Why This Part Is Different
This is an advocacy document. It is the case for the prosecution against Congress governments in power. The selection of material is one-sided on purpose. The evidence in it is not. Every serious claim carries a grade showing how well it is proved, the other side is given its strongest argument, and where the other side wins, this book says so. A second series applying the same test to the government of 2014 onwards is planned.
Thirteen parts of this book have been a list of things done badly. This one is not, and you should know that before you start.
Between 1991 and 1996 a Congress government did the single largest useful thing any Indian government has done since 1947. It took an economy that had been strangled for forty years by its own rules and it cut the rules away. Part Four of this series spent sixty-eight pages explaining how much those rules cost the country. It would be dishonest to spend sixty-eight pages on the disease and then hurry past the cure.
The same government, eighteen months later, sat in Delhi while a mosque was pulled down in Ayodhya by a crowd of a hundred and fifty thousand people, after its own intelligence service had warned that this might happen, after the state government had told the Supreme Court in writing that it did not need central troops, and after nine days had passed in which somebody could have decided otherwise. Around two thousand people died in the riots that followed. That government then set up a commission to find out what had gone wrong, and wrote the commission’s rules so that it was not allowed to examine the central government at all.
Both of those things are true. They were done by the same Prime Minister, the same Cabinet, the same party, inside the same eighteen months. This part does not try to make them cancel out. It tells you what was done, in order, and it puts the good first — deliberately, and for a reason given in Chapter Two.
Almost every account of these years, friendly or hostile, assumes that a government with a big majority can do more than a government without one.
Test it here. Rajiv Gandhi had four hundred and fourteen seats. In 1989 he tried to put village councils into the Constitution. The bill passed the lower house and was defeated in the upper one, and it died. Narasimha Rao had two hundred and thirty-two seats and no majority at all. In December 1992 he passed the same reform, on substantially the same design, and it is in the Constitution today.
Part Twelve found that a majority is the capacity to legislate, not the capacity to decide. Part Fourteen finds the other half of it: a government without a majority has to persuade somebody, and persuading somebody is a form of thinking. The 1991 reforms were designed by a government that could be thrown out any Tuesday. That is not a coincidence, and it is uncomfortable for everybody — because the same weak government also lost the building.
One more warning. This part contains the highest proportion of credit to charge in the whole series. That is a fact about what happened between 1991 and 1996. It is also a fact about how this book selects material, and Chapter Nine turns that question on the book itself.
1A Government Nobody Expected
On the twenty-first of May 1991, Rajiv Gandhi was killed by a bomb at an election rally in Tamil Nadu. Part Twelve of this series ended there. This part begins in the month that followed, and the first thing to understand is that almost nothing about it was planned.
The election was interrupted by the assassination and finished in June. Congress won two hundred and thirty-two seats. It needed two hundred and sixty-nine for a majority. It did not have one, and it did not have a leader either, because the man who would have been Prime Minister was dead.
The party turned to P. V. Narasimha Rao. He had been a minister under Indira Gandhi and under Rajiv Gandhi — Home, Defence, Foreign Affairs. He was sixty-nine. He had not been given a seat to contest. He had, by several accounts, already shipped his personal library out of Delhi. He was sworn in as Prime Minister on the twenty-first of June 1991 and had to win a by-election later that year to stay in Parliament at all.
Before going further it is worth being exact about the number of seats, because two different numbers circulate and a reader who meets the other one somewhere else will stop trusting this book.
You will see Congress’s 1991 result given as 232 seats in some places and 244 in others. Both are correct and they are counting different moments.
Voting happened on the twentieth of May and the twelfth and fifteenth of June 1991 — but not in Punjab, where the situation described in Part Ten meant the poll was postponed to the nineteenth of February 1992. So the June 1991 count, on which the government was formed, is 232 out of 521 seats actually filled. Add Punjab and the by-elections that followed and the figure rises to about 244.
This book uses 232, because that is the number the government had on the day it took office and started making decisions. Where the later figure matters, it is given.
Two hundred and thirty-two out of five hundred and forty-five is not a majority. It is not close to one. What it means in practice is that the government could be voted out on any day that the opposition parties agreed with each other, and everybody in Delhi knew it.
1.2 — What was in the till
Part Four of this series set out the economic position in detail and this part will not repeat it. The short version is that in the summer of 1991 India was close to being unable to pay for imports.
The country’s foreign exchange reserves — the foreign money a country keeps to buy things from abroad — were down to about one point two billion dollars in January 1991, and roughly half that by June. In May, under Chandra Shekhar’s caretaker government, twenty tonnes of India’s gold were flown to a bank in Zurich as security for a loan. In July, forty-seven tonnes went to the Bank of England. Sixty-seven tonnes of gold, raising about six hundred million dollars.
When India buys oil, machinery or medicine from another country, the seller does not want rupees. He wants dollars, or pounds, or yen. So the Reserve Bank keeps a store of foreign money to pay with.
If that store runs out, India cannot buy anything from abroad. Not fuel. Not fertiliser. Not the parts that factories need. This is what people mean when they say a country has “defaulted”: it has run out of the money that outsiders will accept.
Numbers like “one point two billion dollars” do not mean much on their own. They sound enormous. The useful way to hold this number is to ask how long it would last.
India’s reserves in June 1991 covered about two to three weeks of imports.
Think of a household that spends thirty thousand rupees a month and has about twenty thousand rupees left, with the rent due, the school fees due, and no salary arriving. That household is not poor in some general way. It is a fortnight from something breaking.
That was the position of a country of eight hundred and forty million people, run by a government that could be voted out on a Tuesday.
1.3 — Why the weakness matters
It is tempting to treat the arithmetic as background — the sort of thing a book mentions once and moves past. It is not background. It is the single most useful fact for understanding everything else in this part, in both directions.
A government with two hundred and thirty-two seats cannot force anything through. It has to find votes for every measure, from parties that do not agree with it and do not owe it anything. That is why the reforms of 1991 were done the way they were done — fast, by executive order where possible, and without a public argument that would have lost. Chapter Two is about that.
It is also why, in December 1992, the government’s defence for not acting in Ayodhya was that it did not have the political strength to act. Chapter Nine takes that defence seriously, because it is the strongest one available and it is not obviously wrong.
And it is why, in July 1993, when the government finally faced a vote it could lose, somebody appears to have gone looking for votes to buy. Chapter Eight is about that.
One government, one weakness, three completely different outcomes. That is the shape of this part.
1.4 — What else was already burning
The empty treasury is the part of June 1991 that everybody remembers. It was not the hardest thing the new government inherited, and the rest of that inheritance explains most of what happens in Chapter Four.
Four fires were already lit.
Mandal. In August 1990 the V. P. Singh government had announced that twenty-seven per cent of central government jobs would be reserved for Other Backward Classes, acting on a report written in 1980. The reaction was violent. Students set themselves on fire. The matter was in the Supreme Court and stayed there until November 1992.
Mandir. Part Twelve ended with the Faizabad campaign launch of November 1989 and the foundation-laying ceremony that followed. What came after was not in Part Twelve: in September and October 1990 L. K. Advani travelled across northern India in a vehicle got up as a chariot, campaigning for a temple at the disputed site. He was arrested in Bihar in October 1990. The BJP withdrew support from the government, which fell. In the June 1991 state election the BJP won Uttar Pradesh, and Kalyan Singh became Chief Minister.
Indian political writing from this period uses these two words constantly, usually together, and rarely explains them.
Mandal is shorthand for job reservation for the Other Backward Classes, after the commission chaired by B. P. Mandal that recommended it. It sorts people by caste.
Mandir means temple, and is shorthand for the Ayodhya movement. It sorts people by religion.
The 1991 election was fought over both at once. That is why it is usually called the Mandal–Mandir election, and it is the reason a party that had two seats in 1984 had a hundred and twenty by 1991.
Punjab. Part Ten covered these years in full. In June 1991 Punjab was under President’s Rule, had been for four years, and would not hold an election until February 1992 — the election with roughly twenty per cent turnout.
Kashmir. The insurgency described in Part Seven had begun in earnest in 1989–90. By 1991 the state was under President’s Rule, the Pandits had largely left the Valley, and the Armed Forces (Special Powers) Act, whose full history is in Part Thirteen, had been extended to Jammu and Kashmir in 1990.
Set those four against two hundred and thirty-two seats and the position becomes clear.
In June 1991 the new government had two states under direct rule from Delhi, an insurgency in a third, a caste dispute before the Supreme Court, a religious movement organising openly around a specific building, and about a fortnight of money.
Any one of those would define a government today. This one arrived with all five, no majority, and a Prime Minister who had not expected to be in Parliament at all.
This is not offered as an excuse. It is offered because Chapter Nine takes the defence seriously, and a defence cannot be taken seriously if the reader has not been told what the government was defending against.
1.5 — Who was in the room
A short list of the people who matter in the pages that follow, because this part has more named individuals than most.
| Name | Position, 1991–1996 | Where they appear |
|---|---|---|
| P. V. Narasimha Rao | Prime Minister, 21 June 1991 – 16 May 1996. Also held Defence from March 1993 and Foreign Affairs from March 1992 to January 1994. | All chapters |
| Manmohan Singh | Finance Minister. An economist, not a politician; had never won an election. | Chapter Two |
| Shankar Dayal Sharma | President from 25 July 1992. R. Venkataraman before him. | Chapters Four, Six |
| Kalyan Singh | Chief Minister of Uttar Pradesh (BJP) until 6 December 1992. | Chapters Four, Five |
| Sukh Ram | Minister of State for Communications, January 1993 – May 1996. | Chapter Seven |
| Buta Singh | Minister; convicted alongside Rao in 2000, acquitted 2002. | Chapter Eight |
| Arjun Singh, N. D. Tiwari | Senior Congress leaders; left the party in May 1995. | Chapters One, Nine |
Rao holding the Defence and Foreign Affairs portfolios himself is not a detail. A Prime Minister who keeps three ministries is a Prime Minister who does not have people he can spare or trust with them.
On the twenty-first of June 1991 India got a Prime Minister who had been about to retire, leading a party with 232 seats out of 545, with about two to three weeks of foreign money in the bank and sixty-seven tonnes of the national gold already pledged abroad.
Every single thing in this part — the reform, the demolition, the money, the vote — happens inside that weakness. Hold on to it.
2What Was Done in Six Months
This chapter is placed here, second, before anything goes wrong. That is a deliberate choice and it is worth explaining, because the ordering of a book is an argument in itself.
Part Ten of this series put the militants’ own killings in Punjab before the chapters on what the state did, and said in the text why: placed afterwards, it would have read as an excuse offered late. The same reasoning applies in reverse here. If the reforms of 1991 appeared in Chapter Nine, after the demolition and after the scams, they would read as a sweetener — a good thing produced at the end to soften a bad account. So they go here, at full length, before the charge sheet starts.
2.1 — What was taken apart
Part Four described the licence system in detail. In one paragraph: to build a factory in India, or expand one, or change what it made, or import a machine for it, you needed permission from the central government. The permission was slow, it was rationed, and the people who got it best were the people who already had it. Part Four’s finding was that this system, not any single scandal, was the largest cost in this entire series.
Between July 1991 and the end of that year, most of it was removed.
A rule that said: before you may build or expand a factory, you must get a paper from Delhi saying you are allowed to.
