Methodology

How we build a case

This page describes the standard every case on this site is held to. It exists so any claim we make can be checked against the same process, rather than taken on trust.

What counts as a case

We cover financial frauds, procurement scandals, and systemic governance failures that are backed by an official finding — a CAG audit, a court order, a filed chargesheet, or a government body's own review — not by rumour or unverified allegation alone.

Source hierarchy

Sources are used in this order of priority:

CAG performance audits and other official government reports
Court orders, judgments, and filed chargesheets
Parliamentary records and ministry disclosures
Reputable, named journalism and legal-affairs reporting

We were not able to link directly to a certified court judgment for every case — where we rely on journalism's reporting of a court order rather than the order itself, that source is labelled “independent,” not “primary,” even though it concerns an official action.

Confidence levels

Every verdict carries a confidence rating, because not all evidence is equally solid:

High — Multiple independent sources agree, including at least one primary source, and the core facts are not seriously contested.
Medium — The core facts are documented, but a key figure is a disputed estimate, or the matter's legal outcome is still pending.
Low — Sources conflict materially on scale or on basic facts, and no institutional finding has resolved the disagreement. We report the conflict rather than pick a side.

Case status labels

These labels describe an institutional or legal status, not a moral judgment. In particular, “Wrongdoing Confirmed” and “Convicted” are different claims — the first means an official body found irregularity or illegality; the second means a court found a specific person criminally guilty. A case can have the first without the second, as several on this site do.

🔴Convicted

At least one accused has been convicted by a court. Appeals may still be pending.

🟠Wrongdoing Confirmed

An official body (CAG, Supreme Court, a ministry audit) has found irregularity or illegality, but no criminal conviction has been secured.

🟡Under Trial

Charges have been filed and proceedings are active. No verdict yet — the accused are presumed innocent.

⚪Acquitted / Closed

A court acquitted the accused, or the case was closed/discharged without a conviction. An appeal may still be pending.

🟣Disputed Facts

Key figures or claims are genuinely contested between sources, with no institutional finding to settle them.

Presumption of innocence

Under Indian law, an accused person is presumed innocent until convicted by a court. Several cases on this site are under trial, on appeal, or contested in extradition proceedings abroad. We report what has actually been filed, charged, or decided — a chargesheet, an arrest, a bail order, a conviction, an acquittal — and we do not use language implying guilt beyond what a court has established. Where we quote a serious allegation that has not been judicially tested, we attribute it to its source (a CBI FIR, an ED complaint, a CAG finding) rather than stating it as this site's own finding of fact.

Ranking: biggest to smallest

Cases are ordered using editorial judgment that weighs three things together: the scale of public money implicated, the strength of the official finding (a Supreme Court ruling carries more weight than a single contested estimate), and the case's national significance. This is deliberately not a pure sort by rupee amount, because the figures involved are not measured the same way — a CAG “presumptive loss” estimate (an opportunity-cost calculation), a CBI chargesheet's alleged fraud amount, and a court-confirmed cost overrun are three different kinds of number, and treating them as directly comparable would itself be a distortion.

Where a case's headline figure is disputed or methodology-dependent, that is flagged directly on the case page, usually in the Gap section.

Three ways to browse

The database can be sorted three ways, and none of them is the “real” order — they answer different questions. “Latest first” sorts by when the wrongdoing was first publicly reported or put on record — an FIR, a CAG or audit report, a raid, or the first news exposé — not by its most recent verdict, bail order, or update, and not by when we happened to publish or last edit the page. That deliberately means an old case can still surface a fresh instance (a newly reported scandal), while a decades-old case that just received a new bail order stays where its scandal actually broke. “Timeline, since 1947” orders cases by when the underlying promise, policy, or deal dates from — earlier still, and often years before the scandal was exposed — including a visible marker for decades with no case tracked on this site yet, so the gaps in coverage are honest rather than hidden. “Ranked, biggest first” uses the editorial ranking described above.

The homepage summary numbers

The “By the numbers” section totals two things that should not be confused with each other. The first — a sum of each case's own headline alleged or estimated figure — mixes a CAG presumptive-loss estimate, several CBI/ED alleged-fraud amounts, and a chit-fund estimate into one number purely to give a rough sense of combined scale. It is explicitly not a claim that this amount of money was actually stolen, and it excludes cases (like Commonwealth Games 2010, AgustaWestland, and the Jeep Scandal) whose headline figure is not this kind of amount at all.

The recovery-rate figure is narrower and more defensible: it compares alleged and recovered amounts only for the specific cases where both exist on a like-for-like basis, and says so. A low recovery rate across those cases is a real, checkable pattern — a single grand “total India has lost to fraud” number would not be.

Charting standards

Every chart uses a shared, zero-based scale for the values being compared. We do not truncate axes or use inconsistent scales to make a gap look larger or smaller than the underlying numbers support. Where two numbers being compared are not directly commensurable — for example, two competing loss estimates rather than a target versus an actual outcome — the chart's note says so explicitly.

What we don't do

We do not fabricate statistics, quotes, or documents.
We do not present a chargesheet, FIR, or allegation as a proven fact.
We do not speculate about a person's motives or state of mind.
We evaluate documented decisions, spending, and legal outcomes — not unverified character claims.

Corrections

If a source is found to be wrong, outdated, or superseded by a later ruling, the case is updated and the change is reflected in its “Updated” date. Case status in particular can change quickly — bail orders, chargesheets, and appeal outcomes are common — and we aim to revisit high-profile cases when a significant development is reported.

← Browse the ranking