Where We Left Off
Chapter Seven is the longest credit chapter in this series. Chapter Four rests on a claim that a named witness inside the Prime Minister's Office denies, and it is built so that the charge does not depend on it.
Part Eight asked what happens when a party becomes an instrument of one office. This part is the test of that finding, because the office changed hands.
Four things carry into this part.
Part Eight’s central finding. The 1969 split transferred control of candidate selection, and after 1971 every Congress member of parliament owed his seat to one office. Part Eight said the check that could remove a leader had gone from unusable to unimaginable. Chapter Eight of this part asks whether that survived the person, and the answer decides whether Part Eight was a biography or a structural finding.
The election of December 1984. Part Eleven established what had happened in Delhi weeks earlier and what followed for forty-one years. This part begins with the election held in its immediate aftermath, and it does not re-argue Part Eleven’s material.
A warning is received only where somebody has the authority to act, an incentive to believe, and a decision it changes. Part Six’s test, applied to a road in 1958. Chapter Three applies it to a letter written by a retired Supreme Court judge in February 1986, and Chapter Six to the professional advice before a deployment in 1987.
The military content is the part that gets overruled. Part Six found it in the Forward Policy of 1961 and Part Ten found it at Amritsar in 1984. Chapter Six is the third instance in five years, and by then it is not a coincidence.
One note on the shape of this part. It contains the longest chapter of credit items in the series, and that is not balance for its own sake. The five years covered here produced a set of things that worked and are still working, and a book that recorded only what went wrong would be describing a different government from the one that existed.
How to read this book
The Six Boxes
Six coloured boxes run through the series, each doing one job. Here they are with examples from the material ahead.
A Word Box explains a hard word the moment it first appears.
Defection: a member of a legislature leaving the party on whose ticket they were elected, usually to join another, and usually in exchange for something.
It was endemic in Indian state politics from the 1960s. Governments fell and were formed in weeks, and a phrase entered the language — Aaya Ram, Gaya Ram, “Ram came, Ram went” — from a legislator who changed party three times in a fortnight.
Why it matters here: the law passed to stop it is one of the two most consequential things this government did to the Indian constitution, and Chapter Two argues that what it stopped and what it created are different things.
An In Real Terms box turns a number too big to picture into something with a body.
Four hundred and fourteen seats out of 514 contested is about four in every five.
It is more than Nehru ever won. More than Indira Gandhi won in 1971 after a war. It remains the largest majority any party has held in the Indian Parliament, and no party has come within sixty seats of it since.
With it, a government can amend the constitution at will, pass any law it chooses, and lose every opposition vote in both houses without noticing.
The question this part asks is what it was used for, and the answer is that a great deal of it was used to reverse a Supreme Court judgment about one woman’s maintenance.
A How We Actually Know This box shows the physical evidence and then says what it cannot prove.
This part divides sharply between what is documentary and what is testimonial, and the division falls in one place.
Documentary and solid: the election result, the Tenth Schedule, the Shah Bano judgment and the Act that answered it, the date and text of the order unlocking the gates at Ayodhya, the shilanyas of November 1989, the Indo-Sri Lanka accord, every court decision in the Bofors matter, and the whole of Chapter Seven’s list of what was built.
Testimonial and contested: who decided the unlocking, and why. There are several accounts, they disagree, and one of them comes from an official who worked in the Prime Minister’s Office and states that the Prime Minister did not know.
Chapter Four is built so that its charge does not rest on resolving that. What it rests on is 1989, which is not in dispute by anybody.
An Argument box appears where serious people disagree, each side at its strongest.
The framing question. Both answers are held by people who lived through it.
Four hundred and fourteen seats is the capacity to do anything, and it was spent on two decisions in one month of 1986 whose consequences ran for the next thirty-eight years, on a deployment that killed over a thousand Indian soldiers and produced nothing, and on a defence procurement scandal that destroyed the government’s own authority. The reforms that mattered — telecommunications, computers, education — did not need 414 seats. They needed forty.
Count what still exists. A telecommunications system that reached villages. A computer policy that produced an industry employing millions. The voting age lowered to eighteen. Three peace accords in two years, of which one ended a twenty-year insurgency that has not restarted. An environment statute after Bhopal. Panchayati Raj proposed and, though defeated, enacted three years later on his draft. That is a larger list of surviving institutions than most five-year governments anywhere produce.
What would settle it: the cabinet records of 1986 and 1987, closed.
Why people care so much: because this is the government that the argument about dynastic politics in India is mostly conducted about, and both sides need it to have been either a promise betrayed or a promise that was never there.
A Hidden Assumption box digs out a belief sitting underneath an argument that both sides accept without discussing.
Everybody who writes about these five years assumes that a large majority is power.
It is in every account. He had the numbers and squandered them. He had a mandate and lacked the will. Both sides treat 414 seats as a quantity of capacity that was then spent well or badly.
A majority is the capacity to legislate. It is not the capacity to decide, and the two are different things that this part keeps separating.
Deciding well requires something a majority does not supply: a process by which a proposal is tested before it is adopted. Somebody who can say the thing will not work, whose objection has to be answered, and who survives making it. Part Three found that the Congress party stopped having such people by 1954. Part Eight found that after 1969 no Congress member of parliament had a career independent of one office.
So what 414 seats bought was an unlimited ability to enact decisions taken by a very small number of people, quickly, without the proposal being tested by anybody with standing to object. Every chapter from Three to Six describes exactly that: a decision made in weeks, on advice from a handful of advisers, with the professional or judicial objection arriving afterwards and from outside.
Which is why the majority produced so little. It is also why Chapter Seven’s list is what it is — the things that worked were the ones that did not require a contested judgment, only money, authority and somebody competent left alone to build them.
The general point: a legislature that cannot refuse anything is not a source of power to a government. It is the removal of the last body that could have told it no.
And every chapter closes with a Remember This box: the chapter in the plainest words available.
414 of 514 seats — the largest majority any party has held in the Indian Parliament, and none has come within sixty of it since.
The evidence divides cleanly: everything is documentary except who decided the unlocking at Ayodhya and why, where accounts conflict and one comes from inside the Prime Minister’s Office.