The paper was called a licence. Getting one took months or years, needed contacts, and could be refused without a reason being given. It did not stop the biggest firms — they had offices in Delhi whose whole job was collecting such papers. It stopped everybody else.
In July 1991 the requirement was abolished for most industries. Not reformed. Abolished.
Alongside that, four other things happened quickly. The rupee was devalued in two steps in early July, by roughly twenty per cent in total, which made Indian exports cheaper abroad and imports dearer at home. Import tariffs, which were among the highest in the world, began to come down. The rules restricting foreign companies from owning shares in Indian ones were loosened. And the government went to the International Monetary Fund for a loan of about two point two billion dollars.
The speed is the part people forget. The government took office on the twenty-first of June. The industrial policy statement came on the twenty-fourth of July, the same day as the Budget.
Thirty-three days from taking office to dismantling a system that had been in place since 1951.
That is roughly the time it takes to get a passport reissued. A structure that four decades of governments had defended, argued for, expanded and refused to touch was taken apart in the time between one month’s electricity bill and the next.
This is either the most impressive thing in this book or the most alarming, and Chapter Nine argues that it is both.
2.2 — Did it work
Yes, and this book is not going to be coy about it. Growth over the following decade ran at roughly six per cent a year, against the three and a half per cent that Part Four recorded for the period 1950 to 1980. India did not default. The reserves recovered. The gold came back.
But a prosecution document has to be careful with a success, because it is exactly where a reader expects to be handled. So here is the honest shape of the thing, including the parts that do not flatter.
The strongest case that 1991 deserves less credit than it gets:
First, it was not a choice. The country was days from default and the IMF loan came with conditions. A government that does the right thing at gunpoint has not demonstrated much. Part Four already made this point about the timing: the argument for reform had been won inside the government’s own committees by 1969, and what changed in 1991 was the bank balance, not the argument.
Second, the 1980s acceleration came first. Growth had already picked up to around five and a half per cent in the 1980s, under Congress governments, before any of this. Some economists argue 1991 continued a trend rather than starting one.
Third, the design was narrow. Industry was freed. Agriculture, labour law, land, electricity and the banking system were mostly not touched, and thirty-five years later several of them still have not been. The reform did the parts that were politically cheap.
Where this argument fails: on the first point, plenty of governments have faced a crisis and done nothing useful — Part Four lists the committees from 1965, 1967 and 1969 whose correct diagnosis went nowhere for twenty-two years. Being forced to act does not tell you what a government will do when forced. On the second, the 1980s growth was borrowed: it was financed by exactly the deficits that produced the 1991 crisis. On the third, the argument is true and it is a criticism of what came later, not of what was done in 1991.
The one thing that should not be claimed for 1991 is that it was a plan for the whole economy. It was an emergency operation on one part of it, done well, by a government that could not have survived a public debate about it.
2.3 — What was not touched
The reform had a shape, and the shape has a pattern. It is easiest to see in a table.
| Reached in 1991–93 | Not reached |
|---|---|
| Industrial licensing — abolished for most sectors | Labour law — untouched; the rules on hiring and dismissal in larger firms remain substantially as they were |
| Import tariffs — cut sharply from among the world’s highest | Agriculture — no equivalent reform; the procurement and subsidy structure Part Five described was left standing |
| Foreign investment — rules loosened | Electricity — state boards left as they were, and their losses grew through the decade |
| The rupee — devalued, then made convertible on trade account | Land — no change to acquisition law; the 1894 Act described in Part Thirteen stayed in force until 2013 |
| Capital markets — SEBI given statutory teeth after the 1992 scam | Public sector banks and loss-making state firms — largely untouched |
The left column is what could be done by executive decision or a Finance Bill. The right column is what would have required legislation, a fight with organised interests, or both. That is the whole pattern.
This is not a criticism of anybody’s courage. It is the arithmetic of two hundred and thirty-two seats. A government that cannot survive a vote does the things that do not need one.
But it has a consequence that this book is obliged to state, because the people in the right-hand column are not the same people as in the left.
The reforms freed the part of the economy where about one in ten Indians worked. They did not touch the part where roughly six in ten worked, which was agriculture.
Put it as a village. The rules changed for the man who wanted to open a small factory. They did not change for the nine of his neighbours who farmed, whose crop prices, water, electricity and land were still governed by the arrangements Part Five described.
Growth arrived. It arrived in the places where the rules had been changed, which is what one would expect, and it is the beginning of an argument about regional and rural inequality that is still running.
2.4 — The part that is genuinely uncomfortable
The reforms of 1991 were never put to the country. There was no election fought on them, no white paper argued in public, no vote in which the electorate was asked whether it wanted this. They were announced.
Part Four’s Chapter Nine flagged this and said the consequence would show up later: because the change came under pressure rather than persuasion, the argument for it was never actually had, which is why the case still has to be made from the beginning every time somebody proposes the next stage.
This is not a small point, and it is the one place where the two halves of this part touch.
The best evidence that the 1991 reforms were never argued in public is what the Congress party itself said about them afterwards.
They do not appear as a central boast in the 1996 campaign. The party fought that election on other ground and lost badly — 140 seats, its worst result to that point. Manmohan Singh, the Finance Minister who executed the reforms, contested a Lok Sabha seat in 1999 and lost.
A government that had genuinely persuaded the country of something would have run on it. This one did not, because it never tried.
So the honest verdict on the largest achievement in this book is a split one, and it goes on the page in that form.
The decision was right. The licence system was indefensible, Part Four established that at length, and removing it was the correct act. Full credit, without qualification, to a Congress government.
The execution was right. Fast, under emergency conditions, by a government with no majority. It worked. India did not default.
The method left a debt. Nothing was put to the country. No argument was won in public. The result is that the reform has no political owner and no popular mandate, and every extension of it since has had to start the argument from nothing. That is a real cost and it is still being paid.
That split verdict is the shape of this whole part in miniature: a right decision, taken well, that left behind a problem nobody has since solved.
In thirty-three days, a Congress government with no majority took apart the licence system that Part Four identified as the largest single cost in this series. It was the right decision, it worked, and it goes in this book as credit.
It was also never explained to anybody, never voted on, and never defended in an election. The country was rescued without being asked. Hold that sentence — it comes back in Chapter Four in a much worse form.
3What Else Was Built
The 1991 reforms are the famous achievement of these five years. They are not the only one, and one of the others matters more to this series than it does to most accounts, because this book promised it four hundred pages ago.
3.1 — Village councils, and a debt paid
Part Twelve of this series recorded that Rajiv Gandhi tried in 1989 to put local government into the Constitution. India had village councils, but they existed only because state governments allowed them to. A state could dissolve a council, postpone its election for years, or simply give it no money. The reform was meant to stop that by writing the councils into the Constitution itself.
The bill passed the Lok Sabha and was defeated in the Rajya Sabha in 1989. Part Twelve left it there and said this part owed the ending.
Panchayat means a village council. Panchayati raj means rule by such councils — the layer of government below the state, closest to where people actually live.
Before 1993, these councils had no protection in the Constitution. They existed at the pleasure of the state government, which could abolish them, ignore them, or starve them of money at will. Many states did.
Here is the ending. On the twenty-second and twenty-third of December 1992, the Lok Sabha and the Rajya Sabha passed the Seventy-third and Seventy-fourth Amendments to the Constitution. Seventeen states ratified them. The President gave assent on the twentieth of April 1993, and they came into force on the twenty-fourth of April and the first of June 1993.
They added a new Part IX to the Constitution, Articles 243 to 243-O, and an Eleventh Schedule listing twenty-nine subjects the councils would handle. They made elections compulsory and on time. And they reserved one third of all seats for women, at every level.
The one-third rule brought roughly eight hundred thousand women into elected office in a single round of elections.
That is more elected women than there were in every national parliament in the world put together at that time. It happened at the bottom of the system, in villages, where nobody was watching, in a country where in 1993 most rural women could not read.
It is the largest single act of political inclusion in Indian history and it was passed by a government with two hundred and thirty-two seats.
The dates in that sequence deserve a second look, because they are not comfortable.
Those dates are not comfortable, and they are almost never printed together. The Babri Masjid came down on the sixth of December 1992. Riots followed across the country and around two thousand people died. Sixteen days later the same Parliament passed two constitutional amendments changing how six hundred thousand villages are governed.
Told separately, each is a familiar story. Told together they break the picture most people carry, in which a government is either functioning or it is not. This one was doing the most important constitutional work in twenty years and failing at the most basic function of the state in the same fortnight. Governments are not single things. They are many machines running at once, and the failure of one tells you very little about the others.
3.2 — And what the states did with it
A constitutional amendment is a permission, not a result, and this book would be doing exactly what it criticises if it recorded the passing of the Seventy-third Amendment and stopped there.
The amendment made three things compulsory: that councils exist, that elections be held on time, and that a third of seats go to women. It made almost nothing else compulsory. What the councils would actually do — which of the twenty-nine subjects in the Eleventh Schedule they would control, what money they would receive, which officials would answer to them — was left to each state legislature to decide.
Everybody assumes that putting something in the Constitution settles it.
The Seventy-third Amendment guaranteed that village councils would exist and would be elected. It did not guarantee that they would be given anything to do or any money to do it with, because those two questions were handed to the state governments — the precise bodies whose power the councils were meant to reduce.
Part Thirteen found the identical shape in the Fifth Schedule: protections for tribal land handed for enforcement to the state governments with the strongest interest in transferring it. The finding there was that a right is worth the independence of whoever enforces it.
Here it is again, in a different Part of the Constitution, thirty-eight years later, under a government that meant well. The amendment devolved the compulsory part and left the valuable part discretionary, and three decades of state practice have gone into the gap.
The record since is uneven and it is not this part’s subject, but the direction is not disputed. Elections happen, which is the amendment working. Funds and functions were transferred slowly, partially, and differently in every state, and in many states the money that reaches a council is tied to schemes designed elsewhere.
None of that takes away from what was done in December 1992. It does mean the credit has to be stated precisely: the amendment made local democracy permanent, and left local government optional.
3.3 — A siege that ended without a shot
In October 1993, armed militants occupied the Hazratbal shrine in Srinagar. Hazratbal holds a relic believed by Kashmiri Muslims to be a hair of the Prophet. It is the most sacred building in the Valley.
This series has already described what happened the last time an Indian government sent troops into a shrine. Part Ten set out Operation Blue Star in detail: the Akal Takht destroyed, the Sikh Reference Library burned, a wound that has not closed in forty-two years. Nine years later, a government faced the same problem in a different religion’s holiest building.
It did not storm it. The siege lasted around a month. Wajahat Habibullah, then Divisional Commissioner and a civil servant rather than a soldier, negotiated. Food was sent in. The militants surrendered. The shrine was not damaged and the relic was not touched.
Habibullah wrote his own account of the Hazratbal negotiation and it is unusually checkable, because he was injured immediately afterwards — a military truck hit the car he was travelling in with General Zaki, fracturing his skull.
He is also a hostile witness for parts of this book. The same man appears in Part Twelve, where his memoir contradicts the charge that Rajiv Gandhi ordered the Ayodhya locks opened. A source who damages your case in one chapter is worth more, not less, when he supports it in another.
