | The agent stops visiting; recent entries never post | A shortfall between the route's collections and what was remitted |
Only the last is theft as a depositor would recognise it on the day. The first three look identical to a normal collection, and stay invisible until somebody compares two records held in two different places.
Why the gap exists at all
In a paper round the branch knows nothing of a collection until the agent returns at the end of the day and remits. Until then the only evidence that money changed hands is the receipt in the depositor's hand and the carbon copy in the agent's book. Passbook and ledger can therefore disagree — legitimately and routinely — for a day or longer, and neither side has any automatic way to notice when the disagreement is not legitimate.
RBI's Branch Authorisation Directions of 4 December 2025 narrow that window rather than closing it, by requiring a co-operative bank's doorstep scheme to credit collected cash the same day or the next working day. That is a compliance boundary, not a technical impossibility. The window still has to be watched.
What the record actually shows
Reporting on co-operative-sector fraud is plentiful; reporting that isolates doorstep collection is not. These are the cases we could load and read.
| Case | Reported | What was alleged | Level |
|---|---|---|---|
| A credit co-operative society at Indora, Nagpur | 30 May 2025 | ₹4,54,68,237 of embezzlement identified in a five-year special audit covering April 2018 to March 2023; 21 office bearers and directors booked, including for issuing illegal commissions to agents and unauthorised payments to recovery officers | Board and management |
| A women's credit co-operative society at Vaijapur, Chhatrapati Sambhajinagar | 26 June 2025 | ₹5.83 crore of pigmy and fixed deposits allegedly not returned after a branch shut abruptly; case against the chairperson, vice-chairperson and 13 board members. The complaint was filed by one of the society's own pigmy agents | Board |
| New India Co-operative Bank, Mumbai | February to March 2025 | ₹122 crore alleged to have been taken from the bank's own safes at two offices; the general manager and head of accounts, a former chief executive and others arrested. Detected during an RBI inspection | Senior management |
Two things follow, and both cut against the easy story.
The New India Co-operative Bank case has nothing to do with doorstep collection. The money is alleged to have been removed from the bank's own strongrooms by the people who ran it, and it surfaced in a regulatory inspection, not through anything a depositor or agent could see. It belongs in the opposite column from agent risk: no passbook discipline would have caught it, and no collection device would have prevented it. We include it because it is the case the sector is most often asked about, and because using it to make a point about agents would be wrong.
In the Vaijapur case the pigmy agent was the complainant. Agents sit closer to the depositor than anyone at the branch does, and they are often the first to find out that a society has stopped honouring its obligations.
The Nagpur case is the one that touches agent economics directly, and even there precision matters: the report describes illegal commissions paid to agents, not specifically to pigmy agents, and the accused are the society's own office bearers and directors. A separate RTI reply recorded criminal cases against 62 individuals across 19 credit co-operative societies in that district between April 2022 and May 2026, and names no agents either.
The sector numbers, and their limits
| Year | Frauds reported by urban co-operative banks | Amount |
|---|---|---|
| 2014-15 | 478 | ₹19.8 crore |
| 2015-16 | 187 | ₹17.3 crore |
| 2016-17 | 27 | ₹9.3 crore |
| 2017-18 | 99 | ₹46.9 crore |
| 2018-19 | 181 | ₹127.7 crore |
| Total | 972 | ₹221 crore |
Those figures come from an RBI reply to an RTI application, reported by PTI in January 2020. They cover every kind of fraud at an urban co-operative bank — loans, cheques, staff defalcation and doorstep collection together — over five years ending March 2019, and predate the current supervisory framework. They are not a measure of agent fraud, and we found no published figure, at any date, that isolates doorstep or pigmy collection losses.
What already protects a depositor
The doorstep banking rules, if your institution is a co-operative bank. The receipt, the same-day or next-working-day credit, the advice stating the date of credit, and the bank's own cash limits are each a control on the blind window. They are set out in full, along with which institutions they bind and which they do not, in our guide to the rules behind pigmy collection. The advice of the credit date is the one most often skipped in practice, and it is the one that makes delayed crediting visible.
Deposit insurance, if your institution is a bank. DICGC cover extends to state, central and primary co-operative banks up to ₹5,00,000 per depositor per bank, principal and interest together. The DICGC states equally plainly that primary co-operative societies are not insured by it. The first two cases in the table above involved societies, not banks.
What a depositor should do
- Establish whether it is a bank or a society, and therefore whether you are insured. Ask for the licence, not the signboard.
- Reconcile the passbook against the institution's own record, at the branch, at least once a quarter. A passbook alone proves what you were told; only the comparison proves what was posted.
- Keep every receipt until you have seen the matching credit. In an under-recording or unposted-receipt case, it is the only evidence that exists.
- Insist on the advice of the credit date at the time of collection, and turn on SMS alerts if the institution offers them.
- Report a missed visit to the branch, not to the agent. An agent who has stopped appearing is the one situation where a day of delay matters.
What an institution should control
- Same-day or next-working-day crediting, measured rather than declared. Track the value-date lag as a metric with an owner.
- Agent-wise and member-wise reconciliation, daily, with exceptions escalated rather than carried forward to the next day's sheet.
- Per-agent, per-day cash limits, which the Directions require of a bank's doorstep scheme in any case.
- Route rotation and surprise verification of a sample of members' passbooks against the ledger. Rotation is what makes a long-running under-recording pattern surface.
- A Board-approved fraud risk management policy. RBI's Master Directions of 15 July 2024 require every urban, state and central co-operative bank to have one, with early-warning signals, staff accountability and a Special Committee of the Board to monitor fraud cases. Tier 1 and 2 urban co-operative banks, and state and central co-operative banks with deposits below ₹1,000 crore, may run a Committee of Executives instead.
- Capture at the doorstep. A handheld thermal-printer terminal, or an agent app that records the receipt at the point of collection and queues it offline until it syncs, removes the blind window rather than shortening it: the record exists before the agent leaves the street. These are categories of technology, not recommendations, and we name no product here.
The risk that does not get counted
Agents carry cash along a known route at a known time — a description of a target. In August 2025 Deccan Herald reported an arrest in Udupi for stealing cash from a pigmy collection agent, with ₹32,000 in cash and a mobile phone recovered.
That is one reported incident, and one incident is not a pattern. But cash limits and route rotation are usually discussed as protections for the ledger, and they protect the person carrying the bag just as much. An institution reviewing its collection controls should count the agent among the people those controls exist for.
What we could not establish
- How often agent-level defalcation actually happens. No published figure isolates doorstep or pigmy collection losses from other bank frauds, at any date we could find. Anyone quoting a rate is estimating.
- Whether the Nagpur society's "illegal commissions to agents" were pigmy commissions. The report says agents; we have not assumed further.
- A current fraud breakdown for urban co-operative banks. The table above ends in March 2019 and comes from an RTI reply, not a routine published series.
- The full facts of the Udupi theft. The publisher's page served us only the headline and standfirst, so the amount collected before the theft is not established.
- How many institutions have moved collection off paper. No survey measures the paper, device and app split, so the size of the blind window across the sector is unknown.