A majority is the capacity to legislate, not the capacity to decide. A legislature that cannot refuse anything is the removal of the last body that could have said no.
1Four Hundred and Fourteen
The largest mandate in Indian history was delivered six weeks after the events of Part Eleven, to a man who had been a commercial pilot four years earlier.
1.1 — How he arrived
Rajiv Gandhi was not a politician by intention. He flew for Indian Airlines and had stayed out of public life while his younger brother Sanjay — whose position Part Nine’s Chapter Seven described — occupied the space beside their mother.
Sanjay died in an air crash in June 1980. Rajiv entered Parliament in 1981.
On 31 October 1984 his mother was assassinated. He was sworn in as Prime Minister that evening. Part Eleven covers what happened in Delhi over the following week and this part does not repeat it.
The general election was held in December 1984.
1.2 — The result
Congress won 414 seats, on a vote share of around 49 per cent — the highest either figure has ever reached for any Indian party.
Two things produced it, and both should be said.
Sympathy. The election followed an assassination by six weeks, and the campaign was conducted in its shadow. That is a real and sufficient explanation for a very large part of the result and nobody disputes it.
Expectation. He was forty, had no history in the machine, and was widely understood to represent something different from the decade Part Nine describes. The press called him Mr Clean, which was not a slogan his party invented and which he did not discourage.
What the majority meant in constitutional terms is straightforward. With 414 of 514 seats he could amend the constitution at will, pass any statute, and lose every opposition division in both houses without it registering.
1.3 — What was promised
Part One’s test is a government’s own stated targets, and the stated programme here was reasonably specific.
Modernisation, and specifically technology — telecommunications, computers, and the phrase that attached to the government from the beginning, taking India into the twenty-first century.
Cleaning up public life, including the party. He said so at the party’s centenary session in Bombay in December 1985, in unusually direct language about power brokers within Congress.
Settling the outstanding conflicts. Accords were signed with Punjab in July 1985, with Assam in August 1985 and with Mizoram in June 1986, and Chapter Seven takes them properly.
Devolution to local government, which became a constitutional amendment bill in 1989.
That is a serious programme and it is the standard this part scores against.
Mandate: the authority a government claims from an election result to do particular things.
The word does more work in politics than it can support. An election produces a number of seats. What it does not produce is a list — voters choose a party for reasons that are not itemised and cannot be read back out of a total.
A very large majority is therefore not a very large mandate. It is a very large permission, and the difference matters when the government uses it for something nobody voted about.
Why it matters here: Chapter Three describes 414 seats being used to reverse a Supreme Court judgment that had not been mentioned in the campaign, because the campaign preceded it by four months.
What that number means is easier to see beside the majorities other Indian governments have worked with.
Set 414 against what other Indian governments have had to work with.
Nehru’s largest was 371. Indira Gandhi’s post-war landslide of 1971 was 352. The reforming government of 1991, which dismantled the system Part Four describes, had 244 and no majority at all.
Which is the awkward comparison. The government with the fewest seats in that list changed the direction of the Indian economy. The government with the most spent a substantial part of its authority reversing a maintenance order.
Seats are not the constraint, and they never have been.
Against that arithmetic, here is the programme this part scores.
| What was promised | What arrived |
|---|---|
| Technology and modernisation | Delivered. Telecommunications, computers and software. Chapter Seven. |
| Cleaning up public life | Not delivered. The government left office under the largest corruption allegation in Indian history, unresolved to this day. Chapter Five. |
| Settling the outstanding conflicts | Mixed. Three accords in two years. Mizoram worked completely; Assam partly; Punjab’s central term was never implemented. And a new conflict was entered in Sri Lanka. |
| Devolution to local government | Attempted and defeated, then enacted in 1992 on the same design. Chapter Seven. |
| Not promised, and done | The Tenth Schedule, the Shah Bano reversal, and the decisions of Chapter Four. None was in the 1984 campaign. |
The last row is the one worth sitting with. The four most consequential things this government did to the Indian constitution and to Indian public life were none of them things anybody voted for in December 1984.
414 of 514 seats on about 49 per cent of the vote — the highest either figure has reached for any Indian party. It came six weeks after an assassination, and sympathy explains a great deal of it.
The stated programme was technology and modernisation, cleaning up public life, settling the outstanding conflicts, and devolution to local government. That is what this part scores against.
A large majority is a large permission, not a large mandate. An election produces a number of seats and not a list.
2The Anti- Defection Law
It solved a real problem within months of taking office, and it is the most consequential thing this government did to the Indian constitution. Almost nobody discusses what it did.
2.1 — The problem
Through the 1960s and 1970s Indian state politics was disfigured by defection. Legislators elected on one party’s ticket crossed to another, governments fell within weeks of being formed, and the price of a legislator was widely understood to be a number.
This was not a marginal problem. Part Five’s chapter on the closing of political routes and Part Eight’s on the 1969 split both describe a system in which a state government’s survival depended on holding individual members, and in which the centre’s ability to engineer a change of government — Part Ten’s Chapter Seven describes one in 1984 — rested on exactly this.
Something had to be done about it and previous attempts had failed.
2.2 — What was done
The Constitution (Fifty-Second Amendment) Act, 1985 inserted the Tenth Schedule.
Its rule is simple. A member of either house of Parliament or of a state legislature is disqualified if he voluntarily gives up membership of his party, or if he votes or abstains contrary to any direction issued by his party, unless the party condones it within fifteen days.
Read the second limb again, because that is the whole of this chapter.
The law did not merely stop members changing party. It made it a disqualification for a member to vote against his party’s instruction on anything at all — not only on confidence motions, not only on the budget, but on any question on which the party issues a direction.
An exception was provided for a split involving a third of a party’s members, which produced its own trade and was tightened in 2003 to require a two-thirds merger.
Whip: both a party official and the written direction that official issues telling members how to vote.
Every parliamentary system has them. In Britain a member who defies a three-line whip may lose the party’s support, be denied office, and be deselected at the next election. Those are severe consequences and they are political.
India is the outlier. Since 1985 defying a whip carries a legal consequence: disqualification from the house. The member does not lose his party’s favour. He loses his seat.
Why it matters here: that difference is the entire distance between a legislature with a whip and a legislature that is one.