Hazratbal goes in this book as a success, without qualification. It is the correct handling of a situation that had been handled catastrophically in 1984, by a government of the same party, and the difference was that somebody decided to wait.
It also has to be set against what else happened in Kashmir in those same weeks, because this book does not get to keep the good half of a month.
Bijbehara, 22 October 1993
On the eighth day of the Hazratbal siege, Border Security Force personnel opened fire on a protest march in the town of Bijbehara. Around fifty people were killed. Court-found in the sense that Indian courts later awarded compensation; the official account was that the force had been fired on first.
Human Rights Watch rejected that account, citing the United States State Department’s 1993 country report, which recorded that despite the claim of an ambush, one BSF sub-inspector was injured. Fifty dead on one side; one injury on the other.
Part Thirteen set out the rule that makes this possible: Section 6 of the Armed Forces (Special Powers) Act, which bars prosecution without permission from the central government. That permission was not given here either.
Charar-e-Sharief, May 1995
In December 1994, militants under a Pakistani commander occupied the shrine at Charar-e-Sharief, a wooden building around five hundred years old and the second most revered site in the Valley. In May 1995 it burned to the ground.
Who burned it is genuinely disputed, and this book will not pretend otherwise. Alleged in both directions: the Indian government said the militants fired it as they escaped; local residents said the army did it. Habibullah — the same man who is a credit witness three paragraphs above — is critical of how the army handled it and inclines to the local account.
Part Six established the discipline that applies here. Where the state will not release what it knows and the only other source has its own case to make, the grade stays at Alleged and the book does not promote a source because its conclusion is convenient. The charge that survives on the evidence is narrower: that having successfully waited out one shrine occupation in 1993, the government did not repeat the method in 1995.
3.4 — The rest of the credit column
Two more items, stated briefly because they are outside this book’s subject but belong in an honest ledger. Diplomatic relations with Israel were normalised in 1992, ending forty years of a position that had cost India something and gained it little. And what became known as the Look East policy turned Indian trade and diplomacy towards South-East Asia for the first time.
One statement from these years should also be recorded, because Part Seventeen of this series will collect the damaging things Congress leaders said and this is the opposite case — a thing said that was not damaging at all, and was simply not followed by anything.
The defence says: Rao told Parliament in 1995 that on autonomy for Jammu and Kashmir, “the sky is the limit” — that short of independence, the Constitution had room for almost anything. That is a more generous offer than any Indian government has made before or since, and it shows a Prime Minister genuinely looking for a settlement.
The reply: nothing followed it. No proposal was drafted, no talks were opened on that basis, no bill was brought. Part Seven of this series traced Article 370 being hollowed out year by year by Presidential Orders while the word “temporary” sat at the top of it. The 1995 statement joins that record as one more sentence that was true when spoken and had no consequences at all.
Where the defence still has something: a government fighting an insurgency, with no majority, that says in public that autonomy is negotiable, has taken a real political risk for nothing. It should be recorded that he said it.
Taken together, these months are the hardest thing in this book to hold in one hand — which is exactly why they are placed before the chapters that follow rather than after them.
The same government that lost the mosque wrote village councils into the Constitution sixteen days later, and put eight hundred thousand women into elected office. It ended the Hazratbal siege without firing into the shrine, which is what Part Ten shows the alternative looks like.
It also presided over Bijbehara, where fifty people died and one soldier was injured, and over the burning of Charar-e-Sharief, where it will not release what it knows.
The credit and the charge are not two different governments. They are the same one, in the same months.
4Nine Days
This chapter is not about who pulled the mosque down. That question has been asked for thirty-four years by courts, commissions and journalists, and the honest answer is set out in Chapter Ten: nobody has established it, and after this long, nobody is going to.
This chapter is about a different question, and the difference is the whole point.
Every account of the sixth of December 1992 assumes that the question is who ordered the demolition.
Notice what that assumption does. It puts the entire matter into a category — proving a secret instruction — where the evidence will never exist, where thirty-four years of inquiry has produced nothing, and where both sides can therefore say whatever they like for ever.
Ask instead: what did the central government do with the nine days between the twenty-seventh of November and the sixth of December? That question has an answer. It is written down. It is in affidavits, in intelligence reports, in the record of a Supreme Court hearing, and in the deployment orders of the central police forces.
Part One of this series set the rule and this is where it earns its keep: a charge built on documents the state produced itself cannot be answered by disputing the documents. The conspiracy question is unanswerable. The nine days are not.
4.1 — What the position was on the twenty-seventh of November
By late 1992 the dispute at Ayodhya was in the Supreme Court, and the court had ordered that nothing be built or demolished at the site. Muslim groups had filed contempt petitions saying Uttar Pradesh was already disobeying. A large gathering was announced for the sixth of December.
Two facts about that moment are load-bearing and both are on the official record.
The first is the affidavit. On the twenty-seventh of November 1992, the government of Uttar Pradesh told the Supreme Court in writing that it was fully able to protect the structure and that central forces were not needed.
A written statement given to a court, sworn to be true. Lying in one is a criminal offence.
An affidavit is not a press statement or a political promise. It is the highest form of assurance a government can give to a court, and it is why what happened nine days later became a contempt matter rather than merely a scandal.
The second fact is what the central government itself had put in front of the same court.
The Attorney General appearing for the Union handed the Supreme Court reports from the Intelligence Bureau — India’s domestic intelligence service — indicating that the structure was at risk of being demolished.
We know this from K. K. Venugopal, who appeared in those proceedings for Kalyan Singh and later became Attorney General of India himself, and who set it out in his own memoir. He records that the court did not give weight to the intelligence material, apparently because of the state government’s assurances.
This is the strongest kind of source available for a claim like this: a lawyer describing, against his own former client’s interest, what his own side was told.
Put those two together and the position on the twenty-seventh of November is this. The central government possessed intelligence that the mosque might be destroyed. It gave that intelligence to the Supreme Court. The state government, run by a party whose national campaign was built on this site, filed a sworn statement saying central troops were not needed. The court accepted the state’s assurance over the Union’s own intelligence.
And then nine days passed.
4.2 — How the site got there
The nine days did not come out of nowhere, and one step in the sequence is usually left out.
In October 1991 the new BJP government of Uttar Pradesh acquired around two and a half acres of land surrounding the disputed structure, stating that the purpose was tourism and amenities for pilgrims. The acquisition was challenged. In 1992 courts restrained construction on it.
Through 1992 the ground around the mosque was levelled and a platform built. Contempt petitions followed in both the Allahabad High Court and the Supreme Court, alleging that the state was already disobeying orders. The Supreme Court, on an application by counsel for the Chief Minister, confined the contempt question to events up to July 1992 on the ground that fresh allegations were arriving faster than they could be answered.
By November 1992 the position was therefore this: the state had acquired the surrounding land, the site had been prepared, contempt proceedings were already running, and a gathering of unknown size was announced for the sixth of December.
The sequence above is not reconstructed from journalism. It comes from court records — acquisition notifications, interim orders and contempt filings — all of which are public documents produced at the time by parties in conflict with each other.
That is the strongest kind of record this series uses. Part One set out why: a document produced by two sides who disagree, in a forum where lying is punishable, is worth more than any number of later accounts by people describing their own conduct.
Against that background, a sworn statement that the state was fully competent to protect the structure was not a routine assurance. It was an assurance from a government that was already answering contempt allegations about the same site.
4.3 — What the central government could have done
Three things were available, and it matters that they were ordinary powers, not extraordinary ones.
It could have deployed central forces at the site on its own authority. It had assembled a very large number of paramilitary personnel in the area — figures around one hundred and fifty thousand are commonly given — and they were not sent in.
It could have imposed President’s Rule in Uttar Pradesh under Article 356, dismissing the state government before the gathering. Part Three of this series described how that power was invented as an instrument in 1959; Part Six, Part Eight and Part Nine described its use. It had been used dozens of times by then, frequently on far thinner grounds.
It could have gone back to the Supreme Court and told it that the Union did not accept the state’s assurance. It did not do that either.
Between the affidavit and the demolition there were nine days. Two hundred and sixteen hours.
In the same government’s first nine days in office, in June 1991, it had assembled a Cabinet and begun drafting the industrial policy that would be announced five weeks later.
This is not a government that was incapable of acting in nine days. It is a government that acted in nine days when it wanted to.
4.4 — The sixth of December
The sixth of December 1992 was a Sunday. Around one hundred and fifty thousand people gathered. The structure was brought down within hours, by hand and with tools.
Riots followed across the country — Bombay, Delhi, Surat, Bhopal, Ahmedabad, Kanpur, Hyderabad. The commonly cited national toll is around two thousand dead, the majority of them Muslim. Part Eleven of this series already examined the Bombay violence of December 1992 and January 1993 as part of its comparison set, and this part does not repeat that account.
The judges of the Supreme Court assembled that afternoon at the Chief Justice’s house. Contempt proceedings against Kalyan Singh, who had resigned as Chief Minister, produced a token punishment — one day, lasting until the court rose — and a fine.
A sworn undertaking to the highest court in the country was broken, a building the court had ordered protected was destroyed, and around two thousand people died in what followed.
The punishment for the breach was a fine, and imprisonment lasting from the moment it was pronounced until the judges got up and went home.
A person caught travelling without a ticket on the Delhi Metro pays more and has a longer afternoon.
4.5 — What this chapter charges, and what it does not
This book will not say that the central government arranged the demolition, or wanted it, or ordered it. There is no evidence for any of that, the inquiry that ran for seventeen years did not find it, and Chapter Ten sets out why it never will be found.
Part Twelve refused to assert that Rajiv Gandhi ordered the Ayodhya locks opened, on the ground that a series which will repeat an unproved allegation against one man cannot demand proof when the allegation runs the other way. The same rule applies here and it applies to a Congress government’s benefit.
What is charged is narrower and it is not answerable.
Not established: that the central government wanted, ordered, arranged or connived at the demolition. Alleged and never proved, by anyone, in thirty-four years.
Established, on the government’s own documents: that it held intelligence indicating the structure was at risk; that it placed that intelligence before the Supreme Court; that it then accepted, without independent verification, a state government’s sworn assurance that central force was unnecessary — from a state government whose party’s national campaign was built on that site; and that in the nine days available it deployed nothing, dismissed nobody, and did not return to the court.
The charge is a decision, not a conspiracy. Somebody decided to rely on that affidavit. That decision is the thing to examine.
And it was never examined. That is Chapter Five.
That last line is not a rhetorical flourish. The next chapter is about the machinery that was built, ten days later, to make sure the decision was never examined.
Do not argue about who ordered the demolition. Nobody knows and nobody will.
Argue about nine days. On the twenty-seventh of November 1992 a state government swore to the Supreme Court that it did not need central troops. The central government had intelligence saying the mosque might come down — it had given that intelligence to the court itself. It then did nothing for nine days.
Every part of that is written down, and none of it depends on believing anybody’s account of a secret meeting.
5The Commission That Was Not Allowed To Ask
Ten days after the demolition, on the sixteenth of December 1992, the central government appointed a commission of inquiry. It was headed by Justice M. S. Liberhan, a retired High Court judge. It was told to report in three months.
It reported on the thirtieth of June 2009.
Seventeen years. Forty-eight extensions. Three hundred and ninety-nine sittings. Around a hundred witnesses. Roughly eight crore rupees.
A child born on the day the Liberhan Commission was appointed was old enough to vote before it reported.
