Accountability Ledger
Of the people. By the people. For the people.

Why this is published. This ledger exists so that any citizen can see how public office was exercised, what it cost, and whether anyone was ever made to answer for it. It concerns the public conduct of people who held public power. It does not concern their private lives, and it never will. It is published in the public interest, which in a democracy is the entire reason a citizen is permitted to ask in the first place. The same test is applied to every government in scope, with no party affiliation and no money from any of them: was anyone made to answer? A government that finds that question intolerable has told you something about itself, not about the question.

This is an opinion and analysis resource that makes the accountability case. Holding power to account is the right of every citizen of a democracy, and this document exercises that right without apology. Read it as an argument built from the public record, not as a neutral encyclopedia.

Its force is its accuracy, so fact and allegation are kept strictly apart. Court verdicts, official audits, government admissions and documented death tolls are stated as fact. Where a scam or personal wrongdoing is alleged, it sits in a marked "Contested / alleged" field, attributed to the opposition, to reporting, or to a court process, and never asserted here as proven. Some matters remain in court. Each case also records the government's own defence, because an indictment that survives the counter-argument is the one that cannot be waved away.

"Ministers responsible" names who held the portfolio. That is a matter of record, not by itself a charge of personal guilt. Several failures belong to state governments and are labelled as such rather than pinned on the Union. The dashboard sums deaths and costs recorded in cases under a portfolio; it never attributes those figures to an individual.

Sources carry a tier badge. T1 primary record (court, audit, gazette, official publication). T2 independent reporting or documented research. T3 partisan or advocacy; useful as a pointer, never as the only proof. Where an archived copy exists, it is linked alongside the live URL.

Scope is deliberate, and it is not one-sided. This ledger runs from 2000 onward and covers every government that has held power in that time: the NDA under Vajpayee, the UPA under Manmohan Singh, and the NDA since 2014, along with state governments of every party. A ledger that begins in 2014 can be waved away as opposition material, and it should be. This one records the defence-deal sting that brought down a BJP president in 2001, the spectrum and coal allocations of the UPA years, illegal mining under a BJP state government, health funds looted under a BSP one, and counterfeiting protected by a state police force. The test applied is identical in every case and has nothing to do with party: was anyone made to answer?

Being even-handed also means recording the cases that did not survive scrutiny. Some entries here are famous accusations that collapsed: charges the courts threw out, presumptive loss figures that were never tested, ministers who were cleared. Those stay on the ledger, marked as such, because an accountability record that only remembers the accusations and forgets the acquittals is propaganda with footnotes. Where a court cleared someone, this ledger says so in the same breath as the allegation. On each case card, the government's answer sits beside what the record shows.

It also means recording the resignations. The counter at the top of this page is not a rhetorical zero; it is counted from the cases themselves. Every time an office-holder actually left over something on this file, it is logged against that case with the office and the date, and the header adds them up. That is why the number is what it is rather than what would be convenient. Union ministers, state ministers and officials are counted separately, because rolling them together is how a headline figure stops being true. If a minister resigns tomorrow over a case here, the number goes up on its own.

Every number tagged EST is an estimate, often a disputed range. Treat it as order of magnitude and verify before quoting.

The headline cost and human-toll figures are broad, cumulative estimates. They sum only the explicit per-case estimates recorded in the data file, including disputed and range-based figures; cases can overlap and the totals are not an audit or a claim that every amount was stolen. Read them as a scale, not a balance sheet: they mix one-off tolls with annual figures, and money raised with money lost. The two largest single entries are ₹24.85 lakh crore of central fuel excise, which is lawful revenue where the accountability question is how it was split with the states and whether the cess reached the funds it was raised for, and roughly 2.1 million deaths attributed to air pollution in one year, 2021.

Every case on this page carries at least one link you can open and check. That is enforced by the build, not by good intentions: a case that cannot be sourced is held back and does not appear here, however well known it is. Where the record is a court judgment, an audit or a government release, that is what is cited. Where a matter is contested, the citation is to the reporting or the filing that documents the dispute. This is a large catalogue, not a complete one.

If we have something wrong, say so. There is a standing corrections and right-of-reply process, open to anyone, including anyone named on this page.