2.3 — What it created
Consider what a member of parliament has been able to do since 1985.
He may speak against a bill. He may abstain if his party permits it. He may vote as his party directs.
He may not vote according to his own judgment against his party’s instruction, on any subject, at any time, without losing his seat.
Which means that on every division in the Indian Parliament for the last forty years, the outcome has been determined by the decisions of party leaderships and not by the decisions of members. The count is a formality; the whips have already settled it.
That is not a legislature in the sense the word ordinarily carries. It is closer to an electoral college that meets continuously — a body whose function is to register the arithmetic of a general election rather than to deliberate.
And note where the power went. Not to Parliament as against the executive, and not to the executive as against Parliament. It went to whoever controls the party whip, which in a party organised as Part Eight describes means one office.
Imagine you are elected to represent a constituency. A bill comes before the house that will do serious damage there — a plant sited badly, a subsidy withdrawn, a district split.
Your constituents are against it. You are against it. You said so during the campaign.
Your party issues a direction to vote for it.
If you vote as your constituents wish, you cease to be a member of parliament. Not at the next election — immediately, on disqualification.
There is no version of that arrangement in which you represent the constituency. You represent the whip, and the constituency’s only remaining instrument is to elect somebody from a different party at the next election, who will be in exactly the same position.
None of which requires interpretation, because the operative words are in the constitution.
The Tenth Schedule is constitutional text and anybody can read it. The two limbs of the disqualification, the fifteen-day condonation window and the exception for splits are all set out in it.
That makes this chapter’s central claim unusually secure: it is not an inference about how the law has worked in practice but a statement of what it says.
What the text cannot show: how often the second limb has actually been invoked. Disqualification proceedings are decided by the presiding officer of the house, and there is no consolidated public record of directions issued, defiances, or condonations.
That absence matters and cuts both ways. It means nobody can quantify how many members have been disqualified for dissent — and it also means nobody can quantify the far larger number who never dissented because they knew what would follow. The second figure is the effect of the law and it is unmeasurable by design.
Whether that trade was worth making is argued about by two groups who almost never overlap.
It is defended by almost every Indian party and criticised by almost every constitutional scholar, which is an unusual combination.
The problem was real, corrosive and getting worse, and no lesser measure had touched it. Since 1985 Indian governments have been able to govern for their terms instead of buying legislators monthly, and the trade in members that disfigured state politics for two decades largely stopped. Party discipline is not an aberration — every parliamentary democracy has whips, and in Britain defying one is career-ending too. India merely made explicit what other systems achieve by other means.
Britain’s whips end careers; they do not end memberships, and a British MP who defies one remains an MP. That difference is not technical. Every other system leaves a member the ultimate ability to vote his conscience at a cost he chooses to bear, and it is that residual ability which makes a legislature a check on an executive at all. India removed it. And it did so far beyond what the problem required: stopping defection needs a rule about changing parties, not a rule about voting on ordinary legislation.
What would settle it: comparing India’s record on executive accountability with parliamentary systems that discipline without disqualifying. The comparison is not favourable and has not been made systematically.
Why people care so much: because every party in India benefits from it while in office, which is why forty years of criticism from outside politics has produced no amendment.
Both sides of that argument are describing a law about legislators who change parties. It is worth checking whether that is what it is.
Everybody arguing about the Tenth Schedule assumes that defection is the problem it addressed.
Its defenders say it stopped defection. Its critics say it stopped defection at too high a price. Both are describing a law aimed at legislators who change parties, and disagreeing about the collateral damage.
Look at what the two limbs actually cover. The first — voluntarily giving up party membership — is about defection. The second — voting or abstaining against a party direction — is not about defection at all. A member who votes against his party on a single bill and remains in it has defected from nothing.
The second limb is a rule about dissent, and it was included in a law about defection, and it is the limb that operates on every sitting day.
This matters because it changes what the law is for. A rule against defection protects a government from a market in legislators. A rule against dissent protects a party leadership from its own members — which is a different beneficiary, and one that has nothing to do with the instability of the 1960s.
And it explains the puzzle in the argument box: why every Indian party defends a law that constitutional scholars almost uniformly criticise. They are not defending the first limb. Nobody attacks the first limb. They are defending the second, which converts a member of parliament from a person with a vote into a person with a seat, and hands the vote to whoever signs the whip.
Part Eight found that after 1969 no Congress member had a career independent of one office. The Tenth Schedule generalised that to every party in India, permanently, by statute, in 1985 — and it is the single most durable thing this government did.
One thing to carry forward. Every part of this series so far has described a check being removed by somebody with a reason. This is the only occasion on which a check was removed from every party in India at once, by statute, permanently — and it was done by a government that had criticised its own party’s brokers of power three months earlier.
The Tenth Schedule disqualifies a member who leaves his party or who votes against a party direction on anything at all. The second limb is the one that operates every sitting day.
Since 1985, a member of the Indian Parliament cannot vote according to his own judgment against his party without losing his seat. Every division is settled by whips before it is taken.
A rule against defection protects a government. A rule against dissent protects a party leadership from its own members — and that is the limb every party defends.
3Shah Bano
A sixty-two-year-old woman won a maintenance order of a few hundred rupees a month. Four hundred and fourteen seats were used to take it away, and thirty-eight years later the Supreme Court gave it back.
3.1 — The case
Shah Bano Begum was divorced by her husband after more than forty years of marriage. She applied for maintenance under section 125 of the Code of Criminal Procedure — a religion-neutral provision under which any wife unable to maintain herself may claim from her husband.
On 23 April 1985 a five-judge bench of the Supreme Court, presided over by the Chief Justice, held that she was entitled to it. Section 125 applies to all citizens, the Court held, and is not displaced by personal law.
The sums were small. The principle was not: a divorced Muslim woman had the same access to a general maintenance provision as any other Indian woman.
3.2 — What happened next
From August 1985 there was substantial agitation against the judgment, led by conservative Muslim organisations who read it as judicial interference with personal law.
On 15 January 1986, at a conference in Delhi, the Prime Minister announced his intention to legislate to nullify it.
On 25 February 1986 the Muslim Women (Protection of Rights on Divorce) Bill was introduced in the Lok Sabha.