It is the longest-running commission of inquiry in the history of the Republic. The three-month deadline was extended, on average, once every four months for seventeen years.
The delay is what everybody remembers about the Liberhan Commission, and the delay is not the important thing about it. The important thing is one line in the document that set it up.
5.1 — The terms of reference
A commission of inquiry cannot investigate whatever it likes. It is given a written list of questions by the government that creates it, and it is confined to them.
The written instructions that create an inquiry and tell it what it is allowed to look at.
They are drafted by the government. They are not negotiable by the commission. Anything not in them is outside the inquiry’s power, however obviously relevant it is — the commission cannot expand its own remit, and if it strays, its findings on that ground carry no weight.
The terms of reference are therefore the single most important document in any inquiry, and almost nobody reads them.
The Liberhan Commission was told to examine five things: the sequence of events at the site on the sixth of December; the role of the Chief Minister, ministers and officials of the Government of Uttar Pradesh; the deficiencies in the security arrangements as prescribed or operated by the Government of Uttar Pradesh; the assault on journalists that day; and any other matter related to the subject of inquiry.
Read that list again and notice what is not in it.
Examining the role of the central government was not part of the Liberhan Commission’s mandate.
The inquiry into the destruction of the Babri Masjid was set up by the central government, and was not permitted to inquire into the central government.
Every question in Chapter Four — the intelligence, the affidavit, the nine days, the forces not deployed, the Article 356 power not used — was outside the terms of reference of the only official inquiry ever held.
This is not the first time this series has met that device. It is the second, and the first one is twenty-nine years older.
Everybody assumes that a commission of inquiry is an instrument for finding things out. It is set up when something goes wrong; it has a judge; it takes evidence; therefore it is on the side of the truth.
But an inquiry’s terms of reference are written by the body most likely to be inquired into. That body decides, before a single witness is heard, which questions exist.
Part Six of this series found exactly this in 1963. The Henderson Brooks–Bhagat report on the 1962 defeat was given terms of reference that excluded Army Headquarters and the civilian direction of the army — that is, excluded the two places where the decisions had actually been taken. Its findings therefore land entirely on the field commanders.
Now the same shape appears in 1992, under a different Prime Minister, in a different decade, about a completely different subject. Two commissions, twenty-nine years apart, each scoped so that it could not reach the government that scoped it.
A device that appears twice, that far apart, is not an accident and it is not a conspiracy either. It is a technique that works, and it is available to any government, which is why Chapter Nine proposes the only rule that would stop it.
5.2 — What the report did find
It found a great deal, and the findings are serious. Commission-found throughout, by the Liberhan Ayodhya Commission of Inquiry, reporting 30 June 2009.
It held sixty-eight political leaders, and leaders of the RSS, the VHP and the Bajrang Dal, together with senior civil and police officials, accountable for the events leading to the demolition. It found the destruction had been carried out in a manner it called duplicitous and unworthy of a democratically elected government. It indicted Kalyan Singh for diluting the security around the structure. It found that the executive, the political class and the bureaucracy took no effective steps either to prevent the demolition or to arrest those responsible.
It also found that intelligence agencies at both state and central level had either been badly over-optimistic or had withheld material from the commission — and said that, not being a detective agency, it had been forced to work with fragments.
And on the central government it said this: that the Rao government had believed the assurances given by the BJP and its Uttar Pradesh government, and had been, in the report’s phrase, day-dreaming.
5.3 — Which is an exoneration, and this book has to say so
The commission did not target Narasimha Rao. It accepted the central government’s argument that in a minority government’s position, central forces could not simply be deployed into a state against that state’s stated wishes, and that President’s Rule could not be imposed on the basis of rumour.
That is the defence’s case, it came from the only official inquiry ever held, and it must be stated at full strength.
The defence says: a commission headed by a High Court judge sat for seventeen years, heard a hundred witnesses and examined this precise question. It concluded that the Prime Minister was credulous, not complicit. If the strongest word an inquiry can find after seventeen years is “day-dreaming”, the charge in Chapter Four is a charge of poor judgement, and poor judgement is not a crime.
It adds: what exactly was the alternative? Sending central paramilitary into Uttar Pradesh over the objection of an elected state government, days before a religious gathering of a hundred and fifty thousand people, would itself have risked mass casualties. Dismissing an elected government pre-emptively, on intelligence, would have been the precise abuse of Article 356 that thirteen parts of this book have condemned. A government that had done either would be facing a different chapter of this same series.
Where the defence wins: on the conspiracy charge, completely. Chapter Four already conceded it. And the point about Article 356 is a real one that this book cannot wave away, because it has spent four parts building the case against exactly that power.
Where it fails: the commission that produced the exoneration was not permitted to examine the central government. An acquittal from a tribunal without jurisdiction over the defendant is not an acquittal. And the alternatives were not the only two on offer — the government could have returned to the Supreme Court and told it that the Union did not accept the state’s affidavit. That required no troops, no dismissal, and no political capital. It required a lawyer and an afternoon.
That third option is the one to hold on to, because it disposes of the “minority government” defence entirely. A weak government can still tell a court that it does not believe something.
The report was leaked to a newspaper in November 2009 before being tabled in Parliament, and the leak itself became the story — the opposition alleged it had been released deliberately to break their unity on an unrelated sugarcane dispute.
That matters for a reason that has nothing to do with sugarcane. By 2009 the report’s contents were a tactical instrument in a fight about something else. Seventeen years is long enough for an inquiry into two thousand deaths to become a news-cycle weapon, and the parties on both sides treated it as one.
5.4 — What was done with it
When a commission of inquiry reports, the government must table the report in Parliament together with a memorandum saying what it intends to do about the findings. The Liberhan report was tabled on the twenty-fourth of November 2009, seventeen years after the demolition, with such a memorandum attached.
Nothing followed. No prosecution was begun on the strength of it. No official named in it faced disciplinary action as a result of it. The criminal case that already existed continued on its own separate track and ended in acquittals in 2020.
Seventeen years, three hundred and ninety-nine sittings, around eight crore rupees, and no consequence attached to any finding.
Divide it out: roughly one sitting every fortnight for seventeen years, to produce a document whose operative effect was a debate in Parliament and a press cycle.
Part Eleven found that four commissions and nine committees on November 1984 had produced a record and not accountability, and said that establishing a record has functioned as a substitute for accountability rather than a step towards it. This is that finding again, in a different decade, about a different event.
By the time the report arrived, in other words, it had stopped being an inquiry into a demolition and become a piece of parliamentary furniture. Seventeen years does that to a document.
The Liberhan Commission took seventeen years and forty-eight extensions, and that is the least interesting thing about it.
The interesting thing is one line in the instructions: the role of the central government was not within its mandate. The government set up the inquiry into the demolition and left itself out of it.
This series has seen that device before. Part Six, 1963, Henderson Brooks: terms of reference excluding Army Headquarters and the civilian direction of the war. Twice, twenty-nine years apart. It works both times.
6The Fifteenth of December
Nine days after the demolition, the central government did the thing it had not done in the nine days before it.
On the fifteenth of December 1992, President’s Rule was imposed on Madhya Pradesh, Rajasthan and Himachal Pradesh. Three elected state governments, all run by the BJP, dismissed at once. Their assemblies were dissolved. Uttar Pradesh, where the demolition had actually happened, had already gone the same way.
The stated ground was that these governments could not be trusted to uphold a secular constitution, having failed to implement the ban on the organisations involved.
Article 356 lets the central government dismiss an elected state government and rule the state directly from Delhi, if the President is satisfied that the state cannot be governed according to the Constitution.
Part Three of this series described how this power was turned into an instrument in 1959, when the elected Kerala government was dismissed. By 1992 it had been used around ninety times. Almost every use was against a state run by a party other than the one in power at the Centre.
So this is Article 356 again, and by this point in the series a reader knows what to expect: another dismissal, another abuse, another entry in the ledger. This chapter does not go that way, and it is important that it does not.
6.1 — Bommai
The dismissed governments went to court. Their cases were joined with an older one brought by S. R. Bommai, whose Karnataka government had been dismissed in 1989 — under a Congress government at the Centre, in Part Twelve’s period — without being allowed to prove its majority on the floor of the assembly.
On the eleventh of March 1994, nine judges of the Supreme Court gave judgment in S. R. Bommai v. Union of India. It is one of the two or three most important constitutional decisions in Indian history.
What it laid down changed Article 356 permanently.
| Before Bommai | After Bommai |
|---|---|
| The President’s satisfaction was treated as effectively beyond review. | Subject to judicial review. The material behind it can be examined for relevance and good faith. |
| A government could be dismissed on a Governor’s report that it had lost its majority. | Majority is tested on the floor of the House. A Governor’s opinion is not a substitute. |
| Assemblies were routinely dissolved at once, making the dismissal irreversible in practice. | Dissolution before Parliament approves is invalid. A struck-down proclamation means the assembly is restored. |
| No stated limit on the grounds. | Secularism held to be part of the basic structure, and available as a ground. |
Bench: Pandian, Ahmadi, Kuldip Singh, Verma, Sawant, Ramaswamy, Agarwal, Yogeshwar Dayal and Jeevan Reddy JJ. The leading judgment was written by Verma J. It remains the governing authority thirty-two years later.
6.2 — And it upheld the December 1992 dismissals
Here is the part that a one-sided book has to state plainly or lose the reader.
The Court did not strike down the dismissals of Madhya Pradesh, Rajasthan and Himachal Pradesh. It upheld them, on the secularism ground it had just recognised. The government that had done nothing for nine days before the sixth of December acted on the fifteenth, and the Supreme Court held that it had been entitled to.
And it went further in the opposite direction. Bommai’s own dismissal — the Karnataka case from 1989, under a Congress government — was held unconstitutional.
The defence wins this one outright.
The single most important judgment ever delivered on the abuse of Article 356 arose from a Congress government’s use of it, upheld that use, and struck down an earlier Congress government’s use of it from Part Twelve’s period.
This book has spent four parts building a case against Article 356 as an instrument of central control. On the fifteenth of December 1992 it was used against three governments of an opposing party, and the Supreme Court of India, nine judges sitting, said that was lawful.
Court-found, in the government’s favour. It goes in the ledger as a defence win, and Chapter Nine counts it.
Handling this honestly is not generosity. It is the reason the rest of the book can be believed. A document in which every single item is a scandal is a document that has been engineered, and any reader who knows one thing about Bommai will spot the omission and stop reading.
6.3 — What the judgment actually changed
A constitutional decision is only as good as what happens afterwards, and this series has recorded several that changed nothing. This one changed a great deal, and the change is measurable.
Before 1994, Article 356 had been used roughly ninety times in forty-four years — about twice a year, overwhelmingly against states governed by parties opposed to the one in power at the Centre. Part Three traced its invention as an instrument in Kerala in 1959; Part Eight and Part Nine traced it through the decades that followed.
After Bommai the rate falls sharply and stays low. The floor test requirement removed the most common device, which was a Governor reporting that a government had lost its majority without the government ever being asked to demonstrate otherwise. The rule that an assembly cannot be dissolved before Parliament approves removed the second, which was making the dismissal irreversible before anybody could challenge it.
The test of whether Bommai worked is not what commentators say about it. It is whether governments dismissed after 1994 have been restored by courts, and whether a Governor’s report still functions as the operative document.