On 28 February 1986 — three days later — V.R. Krishna Iyer, a retired judge of the Supreme Court, wrote to the Prime Minister protesting against the Bill and warning of the consequences the country would face as a result.
Krishna Iyer is not a convenient witness for anybody’s politics. He had been a minister in a Communist government in Kerala before joining the judiciary, and Part Eight’s Chapter Eight records him granting the conditional stay that allowed Indira Gandhi to remain Prime Minister in June 1975. He was the most prominent progressive jurist in India.
The Bill became law in May 1986.
Personal law: the body of rules governing marriage, divorce, inheritance and adoption, which in India differs by religious community.
India has no uniform civil code. Hindus, Muslims, Christians and Parsis are governed by different rules in these matters, and that was a deliberate constitutional choice — Article 44 sets a uniform code as a directive principle, which is an aspiration rather than an obligation.
The important distinction, and the whole of the argument in this chapter, is that personal law is not the only law that touches a family. Criminal procedure, property statutes and welfare provisions apply to everybody regardless of community.
Why it matters here: section 125 is criminal procedure, not personal law, and whether it counts as an intrusion into personal law is precisely what the 1985 judgment and the 1986 Act disagreed about.
3.3 — What the Act did
It provided that a divorced Muslim woman is entitled to a reasonable and fair provision and maintenance from her former husband within the period of iddat — roughly three months.
Beyond that period, the obligation passed to her relatives, and where they could not maintain her, to the state Waqf Board.
The effect was to remove divorced Muslim women from a general provision of Indian criminal procedure and place them under a separate arrangement, on the ground of their religion.
3.4 — What the courts did with it
Two things, and they are the reason this chapter has an ending.
In 2001, in a challenge brought by Shah Bano’s own lawyer, the Supreme Court read the Act down. It held that the requirement to make provision within the iddat period meant the provision must be made within it, not that it must be limited to it — and that a fair provision extends for the woman’s life until she remarries.
In July 2024, the Supreme Court held that section 125 applies to all women irrespective of religion, and that the 1986 Act does not displace it. The Court said maintenance is a right and not charity.
Which means the position today is the position of 23 April 1985. The Act stands on the statute book and has been construed into consistency with the judgment it was passed to reverse.
Thirty-nine years.
Everything in this chapter is a published document with a date, and that is unusual even for this series.
The judgment of 23 April 1985 is reported. The announcement of 15 January 1986 was made at a public meeting and reported at the time. The Bill’s introduction on 25 February 1986 is in the parliamentary record. The Krishna Iyer letter of 28 February 1986 exists and has been quoted at length. The Act is on the statute book. The 2001 and 2024 judgments are reported.
There is no version of this chapter that depends on anybody’s recollection or on any inference about motive.
What it cannot show: what the political calculation was. That the government believed it was losing Muslim support is asserted by every account and denied by none, but the deliberations are not on the record — and Chapter Four turns on exactly that gap.
What the Act meant for the woman it was named after is worth stating before anybody argues about principle.
The maintenance Shah Bano had been awarded was a few hundred rupees a month. She was sixty-two, had been married more than forty years, and had no other income.
What the 1986 Act did was tell a woman in that position that after roughly three months her former husband owed her nothing, that she should apply to her relatives, and that if they could not support her she should apply to a religious endowment board.
Section 125 exists to stop exactly that sequence. Its purpose is not fairness between spouses; it is to prevent destitution and vagrancy, which is why it sits in a criminal code and not in a family statute.
Four hundred and fourteen seats were used to remove one category of Indian woman from a provision against destitution.
Whether that was nonetheless the right thing for a government to do is genuinely argued, and the defence deserves its strength.
It is defended today by fewer people than defended it in 1986, which is itself worth noting. The defence deserves stating at strength.
India does not have a uniform civil code, by deliberate constitutional choice. Every community is governed by its own personal law in matters of marriage, divorce and succession, and that arrangement is the basis on which minorities accepted the Indian constitutional settlement. A Supreme Court applying a general criminal provision to override Muslim personal law was, on that reading, doing exactly what the settlement excluded — and a government responding to a community’s near-unanimous objection was doing what a democracy does.
Section 125 is not personal law and never was. It is a provision of criminal procedure whose purpose is to prevent destitution and vagrancy, applying to every Indian regardless of faith. Removing one category of citizen from it, on the basis of the religion of her former husband, is not protecting a minority’s personal law — it is creating an exception to a general law that operates against the poorest members of that minority. And the community was not unanimous: the objection was led by clerical organisations, and Muslim women’s organisations opposed the Act.
What would settle it: nothing evidential. The courts have settled the legal question and the political one is still live.
Why people care so much: because this is the origin of the word “appeasement” in modern Indian politics, and the argument about it has been running for thirty-nine years without either side conceding a paragraph.
One last thing about the shape of it, and it is Part Six’s test arriving in a domestic setting. A retired Supreme Court judge of impeccable progressive standing wrote to the Prime Minister three days after the Bill was introduced, warning what it would do. He had authority, he had standing, and he had no interest to serve. The Bill passed in May.
23 April 1985: a five-judge bench held that section 125 of the Code of Criminal Procedure — a religion-neutral destitution provision — applies to a divorced Muslim woman.
15 January 1986 the reversal was announced, 25 February the Bill introduced, 28 February Krishna Iyer wrote warning against it, and in May 1986 it became law. Maintenance was restricted to the iddat period, then passed to relatives, then to the Waqf Board.
The Supreme Court read the Act down in 2001 and in July 2024 held that section 125 applies to all women irrespective of religion. The position today is the position of 23 April 1985 — thirty-nine years later.
4The Locks
A padlock was broken in forty minutes while a television crew filmed it. Who ordered that is genuinely disputed. What happened in 1989 is not.
4.1 — What happened
The disputed structure at Ayodhya had been locked since 1949, when idols were placed inside it and the premises were attached by a magistrate and closed. For thirty-six years that was the position, and it was the position every government had maintained.
On 25 January 1986 a local lawyer, Umesh Chandra Pandey, applied to the Faizabad court for the gates to be opened.