Both have been tested repeatedly since, and in each case the courts have gone back to Bommai and applied it, including by ordering assemblies restored. A judgment still being used three decades later to reverse the acts of governments of every party is a working rule, not a statement of principle.
So this chapter’s concession runs further than the December 1992 dismissals themselves. The litigation those dismissals produced did more to end the abuse of Article 356 than anything else recorded in the thirteen preceding parts of this book.
What the prosecution can still say: the timing is the charge. The power that was too dangerous to use on the fifth of December, when it might have saved the building, was available and used on the fifteenth, when it could only punish. Both decisions were about the same states and the same party and were taken ten days apart by the same Cabinet.
The defence’s reply, which is strong: those are genuinely different acts. Dismissing a government before it does anything, on intelligence, is pre-emptive and is exactly what Part Three condemned in Kerala in 1959. Dismissing it after a demolition its own affidavit had promised to prevent is a response to a fact. The Supreme Court drew that distinction and it is a real one.
Where this leaves it: the defence has the better of the constitutional argument. What survives is smaller and is about consistency of nerve, not law — a government that believed on the fifteenth that these states were a danger to the Constitution had believed something less than that on the fifth, and the difference between the two dates is that two thousand people had died in between.
So this chapter closes with a concession rather than a charge, and the concession is load-bearing. Chapter Nine counts it among the four or five places in this series where the defence wins on the record.
On the fifteenth of December 1992, three BJP state governments were dismissed under Article 356. In S. R. Bommai, decided on the eleventh of March 1994, nine judges upheld those dismissals — and in the same judgment struck down a 1989 Congress dismissal from Part Twelve’s period, made Article 356 reviewable by courts for the first time, and required majorities to be tested on the floor of the House.
This is one of the small number of items in this series where the defence wins. Chapter Nine counts them. A book where nothing is ever conceded is a book nobody has to answer.
7The Money
Five separate financial scandals attach to these five years. They are usually told as a list, with the sums attached, as though the sums were the point. They are not, and this chapter is arranged to show why.
7.1 — The securities scam, 1992
On the twenty-third of April 1992, the journalist Sucheta Dalal published a report in the Times of India showing that around five hundred crore rupees had been taken out of the State Bank of India on the strength of receipts for government bonds that did not exist.
When two banks lent each other money for short periods, they used government bonds as security. Physically moving the bonds was slow, so instead the lending bank issued a piece of paper — a bank receipt — saying in effect: the bonds exist, we are holding them for you.
The entire system ran on the assumption that nobody would issue such a paper for bonds that did not exist. Harshad Mehta, a stockbroker, arranged for exactly that. Two small banks issued receipts against nothing. Large banks accepted them and handed over real money.
The money went into shares. Between April 1991 and April 1992 the Bombay index rose from 1,194 to 4,467 — a gain of about two hundred and seventy-four per cent, the largest in its history. Small investors, watching prices climb, followed. When the receipts were discovered the market collapsed and they were the ones holding the shares.
The amount is given differently in different places and this book has to be careful with it.
You will see the securities scam given as about ₹4,000 crore and as about ₹1,000 crore. They measure different things.
The larger figure, usually stated precisely as ₹4,025 crore, is the total value of funds diverted out of the inter-bank securities system on fraudulent paper. The smaller is an estimate of how much of that was actually pushed into the stock market.
The Joint Parliamentary Committee that investigated it is the primary source for the mechanism. This book uses the ₹4,000 crore figure and says what it counts, because a number quoted without its definition is the thing Part One warned about: an argument about meaning conducted as though it were an argument about arithmetic.
Mehta was arrested on the ninth of November 1992. Seventy-two criminal cases and around six hundred civil actions followed. He died in 2001. The chairman of one bank caught in the scandal killed himself.
And in July 1993, Mehta held a press conference, produced a suitcase, and said he had delivered one crore rupees in cash to the Prime Minister.
Alleged. Never proved. No court ever found it, no inquiry established it, and Mehta at that point was a man facing seventy-two prosecutions with an obvious interest in enlarging the story.
It goes in this book graded, and with this observation: the reason the claim survives is the suitcase, not the evidence. A physical object at a press conference gave the allegation a picture, and the picture has outlived every attempt to test the claim.
Part One’s chapter on fabrication makes the strategic argument: allegations like this are Congress’s best defence, because they let every documented charge be waved away as more of the same. The urea contract in section 7.4 ended in convictions. The suitcase did not. Keeping them in separate columns is what makes the first one impossible to dismiss.
7.2 — What Parliament did about it
Parliament set up a Joint Parliamentary Committee on the securities scam — only the second such committee in Indian history, the first having examined Bofors in 1988.
A committee drawn from both Houses, set up to investigate a specific matter. It can summon officials and papers.
Two things limit it. Its membership reflects the strength of the parties in Parliament, so the governing party has a majority on it. And its report is a recommendation: no part of it binds anybody, and no mechanism exists to make a government act on it.
Part Twelve recorded what happened to the first one, on Bofors in 1988: government majority, opposition boycott, and no consequence. This was the second.
The committee did serious work. It traced the mechanism, established how the receipts had been issued and accepted, and identified failures of supervision at the Reserve Bank and inside the public sector banks.
The lasting result was regulatory rather than criminal. The Securities and Exchange Board of India, which had existed since 1988 as an advisory body, was given statutory powers in 1992. The bank receipt system was overhauled. Share trading moved, over the following years, from paper certificates to electronic records — which removed at a stroke a whole family of frauds that depended on physical paper.
That is a real reform and it goes in the credit column. But it is worth being exact about what it was and was not.
The 1992 scam produced a better market and almost no punishment.
Trading is now electronic, the regulator has teeth, and a fraud of that exact shape cannot be repeated. Meanwhile the man at the centre of it died in 2001 with seventy-two cases still running, and the bankers who accepted receipts for bonds that did not exist mostly retired.
This is the pattern Part Eleven identified in a far graver setting: the system learns, and nobody is held responsible. Those are not the same thing, and the first is regularly offered as though it were the second.
7.3 — The Jain diaries
In March 1991, a man alleged to be connected to a militant organisation was arrested in Delhi. His interrogation led the CBI to raid the premises of a businessman, S. K. Jain. They seized foreign and Indian currency, two diaries and two notebooks. The books recorded large payments, against initials.
The initials corresponded to those of senior politicians and civil servants, across parties.
Then nothing happened for five years.
The diaries were seized in March 1991, three months before this government took office. Charge sheets were filed in 1996, as it was leaving.
That is the entire life of the government. A document naming its own ministers sat in a CBI cupboard from the first week to the last.
It moved only because a journalist, Vineet Narain, took the government to the Supreme Court under Article 32 and made it move.
The prosecutions that eventually followed collapsed in 1997 and 1998. The courts held that diary entries against initials, without corroboration, were not evidence on which anyone could be convicted. Everybody named walked, including three Congress ministers and, in the entries said to relate to him, the Prime Minister.
The acquittals are not the charge here, and it is important to be precise about that.
The defence says: the courts examined this material and found it worthless. Names in a private diary are not proof of anything. Every accused person was acquitted, across every party, which is what happens when an allegation has nothing behind it. Treating an acquittal as a conviction with extra steps is precisely the dishonesty this book claims to avoid.
The reply: agreed, entirely, on the acquittals. The charge is not that the accused were guilty. The charge is the five years. Evidence sufficient to justify investigation was in the state’s own possession from March 1991, and the state did not investigate until a private citizen compelled it through the Supreme Court. If the diaries were worthless, five years of a government’s own investigators would have established that in 1991 and the reputations named in them would not have been damaged for a decade.
What the defence keeps: nothing in this section supports a claim that anybody took the money. This book does not make one.
What came out of it was not a conviction but a rule. In Vineet Narain v. Union of India, decided on the eighteenth of December 1997, the Supreme Court issued twenty-six directions. It gave the Central Vigilance Commission a supervisory role over the CBI. It struck down the “Single Directive”, the rule that had required political permission before senior officials could even be investigated. It fixed the CBI Director’s tenure at two years so that a Director could not be removed for investigating the wrong person. And it established the technique of keeping a case open and supervising compliance over time.
Court-found, and it is the most important thing this scandal produced.
7.4 — Telecom
Sukh Ram was Minister of State for Communications, with independent charge, from January 1993 to May 1996. In 1996 the CBI raided his houses.
They found around three and a half crore rupees in cash. Some of it was in a shoebox in a Delhi hotel room. Over one crore was at his residence.
He was convicted twice. Convicted in 2002, three years’ rigorous imprisonment and a fine, over a contract for radio equipment supplied at inflated rates, causing a loss of around one crore sixty-six lakh rupees; two officials were convicted with him. Convicted again in 2011, five years and a four lakh fine, for taking a three lakh rupee bribe over a thirty crore rupee cable contract.
Those are convictions, the top of this book’s evidence scale. But look at the dates.
The offence was in 1996. The five-year sentence came in November 2011 — fifteen years later. He was eighty-four.
He was taken to Tihar jail on the nineteenth of November 2011 and released on bail by the Delhi High Court on the twenty-eighth, on grounds of age and heart disease.
Fifteen years to the sentence. Nine days in custody.
In between he had left Congress, formed his own party, helped the BJP take Himachal Pradesh in 1998, and rejoined Congress in 2019.
7.5 — Urea
In 1995, National Fertilisers Limited, a government company, put out a global tender for two hundred thousand tonnes of urea. The contract went to a Turkish firm.
NFL paid the entire sum — one hundred and thirty-three crore rupees — in advance, without a bank guarantee. The money went to an account in Geneva. No urea was ever delivered. The firm was not found in Turkey.
A promise by the seller’s bank: if this company takes your money and does not deliver, we will pay you back.
It is the most basic protection in international trade. No private business of any size sends a hundred and thirty-three crore rupees abroad without one. It is not a sophisticated safeguard that somebody clever got around. It is the first thing on the checklist, and it was not there.
The CBI registered the case on the nineteenth of May 1996, days after the government left office, and filed a charge sheet in 1997. The trial ended in 2018.
Convicted: two Turkish nationals, six years each and a hundred crore rupee fine. B. Sanjeeva Rao, a nephew of the Prime Minister, three years and a five crore fine. Prakash Chandra Yadav, son of a Union minister, three years and a one crore fine. Officials of NFL.
Twenty-three years from the payment to the verdict. And in a related NFL case, the CBI filed a closure report in January 2021 — twenty-two years after the case was registered — which the court rejected.
Two more cases against the Prime Minister himself should be recorded and their outcomes given. In the St Kitts matter he was charged with helping to forge documents showing that V. P. Singh’s son held twenty-one million dollars offshore with his father as beneficiary. In the Lakhubhai Pathak matter, a businessman said he had paid a hundred thousand dollars for a contract that never came. Charge sheets were filed in 1996. He was acquitted in both, in 2003. They go in the ledger as charges brought and lost.
7.6 — What the five have in common
Set them side by side and the sums are the least interesting column.
| Case | Amount | Outcome | Years elapsed |
|---|---|---|---|
| Securities scam (1992) | ~₹4,000 crore diverted | Prosecutions; Mehta died 2001 with cases running | 9+, unresolved at death |
| The suitcase allegation | ₹1 crore | Alleged; never proved | Never tested |
| Jain diaries | ~₹65 crore across those named | All acquitted 1997–98 | 5 years to charge sheet |
| Telecom (Sukh Ram) | ₹3.6 crore found in cash | Convicted 2002 and 2011 | 15 years to the second sentence |
| Urea (NFL) | ₹133 crore paid in advance | Convicted 2018 | 23 years |
The two cases that ended in convictions are also the two that took longest. That is not a coincidence and section 7.6 explains why.
Every account of political corruption assumes that the seriousness of a case is measured by the amount of money. This is why scandals are named after their sums.
But look at the column that actually varies. The amounts here run from one crore to four thousand crore — a factor of four thousand — and they predict nothing. What predicts the outcome is elapsed time.
Fifteen years to a sentence means the convicted man is eighty-four and is bailed on medical grounds in nine days. Twenty-three years means the money is untraceable and the company has dissolved. Five years to a charge sheet means the evidence has gone cold and the case fails in court.
Delay is not a failure of the anti-corruption process. For these purposes it is the anti-corruption process. A system in which every case is eventually decided, and every case takes two decades, produces exactly the same result as a system with no cases in it — except that it also produces the appearance of accountability, which is more useful to a government than impunity would be.
Part Eleven found the same shape in a different setting: transparency improving while accountability did not, so that establishing a record functioned as a substitute for consequences rather than a step towards them.
7.7 — Why nobody has to decide anything
The useful thing about the delay finding is that it does not require anybody to have done anything wrong.
No minister ordered a case to go slowly. No file says “delay this”. The CBI is understaffed and its officers are transferred. Special courts have too many cases. Adjournments are granted because a lawyer is unavailable. Appeals take their turn. Each individual step is defensible and most are correct.
Part Thirteen described the fifth form of a recurring pattern in this series: an addition nobody has made, where every component exists, nothing is concealed, and no decision exists that anybody has to defend. This is the same shape in a courtroom. There is no decision to delay. There is only the absence of a decision to hurry, and nobody is obliged to make one.
Five money scandals. One ended in nothing because it was never real. One collapsed in court because a diary is not evidence — but sat unexamined for five years until a journalist forced the government’s hand in the Supreme Court.
Two ended in convictions: telecom in fifteen years, urea in twenty-three. The man sentenced to five years served nine days.
Stop looking at the amounts. Look at the years. The amounts differ by a factor of four thousand and tell you nothing. The years tell you everything.
8The Vote
In July 1993 the opposition moved a motion of no confidence. On the arithmetic, the government should have lost. It won, two hundred and sixty-five to two hundred and fifty-one.
Afterwards it was alleged that members of the Jharkhand Mukti Morcha had been paid to vote with the government. One of them, Shailendra Mahato, turned approver — he confessed and gave evidence against the others.
A no-confidence motion is a vote in Parliament on whether the government should continue. If the government loses, it must resign. It is the only routine way, between elections, of removing a government.
An approver is somebody accused of a crime who admits it and gives evidence against the others, usually in exchange for a pardon. Approver evidence is treated cautiously by Indian courts, because a person who has admitted lying and has been promised something has two reasons to say what the prosecution wants.
Nothing happened while the government was in office. Investigation began in earnest in 1996, after it had gone. What followed took thirty-two years and it is the fullest example in this series of a charge moving up and down the evidence scale.
8.1 — The 1998 judgment
The case reached a Constitution Bench of the Supreme Court. In 1998, by three votes to two, it decided something that surprised almost everybody.
Article 105(2) of the Constitution says that no Member of Parliament shall be liable in any court for anything said, or any vote given, in Parliament. The purpose is obvious and good: a member must be able to speak and vote without fear of being prosecuted for it.
The majority held that this protection extended to a member who had taken a bribe and then voted as agreed. The bribe was “in respect of” a vote, so the immunity covered it.
The consequence was absurd on its face, and the Court said so itself years later.
Under the 1998 judgment, the members who took the money and voted as promised could not be prosecuted. Ajit Singh, who was alleged to have taken money and then abstained, could be.
Take a bribe and keep the bargain: protected. Take a bribe and break it: prosecuted.
The law rewarded the more corrupt of two corrupt acts, and it did so for twenty-six years.
The immunity also covered only the takers, not the givers. That is why the case against the Prime Minister continued when the cases against the members who allegedly took the money did not.
8.2 — Convicted, then acquitted
In 2000 a special CBI court convicted P. V. Narasimha Rao and Buta Singh. Convicted — the first time in Indian history a former Prime Minister was convicted of a criminal offence. Rao was sentenced to three years’ rigorous imprisonment and a fine of one lakh rupees. Three other accused, including Ajit Singh, were acquitted.
He appealed and stayed on bail. In 2002 the Delhi High Court set the convictions aside, principally on the credibility of the approver’s evidence. Court-found, in his favour.
He died in 2004 with no conviction standing against him. His body was not allowed inside the Congress headquarters. In February 2024 he was awarded the Bharat Ratna.
8.3 — And then the law was struck down
In 2012, Sita Soren — a member of the Jharkhand assembly, and Shibu Soren’s daughter-in-law — was accused of taking a bribe to vote for a candidate in a Rajya Sabha election. Because the ballot was open, it emerged that she had voted for her own party’s candidate instead. She relied on the 1998 judgment.
The case went up. A two-judge bench referred it to three; three referred it to five; five held that only a bench larger than the one that decided the 1998 case could reconsider it, and referred it to seven.
On the fourth of March 2024, seven judges of the Supreme Court, unanimously, overruled P. V. Narasimha Rao v. State.
Bench: Chief Justice D. Y. Chandrachud, with Bopanna, Sundresh, P. S. Narasimha, Pardiwala, Sanjay Kumar and Manoj Misra JJ. Unanimous.
Bribery is not part of legislative activity. Articles 105(2) and 194(2) do not protect it. Legislators are public servants for the purposes of the Prevention of Corruption Act. The offence is complete when the bribe is accepted, whether or not the vote is ever cast.
The Court noted the absurdity directly: the earlier reading meant a person who took a bribe and did not vote could be prosecuted, while one who took a bribe and voted could not.
The arc of this single case is worth setting out in one place, because Part One promised that where a charge moves up and down the scale over time, the whole movement would be given.
| Date | What happened | Grade at that moment |
|---|---|---|
| July 1993 | Motion survived 265–251 | — |
| 1993–96 | Allegation public; no investigation while in office | Alleged |
| 1998 | Supreme Court, 3:2 — takers immune, givers not | Court-found |
| 2000 | Rao and Buta Singh convicted by special court | Convicted |
| 2002 | Delhi High Court acquits both | Court-found, for the defence |
| 4 March 2024 | Seven judges overrule the 1998 immunity ruling | The law behind the 2002 acquittal is gone |
The acquittal stands. Nothing in 2024 reopens it or touches Rao’s record. What changed is that the reasoning which made the acquittal possible is no longer the law of India.
8.4 — What this case is actually evidence of
A reader could take two opposite lessons from this and both would be wrong.
The first wrong lesson is that Rao was guilty and got off. The 2002 acquittal is a real acquittal by a real court on the evidence, and this book does not treat an acquittal as a conviction with extra steps. Part Twelve refused to do that to Rajiv Gandhi over Bofors and the same rule protects Rao here.
The second wrong lesson is that the system worked. It took thirty-two years to produce a rule that a first-year law student would have written on day one.
What the case is genuinely evidence of is something narrower and more useful. In July 1993, the price of a government surviving a vote was that somebody went looking for members to buy. Whether anybody bought them was never established. What was established, in 1998 and for twenty-six years afterwards, is that if they had, the law protected them.
The defence says: he was acquitted. A High Court examined the evidence and found the approver not credible. The 2024 judgment changed the law prospectively and says nothing whatever about whether this particular man did this particular thing. To bring 2024 into a chapter about 1993 is to use a change in the law as a substitute for evidence that never existed.
That is correct and this book accepts it. The section above does not say Rao took or gave a bribe, and Chapter Ten lists it among the things not established.
What the 2024 judgment is doing in this chapter is different. It is evidence about the state of Indian law between 1998 and 2024 — a period in which the Supreme Court’s settled position was that a legislator who sold his vote and delivered could not be touched. That is a fact about the country, not about the accused, and this series is about what happened to the country.
8.5 — How it ended
The vote of July 1993 bought the government three more years. What it did not buy was a party.
Through 1994 and 1995 the Congress split repeatedly. In May 1995 Arjun Singh and N. D. Tiwari, two of its most senior figures, left and formed their own party. Others followed at state level. Seven cabinet ministers resigned over the course of the term.
In the general election of April and May 1996 the Congress won 140 seats — its worst result since 1947, and eighty-four fewer than in 1991. Rao left office on the sixteenth of May 1996 and resigned as party president that September.
A government that had rescued the country from default, put village councils into the Constitution and doubled the growth rate was reduced to fewer than one seat in four.
No party has ever been punished at the ballot box for a set of economic reforms that worked. This one was punished for everything else in this part, and the reforms were not enough to offset it.
That is worth holding on to when Chapter Nine asks whether a list of failures can be a judgement on a government. In 1996 the voters made exactly that judgement, with far less information than you now have, and they were not obviously wrong.
What happened to the man afterwards has become a small story in its own right, and it belongs here because it says something about the party rather than about him.
When Rao died in 2004, with a Congress government back in office, his body was not taken inside the party’s Delhi headquarters. He was cremated in Hyderabad rather than in the capital. In February 2024 he was awarded the Bharat Ratna, India’s highest civilian honour, by a government of a different party.
The defence says: the treatment of Rao after 1996 shows that the Congress itself disowned this period, which means the party should not be held to it as though it were a continuous actor with a settled character.
The reply is the founding assumption of this whole series. Part One’s opening box conceded that treating Congress as one continuous thing across seventy-nine years is a choice made before the analysis begins, and that it cuts both ways: it lets this book inherit every sin across eight decades, and it lets the defence disown any of them at will.
Part Fourteen is where that cuts hardest, in both directions at once. If the party may disown the demolition because it disowned Rao, it must also give up the 1991 reforms — the largest credit item in nineteen parts. It has spent thirty years trying to keep the second while shedding the first, and it cannot have both.
Both the defence and the prosecution can therefore leave this chapter with something, which is unusual in this book and is the honest result.
Chapter Eight therefore ends where the government did, and the next chapter takes up its defence.
July 1993: the government survived by fourteen votes. Members were alleged to have been paid. Nothing was investigated until the government fell.
Then: alleged → the Supreme Court holds bribe-takers immune if they vote as promised (1998) → convicted, the first former Prime Minister ever (2000) → acquitted (2002) → the immunity ruling itself struck down by seven judges (2024).
Thirty-two years, four reversals, and the man at the centre of it died in the middle. This is what Part One meant by giving the whole arc.
9What Would Have Prevented It
This chapter does two things. It states the defence of this government at its full strength, and then it asks what rule would have changed each outcome — because a failure with no available remedy is a tragedy, and a failure with an obvious unmade remedy is a choice.
9.1 — The defence, at full strength
One. The government had no majority and that is not an excuse, it is a fact. Two hundred and thirty-two seats out of five hundred and forty-five. It could have fallen in any week. A government in that position that had sent paramilitary forces into Uttar Pradesh against the wishes of an elected state government, days before a gathering of a hundred and fifty thousand people, would have been risking a massacre and its own survival at once. This is not a hypothetical: it is what the Liberhan Commission itself accepted.