On 1 February 1986 the District Judge, K.M. Pandey, ordered the locks removed so that Hindu worship could take place. His reasoning was that Muslims would not be affected if the locks were opened and the idols allowed to be seen and worshipped, and that the heavens would not fall.
Three things about the execution.
The locks were broken within about forty minutes of the order.
Doordarshan — the state broadcaster — was present and telecast it nationally.
And officials of the central and the Uttar Pradesh governments had testified in the proceedings that they anticipated no law and order problem.
Attachment and status quo: two ways a court freezes a disputed property while it decides who owns it.
In 1949 a magistrate attached the premises at Ayodhya under the provision allowing a disputed property to be taken into custody where a breach of the peace is likely, and locked it. That is why there were locks.
The point of freezing a property is that nobody gains by the passage of time. Whoever wins the case gets it as it was.
Why it matters here: unlocking a property attached to prevent a breach of the peace, so that one party may worship in it, alters the position on the ground while the title suit remains undecided — and the title suit had by then been pending for thirty-six years and would remain pending for another thirty-three.
4.2 — What followed
On 3 February the High Court directed that the nature of the property not be altered pending further orders.
On 14 February 1986 Muslim organisations observed a day of protest. There was rioting in Delhi, in Meerut, elsewhere in Uttar Pradesh, and in Anantnag in Kashmir.
On 15 February 1986 the Babri Masjid Action Committee was formed.
4.3 — Who decided
This is the part that is genuinely contested, and this part will not pretend otherwise.
The standard account holds that the unlocking was a deliberate counterweight to the Shah Bano reversal — a concession to Hindu opinion to offset a concession to Muslim clerics — arranged through the district court with the assistance of the Uttar Pradesh government, and that the Prime Minister’s advisers, principally Arun Nehru and M.L. Fotedar, drove it.
Against that stands an account from Wajahat Habibullah, an official who worked in the Prime Minister’s Office and later wrote about the period. He records that the Prime Minister was not informed, that he asked the Uttar Pradesh Chief Minister to explain, and that he suspected Arun Nehru and Fotedar were responsible and was having it verified.
Those two accounts are not compatible on the central question, and the second comes from inside the office.
Grade: that the unlocking was arranged rather than spontaneous is Documented — a padlock does not come off in forty minutes with a national broadcaster standing by unless somebody arranged for both. That the Prime Minister ordered it is Alleged.
This part therefore does not assert it, and does not need to.
4.4 — 1989, which nobody disputes
Three years later the same government made two decisions that are not contested by anybody.
In November 1989 it permitted the shilanyas — the laying of foundation stones for a temple — on land adjacent to the disputed structure. Assurances had been given that the status quo would be maintained. The ceremony went ahead.
And on 3 November 1989 the Prime Minister launched his party’s general election campaign from Faizabad, next to Ayodhya. The venue had been changed from the one originally planned. In his speech he promised to establish Ram Rajya — a phrase reportedly not in the drafted text and added there.
A senior Congress leader from Uttar Pradesh, Kamlapati Tripathi, warned that this approach would damage the unity and integrity of the country.
So whatever is true about February 1986, the position in November 1989 requires no inference. A Prime Minister chose to open a national election campaign at Ayodhya, on a religious promise, having permitted a foundation ceremony at the site weeks earlier.
That is the charge in this chapter, and it survives every version of the 1986 dispute.
Forty minutes is not long enough to obtain a certified copy of an order in an Indian district court.
It is not long enough to convene the officials who hold the keys, arrange a police presence, or decide how a crowd will be managed.
It is certainly not long enough to summon a national television crew from Delhi.
Which is why the timing is the most informative fact in this chapter, and why it establishes something narrower and firmer than any of the competing accounts of who decided. Somebody knew in advance. The question this chapter cannot answer is who, and the question it does not need to answer is whether it was the Prime Minister.
All of which is why the standard account of 1986 needs handling more carefully than it usually gets.
Almost every account treats Shah Bano and the unlocking as a matched pair. The evidence for that is thinner than the confidence with which it is asserted.
The reversal was announced on 15 January 1986 and the locks came off on 1 February — seventeen days. Government officials testified that no law and order problem was anticipated, which is what a government says when it wants a thing to happen. Doordarshan was present. Multiple senior Congress figures have described the calculation in exactly these terms, and the political logic is obvious: a government that had just been accused of appeasing one community needed something for the other.
There is no document. The order came from a district judge on an application by a private lawyer, and the Allahabad High Court did not disturb it on the merits. The account from inside the Prime Minister’s Office states that the Prime Minister did not know and was asking who had done it. And every source describing the quid pro quo was written afterwards, mostly by people with a position in the subsequent politics — which is precisely the class of evidence this series has refused to rely on in six previous parts.
What would settle it: the Uttar Pradesh government’s file on the February 1986 proceedings, and the Prime Minister’s Office record. Neither is open.
Why people care so much: because if it was ordered from Delhi, the Congress party opened the door that led to December 1992, and if it was not, then the door was opened by a district judge and the Congress party merely failed to close it — which are different degrees of the same responsibility.
Grading all of that carefully is the point of this chapter, so here is the grading.
This chapter has the most uneven evidence in the part and the unevenness is the reason it is written as it is.
Solid: the application of 25 January, the order of 1 February and its reasoning, the interval before the locks came off, the presence of the state broadcaster, the officials’ testimony that no disorder was anticipated, the High Court’s direction of 3 February, and the events of 14 and 15 February. All reported at the time or on the court record.
Solid: the shilanyas of November 1989 and the Faizabad campaign launch of 3 November. Contemporaneous reporting, undisputed.
Alleged: that the unlocking was ordered from Delhi, and that it was a counterweight to the Shah Bano reversal. Asserted in many accounts, denied in one from inside the Prime Minister’s Office.
What none of it can show: what a district judge was thinking. He gave reasons, they are on the record, and whether they were the operative ones is not knowable from a judgment.
Underneath the whole dispute, though, sits a way of describing 1986 that both sides use without examining.
Everybody arguing about 1986 — those who condemn the pairing and those who deny it — assumes that the two decisions were a balance.
One side says: a concession to Muslim clerics was offset by a concession to Hindu organisations, and that was cynical. The other side says: they were not connected, so there was no such trade. Both are arguing about whether a balance was struck.
Notice what the framing accepts. It takes for granted that the government’s task in these matters is to allocate satisfaction between religious communities — that a concession to one creates a debt to the other, and that governing India well means keeping the account level.
That premise is the thing that produced both decisions, and it is not questioned by anybody in the argument.
Test it against what was actually at stake. Shah Bano was a case about whether a destitute divorced woman could claim maintenance under a general provision of criminal procedure. Ayodhya was a dispute about title to a piece of land, pending in a court since 1950. Neither was a question about the relative standing of two communities. Both were converted into one.
And once a question has been converted, the balance framing makes the conversion permanent. A government that answers a legal question with a communal calculation has told every organisation in the country what kind of question to bring next — which is why the years after 1986 produced a great many more of them, from both directions, and almost none of them about maintenance or about title.
Which is why the argument about whether the two were a deliberate pair is less important than it looks. If they were, a government made a calculation. If they were not, a government made the same calculation twice in three weeks without noticing. The premise is the finding, not the coordination.
Which is why this chapter has been written to rest on 1989 rather than on 1986. The later decisions require no inference, no contested memoir and no closed file, and they were taken by the Prime Minister personally.
1 February 1986: a district judge ordered the locks removed. They came off within forty minutes, with Doordarshan present and broadcasting, after government officials had testified that no law and order problem was expected.
Whether Delhi ordered it is disputed, and an official who worked in the Prime Minister’s Office states that the Prime Minister did not know. This part does not assert it.
November 1989 is not disputed. The government permitted the shilanyas and the Prime Minister opened his election campaign at Faizabad promising Ram Rajya. The charge survives every version of the 1986 argument.
5Bofors
It is the most famous corruption allegation in Indian history. Nobody has ever been convicted of anything in it, under any government, in thirty-eight years.
5.1 — The contract and the broadcast
In March 1986 India contracted with the Swedish arms manufacturer AB Bofors for 410 howitzers, at a price of roughly ₹1,437 crore. The gun itself was and is well regarded; it performed decisively at Kargil in 1999.
The contract contained an express provision against the use of agents or the payment of commissions.
On 16 April 1987 Swedish state radio broadcast a report that Bofors had paid bribes to secure it. Sweden’s national audit body subsequently confirmed that commissions had been paid.
Four days earlier, on 12 April 1987, the Defence Minister V.P. Singh had resigned from the government — over a separate inquiry he had initiated into commissions on a submarine purchase. The two events are frequently run together and were separate. He went on to lead the opposition that defeated the government in 1989.
Agent and commission: in a defence purchase, a middleman who represents the seller in the buyer’s country, and the payment he receives for securing the contract.
The practice is universal in the arms trade and it is not inherently criminal. What makes it consequential is that the money necessarily comes out of the price the buyer pays, and that the middleman’s usefulness consists of access to the people making the decision.
Which is why serious defence contracts prohibit it, as this one did. A bar on agents is a country’s way of saying that its procurement decision will be made on the equipment and not on the relationships.
Why it matters here: the established fact in this chapter is not that money changed hands somewhere in a murky trade. It is that money changed hands in breach of a term the Indian government had itself insisted on.
5.2 — What is established
Being precise here matters more than in almost any other chapter of this series, because the loose version has done forty years of work.
Money was paid. Documents obtained from Swiss authorities by the Central Bureau of Investigation in January 1997, after a long legal effort, established that AB Bofors made payments to intermediaries — including the Indian agent Win Chadha, the Hinduja brothers, and the Italian businessman Ottavio Quattrocchi.
Those payments breached the contract. The agreement prohibited agents and commissions, and commissions were paid.
Quattrocchi was close to the Prime Minister’s family. This is not disputed by anybody.
That is the established core, and it is substantial. A defence contract carrying an express bar on commissions was obtained with commissions paid to people including a family friend of the Prime Minister.
5.3 — What has not been established
Everything else.
A Joint Parliamentary Committee reported in 1988 finding no evidence of bribes; it had a government majority and was boycotted by the opposition, and it settled nothing.
A first information report was registered in January 1990, under the successor government. A charge sheet was filed in October 1999, under a different successor government, naming Quattrocchi, Chadha, the Hindujas, the former Defence Secretary, AB Bofors — and Rajiv Gandhi, who had been dead for eight years.
On 4 February 2004 the Delhi High Court quashed the charges of bribery against Rajiv Gandhi. That was three months before Congress returned to office.
On 31 May 2005 the Delhi High Court quashed the remaining charges against the Hinduja brothers and AB Bofors, and criticised the Central Bureau of Investigation for its conduct of the case, noting the public money it had consumed.
The CBI did not appeal within the period allowed. A private appeal by an individual has been pending in the Supreme Court since.
In 2006 the Indian government’s own law officer asked the British authorities to unfreeze Quattrocchi’s accounts for want of evidence linking them to the payments. In March 2011 a Delhi court accepted a closure report dropping the case against him. He died in 2013.
No court has ever convicted anybody of anything in the Bofors matter.
| Date | What happened | Who was in office |
|---|---|---|
| 1988 | A Joint Parliamentary Committee reports finding no evidence of bribes. Government majority; boycotted by the opposition. | Congress |
| Jan 1990 | A first information report is registered. | National Front |
| Jan 1997 | Swiss documents are obtained establishing payments to named intermediaries. | United Front |
| Oct 1999 | A charge sheet is filed, naming Rajiv Gandhi posthumously among the accused. | NDA |
| Feb 2004 | The Delhi High Court quashes the bribery charges against Rajiv Gandhi. | NDA |
| May 2005 | The Delhi High Court quashes the remaining charges and criticises the investigating agency. No appeal is filed in time. | UPA |
| 2006 | India’s own law officer asks Britain to unfreeze the accounts for want of evidence. | UPA |
| Mar 2011 | A closure report is accepted. The case ends. | UPA |
5.4 — Under whose watch
The failure is not one party’s.
The case was registered under a non-Congress government in 1990. It was charge-sheeted under a non-Congress government in 1999. The charges against Rajiv Gandhi were quashed while a non-Congress government was in office, in February 2004. The remaining charges were quashed in 2005 and the appeal was not filed, under a Congress government. The closure report was accepted in 2011, under a Congress government.