Two. The alternative the prosecution wants was the abuse this book condemns. Dismissing an elected state government pre-emptively, on the basis of intelligence about what it might allow, is precisely the use of Article 356 that Part Three condemned in Kerala in 1959 and that Part Eight and Part Nine traced through the following decades. A government that had done it would be facing a different chapter of this same series, and the prosecution would have written it.
Three. It did the biggest right thing in the book. Chapter Two. It also passed the Seventy-third and Seventy-fourth Amendments, which a government with four hundred and fourteen seats had failed to pass. It ended the Hazratbal siege without firing into a shrine. It normalised relations with Israel and turned Indian policy east.
Four. The Supreme Court agreed with it about Article 356. Chapter Six. Nine judges upheld the December 1992 dismissals in the very judgment that ended the abuse of the power.
Five. Every accusation against the Prime Minister personally failed. St Kitts: acquitted. Lakhubhai Pathak: acquitted. JMM: convicted and then acquitted on appeal. Hawala: never charged. The suitcase: never tested. A man against whom four separate prosecutions were brought and none survived is entitled to have that stated in one place, and this is the place.
The test of whether a defence has been stated fairly is whether its own side would recognise it. Two independent checks apply here.
The Liberhan Commission — appointed by a Congress government but reporting under a different one, seventeen years later, into a subject where nobody had an interest in protecting Rao by then — reached the minority-government conclusion on its own.
And the Supreme Court, in a case brought against the government by the parties it had dismissed, decided for the government on the central constitutional question. Neither of those bodies was doing Congress a favour in 2009 or in 1994.
Six. It did not start this. The strongest point in the defence is the one it is least able to make in public, because making it means indicting its own party.
Part Twelve set out the sequence. In February 1986 a district judge ordered the locks at the disputed site removed, and they were broken open within about forty minutes with Doordarshan present and broadcasting. In November 1989 a Congress government permitted the foundation-laying ceremony at the site, and launched its own general election campaign from Faizabad. Part Twelve graded the claim that Rajiv Gandhi personally directed the unlocking as Alleged and declined to assert it — but the shilanyas and the campaign launch are not disputed by anybody.
By June 1991 the movement was six years old, had already brought down one national government, and had just won Uttar Pradesh. The government of 1991 did not create that. The government of 1986 and 1989 helped to create it, and it was the same party.
Between the locks being opened in February 1986 and the mosque coming down in December 1992 there are six years and ten months.
For four of those years the central government was Congress. For the remaining three it was not.
This defence therefore succeeds and fails at the same moment. It is true that the government of 1991 inherited the problem. It is also true that this book treats Congress as one continuous actor, and on that basis the party inherited the problem from itself.
9.2 — Where the defence loses anyway
On the demolition it loses on one point and only one, but the point is not answerable.
The government had a third option that required neither troops nor a dismissal. It could have gone back to the Supreme Court, before the sixth of December, and said: we do not accept the state’s affidavit; here is our intelligence; we ask for directions. That costs no political capital. It risks no confrontation. It does not require a majority. It requires a law officer and an afternoon.
It was not done, and no defence has ever explained why not.
On the Liberhan Commission the defence does not lose — it is disqualified. An exoneration by a body that was not permitted to examine you is not an exoneration.
On the money, the defence of individual acquittals is sound and the defence of the delays is not, because the delays are not a matter of guilt or innocence at all.
9.3 — The rules that would have changed it
Five, all of them ordinary legislation. None requires a constitutional amendment. None has been enacted in the thirty-four years since.
| The failure | The rule that would have caught it | Status |
|---|---|---|
| Terms of reference excluding the government that writes them | Commissions of Inquiry Act amended so terms of reference are laid before Parliament and may be enlarged by the commission itself on stated grounds | Never proposed |
| An inquiry running seventeen years on rolling extensions | Statutory outer limit; each extension approved by Parliament, not the ministry | Never enacted |
| A state affidavit accepted without verification | A standing duty on the Union to place its own contrary intelligence before the court where it holds any | Never enacted |
| Fifteen and twenty-three year prosecutions | Time limits for corruption trials involving public office, with the case failing against the state rather than the accused if missed | Never enacted |
| Evidence sitting five years before a charge sheet | Statutory time limit from seizure to decision on prosecution, reported to Parliament | Partly addressed by Vineet Narain directions, 1997; never legislated |
Part Nine, Part Eleven and Part Thirteen each ended with a similar table. In every case the remedies were ordinary law and in every case they were not made. Part Nineteen consolidates all of them.
The fifth row is the interesting one, because something did happen there — and what happened was that a court did it, not a legislature. The Vineet Narain directions of 1997 gave the CBI a protected tenure and abolished the Single Directive. That was the Supreme Court legislating because nobody else would.
9.4 — The question turned on this book
This part has a problem that the previous thirteen did not have, and it would be dishonest to reach Chapter Ten without naming it.
This series assumes that a list of what a government did wrong is an account of that government. Part One’s final Hidden Assumption box already conceded that there is no denominator — no measure of everything a government did, against which the failures could be weighed.
Part Fourteen is where the absence of a denominator stops being a technical point and becomes a real problem, because this government did something enormous and good.
Here is the difficulty, stated plainly. How many deaths in the riots of December 1992 are cancelled by taking eight hundred thousand women into elected office? How much does an economy growing at six per cent instead of three and a half offset a mosque? There is no exchange rate. There is not even a currency.
Every reader who has got this far has been performing this calculation privately for several chapters, and every reader is getting a different answer, and none of them can show their working — because there is no working to show.
What this book can honestly claim is narrower than it looks: that each item in it is true, graded, and stated with its defence. What it cannot claim, and has never claimed, is that the list adds up to a verdict on a government. It does not. It cannot. And a reader who takes it as one has been misled by the form of the thing rather than by anything written in it.
That is the most uncomfortable paragraph in this series so far, and it belongs here rather than at the end, where it would look like a disclaimer bolted on.
The defence is strong: no majority, a Supreme Court that agreed about Article 356, four failed prosecutions of the Prime Minister, and the largest economic reform in Indian history.
It loses on one thing. The government could have gone back to the Supreme Court and said it did not believe the affidavit. That required no troops, no dismissal and no majority. It has never been explained.
And the deepest problem in this chapter is not about that government. It is that this book has no way of weighing what it did right against what it did wrong, and neither do you.
10An Honest List of What We Do Not Know
Every part of this series ends here. The list is not a formality and it is not modesty. It exists because a document that claims to know everything is a document that can be destroyed by one thing it got wrong.
10.1 — Not established, and stated as such
Whether anybody ordered the demolition. Not established. Not by the CBI, not by the Liberhan Commission in seventeen years, not by a criminal trial that ran twenty-eight years. This book does not assert it against anybody, of any party.
Why the central government relied on the affidavit. This is the single most important unknown in the part. The decision was taken by somebody, on some advice, for some reason. No account of that decision exists in public. The commission that would have asked was not allowed to.
Whether the Prime Minister took money in any of the five financial matters. Not established in any of them. Two prosecutions ended in acquittal, one in a conviction that was set aside, one was never charged, one was never tested.
Who burned Charar-e-Sharief. Genuinely disputed. The government says militants, local residents say the army, and a senior civil servant who was there inclines to the second. Alleged, both ways.
How much of the 1990s growth was caused by the 1991 reforms. No honest number exists. Growth accelerated. So did growth in several countries that did nothing similar. Part Four declined to put a figure on the counterfactual cost of the licence system and this part declines to put one on the benefit of removing it, for the same reason.
10.2 — The question this part cannot settle
The largest unknown here is not about the demolition. It is about Chapter Two, and it is the reason this part cannot produce a verdict on the government as a whole.
Growth accelerated after 1991. That is not in doubt. What is in doubt is how much of it the reforms caused. Several things happened at once: the licence system went, the rupee was devalued, the world economy grew through the 1990s, a large young population entered the workforce, and the software industry appeared for reasons that had little to do with Indian policy.
Part Four refused to put a number on what the licence system cost, and said so in its Chapter Ten. This part refuses to put one on what removing it gained, for exactly the same reason: no honest number exists, and one could be produced only by choosing a comparison country and hoping nobody checks the choice.
The comparison that would answer this has never been made properly, and it could be.
Indian industries were deregulated at different times over the following decade rather than all at once. That produces something close to a natural experiment: sectors freed in 1991, sectors freed in 1993, sectors freed later, and sectors not freed at all — the right-hand column of Table 2.
Comparing growth, employment and productivity across those groups, controlling for what else changed, would give a real estimate of what deregulation did. The data exists. The Annual Survey of Industries has been collected throughout.
Part Five ended by naming the one calculation that would settle freight equalisation, which nobody has done. This part ends by naming a second one. Both are arithmetic. Neither requires a document anybody is hiding.
10.3 — Where this part may be wrong
Three specific vulnerabilities, stated so that a hostile reader does not have the pleasure of finding them.
The status of the criminal appeal. The special CBI court acquitted all thirty-two accused in the demolition case on the thirtieth of September 2020. The CBI did not appeal. An appeal was brought instead by two residents of Ayodhya, filed in January 2021 and converted from a revision into a criminal appeal. This book has not been able to establish what has happened to it since 2022, and says so rather than guessing. Nothing in this part depends on that appeal, because this part does not charge anybody with conspiracy.
The seat count. Chapter One explains why this book uses 232 rather than 244. A reader who has seen the other figure is entitled to think this one was chosen because it flatters the argument in the front-matter box. It was chosen because it is the number the government had on the day it took office, and the box would survive the other figure too — 244 is still not a majority.
The placement of Chapter Two. Putting the achievement second, before the charge, is a choice this book made and defended in the text. A hostile reader may reasonably say it is a device: the credit is banked early so that the prosecution can proceed with clean hands. That reading is available and this book cannot disprove it.
10.4 — The document that would settle it
Six parts of this series have ended by naming a file the state holds and will not release. Part Thirteen broke the run deliberately, ending instead on the observation that in the tribal belt there was no closed door at all — the numbers existed everywhere and the addition had simply never been done.
This part returns to the door, and the document behind it is a small one.
Somewhere between the twenty-seventh of November and the sixth of December 1992 there is a file. It contains the Intelligence Bureau’s assessments of the risk at Ayodhya, the Home Ministry’s notes on them, and whatever was written down about whether to deploy central forces or to seek President’s Rule in Uttar Pradesh.
It is not a cabinet paper of great sensitivity. It is nine days of routine internal correspondence in a ministry, now thirty-four years old.
It has never been released, and no government of any party has released it. Classification in India, as Part Six established, has no automatic expiry — so silence is the default and requires nobody to decide anything.
That file would answer the central question of this part in an afternoon. Its absence is why Chapter Four had to build its charge out of an affidavit and a memoir instead.
One closing observation about that, and it is the least comfortable thing in this part.
The Liberhan Commission was not allowed to look at the central government. But it also, in its own words, found that the intelligence agencies had either been over-optimistic or had withheld material from it. So even the inquiry that was forbidden to ask the question was separately unable to get the documents that would have answered it.