Every government of every party has held this file. None has produced a conviction.
This chapter’s discipline is to separate three things that are usually merged: what has been established, what has been alleged, and what has been decided by a court.
Solid: the contract and its terms; the Swedish broadcast of 16 April 1987; the Swedish audit finding that commissions were paid; the Swiss documents of January 1997 establishing payments to named intermediaries; and every court decision listed in section 5.3, each of which is reported with a date.
Alleged: that any Indian politician received any part of the money. No court has found it. The one occasion on which a court considered the charge against the Prime Minister, it quashed it.
What none of it can show: whether he knew. He died in 1991, and a man who cannot answer a charge cannot be tried on one — which the Delhi High Court said in 2004 when it quashed it.
This is why the chapter is written as it is. A book that repeated the popular version would be asserting, about a dead man, something no court has found in thirty-eight years — which is exactly the discipline Part Eleven’s Chapter One imposed on itself about November 1984, and it applies here or it applies nowhere.
Set the whole of it against the calendar and the shape of the thing becomes visible.
Thirty-eight years is longer than the entire period from independence to the contract being signed.
In that time the case has been examined by a parliamentary committee, investigated across three countries, charge-sheeted, quashed in two stages, appealed by a private citizen because the state did not appeal, and formally closed.
Two of the accused died before trial. The Prime Minister named in the charge sheet had been dead for eight years when it was filed.
What the country obtained for all of that is a set of documents proving that commissions were paid to three intermediaries, which the Swedish audit authority had said in 1987.
Which suggests the thing everybody argues about is not the thing that happened.
Everybody who discusses Bofors assumes that it was about corruption.
One side says a great scandal was buried. The other says an innocent man was destroyed by an allegation. Both are arguing about whether money reached a politician, and both expect the answer to settle something.
Thirty-eight years, four governments, a Joint Parliamentary Committee, Swiss documents, a charge sheet and multiple High Court judgments have produced no conviction of anybody. That is not a temporary state of the inquiry. It is the outcome.
What Bofors actually established was a political mechanism, and it is the mechanism rather than the money that has shaped Indian politics since.
An allegation, never proved, destroyed the largest parliamentary majority in Indian history in under three years. It did so without a trial, without a finding, and without any court ever determining that an Indian politician received a rupee. And the case then died slowly under every subsequent government, including the ones that had been elected on it.
Both halves matter. The first taught every Indian opposition that a defence procurement allegation is the most powerful instrument available in national politics. The second taught every Indian government that such an allegation need never be resolved, because the party that raised it will not pursue it once it is in office and has its own procurements.
That combination — maximum political effect, minimum evidential requirement, no resolution — is the Bofors legacy, and every large Indian defence controversy since has followed the identical arc.
Which means the question “was he guilty” is the least useful one available. It cannot be answered, it has not been answered, and answering it would not touch the thing Bofors actually did.
One thing this chapter does not do. It does not say the allegation was false. It says no court has found it true in thirty-eight years, that the one court to consider it against the Prime Minister quashed it, and that a book which asserts more than that is doing the thing the mechanism above depends on.
Established: the contract barred agents and commissions; commissions were paid; Swiss documents of January 1997 named the recipients, including a family friend of the Prime Minister.
Not established: that any Indian politician received anything. The charges against Rajiv Gandhi were quashed in February 2004, before Congress returned to office. No court has ever convicted anybody of anything in this matter.
Every government of every party has held the file and none produced a conviction. What Bofors established was a mechanism: maximum political effect, minimum evidential requirement, no resolution.
6Sri Lanka
India deployed an army to disarm a group it had spent years arming, on an assessment produced by the agency that had armed it. Over a thousand Indian soldiers died.
6.1 — What was signed
On 29 July 1987 the Indo-Sri Lanka Accord was signed in Colombo by Rajiv Gandhi and President J.R. Jayewardene.
Its terms were a serious attempt at a settlement. Sri Lanka would devolve power to provincial councils, which required an amendment to its own constitution. Tamil militant groups would surrender their weapons. And India would guarantee the arrangement, providing a peacekeeping force to oversee the disarmament.
The Indian Peace Keeping Force began deploying the following day.
The next morning, at the guard of honour before his departure, Rajiv Gandhi was struck with a rifle butt by a Sri Lankan naval rating. It is a small detail and it is the first accurate indication of how the arrangement was regarded in the country it was signed in.
Peacekeeping: the deployment of a force to hold a settlement that the parties have already agreed to.
The distinction from war-fighting is not about weapons but about consent. A peacekeeping force works because the combatants want the arrangement it is guaranteeing, and it can therefore be small, lightly armed and politically neutral.
Remove the consent of any significant party and the force is not keeping a peace. It is one side in a war, in a foreign country, with a force structure and rules of engagement designed for something else.
Why it matters here: the Indian Peace Keeping Force was named, sized and deployed on the first assumption and was fighting under the second within ten weeks.
6.2 — The premise
The accord rested on a proposition: that the Tamil militant groups, and above all the Liberation Tigers of Tamil Eelam, would give up their weapons.
The basis for believing it was India’s relationship with those groups. Through the early 1980s the Research and Analysis Wing had cultivated, trained and armed Tamil militants — including the LTTE — in camps in Tamil Nadu, as an instrument of pressure on Colombo.
So the assessment that the LTTE would comply came from the agency that had spent years building it, about a client it believed it understood.
The LTTE handed over a quantity of weapons and did not disarm.
6.3 — What the army was asked to do
By October 1987 the force sent to keep a peace was fighting a war, in a foreign country, against an enemy that knew the ground and had been trained partly by India.
Operation Pawan, the attempt to take Jaffna, began in October. The initial assault was launched on short notice, against an opponent whose strength had been underestimated, and the early actions produced heavy casualties. A helicopter-borne insertion at Jaffna University went badly wrong.
The pattern is the one this series has now met three times in twenty-six years.
In 1961 the Forward Policy was decided in the Prime Minister’s office by four men, one of them a soldier, on an intelligence assessment that China would not fight, over the professional objection of Western Command. Part Six.
In 1983 the Vice-Chief of Army Staff advised against an assault on the Golden Temple and was passed over for Chief. Part Ten.