Two locks on one door, fitted by different people, neither of whom had to coordinate with the other. That is how this works. It is not a conspiracy. It is a building designed, over decades, by people mostly acting reasonably, in which certain rooms cannot be entered.
Nobody knows who ordered the demolition and nobody is going to. This book does not claim to.
What is missing is much smaller: nine days of Home Ministry file notes from 1992. Not a state secret. Not an operational matter. Nine days of correspondence about whether to send police somewhere.
It has never been released, by any government, of any party, in thirty-four years. The inquiry that would have asked for it was not permitted to, and the agencies did not give it what it did ask for.
Sources & further reading — Part 14
The State Ledger
Part Five committed every subsequent part to carrying this table: which states were affected by the material in this part, and how. It accumulates across the series and consolidates in Part Nineteen.
| State | What was done, 1991–1996 | Standing |
|---|---|---|
| Uttar Pradesh | The demolition and the affidavit that preceded it. President’s Rule from December 1992; elections November 1993; President’s Rule again October 1995. | The central case of this part |
| Madhya Pradesh | Government dismissed 15 December 1992 under Article 356. Dismissal upheld in Bommai. | Lawful, per nine judges |
| Rajasthan | Government dismissed 15 December 1992. Upheld. | Lawful, per nine judges |
| Himachal Pradesh | Government dismissed 15 December 1992. Upheld. Separately, the telecom convictions of Sukh Ram, its most senior national figure. | Both columns |
| Karnataka | Not a Part Fourteen event, but resolved here: the 1989 dismissal of the Bommai government, under a Congress Centre, held unconstitutional in 1994. | Charge, referred back to Part Twelve |
| Jammu and Kashmir | Hazratbal siege ended without assault, October 1993 — credit. Bijbehara firing, 22 October 1993, ~50 dead. Charar-e-Sharief destroyed May 1995, attribution disputed. Under President’s Rule throughout. | Both columns |
| Maharashtra | The securities scam ran through the Bombay market. The December 1992–January 1993 violence is covered in Part Eleven and is not repeated here. | Cross-reference |
| Andhra Pradesh | The urea prosecutions, including the conviction of a relative of the Prime Minister in 2018. | Charge |
| Every state | The Seventy-third and Seventy-fourth Amendments, in force April and June 1993. One third of local seats reserved for women. | The largest credit item in the ledger |
Timeline
| Date | Event |
|---|---|
| Aug 1990 | Mandal reservations announced by the V. P. Singh government. Violent protests follow. |
| Sept–Oct 1990 | Rath yatra across northern India. Advani arrested in Bihar; that government falls. |
| 21 May 1991 | Rajiv Gandhi assassinated. Part Twelve ends. |
| 21 June 1991 | Rao sworn in. Congress 232 seats; no majority. |
| July 1991 | Rupee devalued. 47 tonnes of gold to the Bank of England. Industrial licensing abolished for most sectors. |
| Oct 1991 | UP government acquires ~2.5 acres around the disputed structure. |
| 19 Feb 1992 | Punjab finally polls, postponed from 1991. Turnout around 20%. |
| 23 April 1992 | Sucheta Dalal publishes the securities scam. |
| 16 Nov 1992 | Supreme Court upholds the 27% OBC reservation in Indra Sawhney — three weeks before the demolition. |
| 27 November 1992 | Uttar Pradesh affidavit to the Supreme Court: central force not necessary. |
| 6 December 1992 | Babri Masjid demolished. Riots follow; around 2,000 dead nationally. |
| 15 December 1992 | President’s Rule in Madhya Pradesh, Rajasthan, Himachal Pradesh. |
| 16 December 1992 | Liberhan Commission appointed. Told to report in three months. |
| 22–23 December 1992 | 73rd and 74th Amendments passed by both Houses. |
| 24 April / 1 June 1993 | The two amendments come into force. |
| July 1993 | No-confidence motion survived 265–251. |
| Oct 1993 | Hazratbal siege resolved without assault. Bijbehara firing, 22 October. |
| 11 March 1994 | S. R. Bommai: Article 356 made reviewable; December 1992 dismissals upheld. |
| May 1995 | Charar-e-Sharief destroyed. Arjun Singh and N. D. Tiwari leave Congress. |
| 1996 | CBI raids Sukh Ram; ~₹3.6 crore recovered. Urea case registered 19 May. Congress reduced to 140 seats; Rao leaves office 16 May. |
| 18 Dec 1997 | Vineet Narain: 26 directions; CVC given supervisory role; Single Directive struck down. |
| 1998 | Supreme Court, 3:2, holds bribe-taking MPs immune if they vote as promised. |
| 2000 / 2002 | Rao convicted; acquitted on appeal. |
| 2002 / 2011 | Sukh Ram convicted, twice. Nine days in custody in 2011. |
| 30 June 2009 | Liberhan reports. 17 years, 48 extensions. |
| 30 Sept 2020 | All 32 accused acquitted in the demolition case. CBI does not appeal. |
| 2018 / 2021 | Urea convictions, 23 years on. Related closure report rejected after 22 years. |
| 24 Nov 2009 | Liberhan report tabled in Parliament with an action-taken memorandum. Nothing follows from it. |
| Feb 2024 | Rao posthumously awarded the Bharat Ratna. |
| 4 March 2024 | Sita Soren: seven judges overrule the 1998 immunity ruling. |
Words Used in This Part
| Affidavit | A written statement given to a court and sworn to be true. Lying in one is a criminal offence. |
| Approver | An accused person who admits the offence and gives evidence against the others, usually for a pardon. Treated cautiously by courts. |
| Article 105(2) | The clause protecting MPs from court action for anything said or any vote given in Parliament. Article 194(2) does the same for state legislators. |
| Article 356 | The power to dismiss an elected state government and rule the state from Delhi. See Part Three, where it was first turned into an instrument. |
| Bank guarantee | A promise by the seller’s bank to repay the buyer if the seller takes the money and does not deliver. |
| Bank receipt | A paper issued by one bank to another confirming it holds government bonds on the other’s behalf. The instrument at the centre of the 1992 scam. |
| Basic structure | The doctrine that some features of the Constitution cannot be amended away. Established in Kesavananda Bharati (1973); see Part Eight. |
| Commission of inquiry | A body appointed by government under the Commissions of Inquiry Act 1952 to investigate a matter. Confined to its terms of reference. |
| Foreign exchange reserves | The store of foreign money a country keeps to pay for imports. India’s covered two to three weeks in mid-1991. |
| Industrial licensing | The requirement to obtain permission from Delhi before building or expanding a factory. Abolished for most sectors in July 1991. |
| No-confidence motion | A vote on whether the government should continue. If lost, the government must resign. |
| Panchayati raj | Government by elected village councils. Given constitutional protection by the 73rd Amendment, in force 24 April 1993. |
| President’s Rule | Direct rule of a state by the Centre under Article 356. |
| Ready forward deal | A short-term loan between banks secured on government bonds, sold and bought back at a higher price. |
| Single Directive | The rule requiring political permission before senior officials could be investigated. Struck down in Vineet Narain (1997). |
| Shilanyas | A foundation-laying ceremony. The one at Ayodhya in November 1989 is described in Part Twelve. |
| Mandal | Shorthand for reservation of government jobs for the Other Backward Classes, after the commission chaired by B. P. Mandal. |
| Mandir | Temple. Shorthand for the Ayodhya movement. With Mandal, it gives the 1991 election its usual name. |
| Rath yatra | A procession on a vehicle got up as a chariot. The 1990 journey across northern India campaigning for a temple at Ayodhya. |
| Joint Parliamentary Committee | A committee of both Houses set up to investigate a specific matter. Its report binds nobody. |
| Eleventh Schedule | The list of twenty-nine subjects added to the Constitution in 1992 that village councils may be given. Which of them a council actually gets is decided by the state. |
| Terms of reference | The written instructions creating an inquiry and limiting what it may examine. Drafted by the government. |
A Word on Sources
Everything dated in this part was checked against a source rather than recalled, and where two sources disagree the disagreement is on the page rather than resolved silently.
The strongest material here is documentary and most of it belongs to the state. The Uttar Pradesh affidavit of 27 November 1992 is a court filing. The Liberhan Commission’s terms of reference and its report of 30 June 2009 are published. S. R. Bommai (11 March 1994), Vineet Narain (18 December 1997), P. V. Narasimha Rao (1998) and Sita Soren (4 March 2024) are reported judgments. Convictions and acquittals are court records.
Two sources are memoirs and are treated as such. K. K. Venugopal’s account of the pre-demolition proceedings is used for what the court was shown, which is a fact about a hearing he attended. Wajahat Habibullah’s account of Hazratbal is used the same way, and his contrary view on Charar-e-Sharief is given as a view, not a finding. Where this part rests on a memoir it says so in the text.
Three things are graded Alleged and nothing in this part depends on them: the suitcase, the entries said to relate to the Prime Minister in the Jain diaries, and the attribution of the Charar-e-Sharief fire.
The status of the criminal appeal against the 2020 acquittals could not be established beyond 2022 and Chapter Ten says so. That is a gap in this part and it is the reader’s to know about.
A note from the author
Thirteen parts of this series have been easier to write than this one, and I want to say why rather than pretend otherwise.
I set out to write the case for the prosecution. That is what this book is, it says so on the cover, and I have not softened it. But a prosecutor who reaches a chapter where the defendant has plainly done something large and good has a choice to make, and there are only two honest ways to take it. He can leave it out, which is fraud. Or he can put it in at full length, in the strongest form he can manage, and accept that his own document is now harder to read as a verdict.
I took the second, and Chapter Two is placed second for that reason. I want to be exact about where that could have pulled the writing, because the rule I set myself in Part One was that the position gets declared at the point where it could distort, not in a disclaimer at the front.
Two places. The first is Chapter Two itself. I found, drafting it, that I wanted to shorten it — to give the reforms three pages and move on to the demolition, where my case is strong. That instinct was about the shape of my argument, not about 1991, and I have written the chapter at the length the material deserves instead. Part Twelve’s author’s note recorded the same pull about Rajiv Gandhi’s credit chapter, and I resisted it there too. Twice is a pattern in me, not in the evidence.
The second is Chapter Six. I have spent four parts of this series building a case against Article 356, and I reached a chapter where a Congress government used it and nine judges of the Supreme Court said it was entitled to. My first draft of that chapter buried the concession in a subordinate clause. It is now a Verdict box with a heading. The rest of the book is worth more with it there than without it, and if I am honest, that is the argument that persuaded me rather than any purer one.
One further thing. Part Ten declared that I am Punjabi and had written on that material before. Nothing in this part touches that. But readers of the earlier parts should know that I came to the Rao years expecting to dislike them and did not — and that this is the only part of nineteen where I have changed my own mind while writing, rather than before starting.
What I have not done is weigh it up. Chapter Nine’s last box says there is no exchange rate between eight hundred thousand women in elected office and two thousand people dead in December 1992, and I meant it. I do not have one. If you finish this part with a settled verdict on this government, it did not come from me, and you should look carefully at where it did come from.
Lovepreet Singh
Part Fourteen of The Congress Record
The Living Archive · misterlove.in
Part Fifteen takes up the terror wave, 26/11, cash-for-votes and the nuclear deal under UPA I, 2004–2009. The comparison with the government of 2014 onwards is Part Eighteen, on the same ruler, and the second series auditing that government on this scale is still promised.