In 1987 an army was committed to a foreign deployment on an intelligence assessment about the intentions of a group that agency had itself created, with the operational planning compressed into weeks.
Three instances, three decades, two parties of the same family in office, and the same structure each time: a political decision resting on an intelligence assessment, with the military content supplied afterwards and under time pressure.
6.4 — What it cost
The IPKF remained until March 1990, withdrawn on the orders of a successor government after the Sri Lankan President asked it to leave.
Something over 1,100 Indian soldiers were killed.
The devolution the accord was signed to secure was enacted in Sri Lankan law and has been implemented only partially in the thirty-eight years since.
The LTTE was not disarmed. It fought the Sri Lankan state for a further nineteen years.
And on 21 May 1991, campaigning in Tamil Nadu, Rajiv Gandhi was killed by an LTTE suicide bomber.
Over 1,100 Indian soldiers died in Sri Lanka.
That is more than India lost in the 1965 war with Pakistan, and it happened in a country India was not at war with, under an agreement India had signed to end a war it was not a party to.
They are commemorated at a memorial in Colombo built in 2008, eighteen years after the withdrawal, by the Sri Lankan government.
There is no equivalent national memorial in India, and the deployment appears in no Indian school syllabus.
Whether any of it was defensible is argued in India almost entirely in one direction, which is itself a reason to state the other case properly.
It is remembered in India almost entirely as a disaster, which is not a complete account of it.
India had a real interest: a civil war on its doorstep, fifty million Tamils in the neighbouring state with strong sympathies, refugee flows, and an opening for outside powers in a country that sits across India’s southern sea lanes. The accord obtained something no purely diplomatic effort had — a Sri Lankan constitutional amendment devolving power to Tamil areas, which remains the only formal concession the Tamil population has ever secured. A guarantee without an enforcement mechanism would have been worthless, and India provided one.
The whole arrangement depended on the LTTE disarming, and the people best placed to judge that were the people who had armed it — which is the least independent assessment available on any question. The force was deployed the day after signature, before the disarmament had been tested. When it failed, an army with no plan for a war it had not been told to expect was ordered to take a city in weeks. The interest was real; the method committed Indian soldiers to enforce an agreement whose central term nobody had secured.
What would settle it: the military appreciation prepared before 29 July 1987, if one was prepared. It has never been released, and whether it exists is itself an open question.
Why people care so much: because more Indian soldiers died in Sri Lanka than in the 1965 war with Pakistan, for an outcome nobody can name, and because it ended with the Prime Minister who ordered it being killed by the organisation it was aimed at.
Underneath both of those positions lies a description of what went wrong that neither side examines.
Everybody who writes about the Sri Lanka deployment — in India and in Sri Lanka — assumes that the LTTE broke the agreement.
The Indian version: an organisation India had helped betrayed it. The Tamil version: an agreement made over Tamil heads was imposed and resisted. Both describe a party that was bound and did not comply.
The LTTE was not a party to the accord. It was signed by two governments — India and Sri Lanka — and the disarmament of the Tamil groups was a term one government promised the other it would deliver.
Read that structure again. India undertook to a foreign state that a third organisation, which had not signed anything, would surrender its weapons. And India then deployed an army to enforce the undertaking, beginning the day after signature.
Nothing was broken, because nothing had been agreed by the party whose agreement mattered. What failed was a prediction, made by India, about what a group it had armed would choose to do — which is the precise thing the front matter said 414 seats cannot buy.
And the framing hides the responsibility. If the LTTE broke an agreement, the fault is the LTTE’s and India was let down. If India guaranteed something it had not obtained, the fault is India’s and it is a fault of process rather than of trust.
The general rule, which this series has met before in a different setting: an undertaking about somebody else’s future conduct is not an agreement. It is a forecast with signatures on it, and the signatures make it look like something it is not.
Four years later the organisation that was to have been disarmed killed the Prime Minister who signed the agreement. This part records that as a fact about consequences and not as a moral about them.
The accord of 29 July 1987 obtained a real concession — Sri Lankan constitutional devolution to Tamil areas — and rested on a premise: that the LTTE would disarm. It did not.
The assessment that it would came from the agency that had spent years arming it. The force deployed the day after signature, before the premise had been tested.
Over 1,100 Indian soldiers were killed. This is the third instance in twenty-six years of a political decision resting on an intelligence assessment with the military content supplied afterwards and under time pressure.
7What Was Built
This is the longest chapter of credit items in this series. Most of what is in it is still working, and one entry ended a twenty-year war that has not restarted in forty years.
7.1 — Telephones
In 1984 an Indian telephone connection took years to obtain, worked unreliably, and existed almost entirely in cities. A trunk call between two towns was booked and waited for.
The Centre for Development of Telematics was established in 1984 to design digital switching equipment in India, for Indian conditions — exchanges that would work in heat and dust without air conditioning, which imported equipment would not.
What followed over the next decade was the rural exchange network and the public call office. The small booth with a yellow sign and a metered telephone, run by an operator, appeared in tens of thousands of villages and small towns.
Whatever else is argued about this government, a very large number of Indians made their first telephone call because of a decision taken in these five years, and the institutional apparatus — a separate telecom commission, the corporatised operators created in 1986 — dates from here.
7.2 — Computers
The computer policy of 1984 cut import duties on hardware, permitted duty-free import of software for export production, and recognised software as an industry rather than as a suspicious import.
Part Four’s whole argument was about a permission system that decided what could be made and by whom. This was a deliberate exemption from it, in a sector nobody yet had a constituency in.
The Indian software export industry begins here. It now employs several million people directly.
7.3 — The vote at eighteen
The Constitution (Sixty-First Amendment) Act, 1988 lowered the voting age from twenty-one to eighteen, taking effect from 1989.
It enfranchised something in the order of fifty million people at a stroke, and it was done by a government that had no particular reason to expect those voters to support it. They did not: the government lost the election in which they first voted.
It is the least discussed constitutional amendment of the period and one of the very few in this series that expanded a right rather than qualifying one.
7.4 — Three accords
Between July 1985 and June 1986 the government signed settlements with three insurgencies or agitations.
Punjab, July 1985. Part Ten’s Chapter Eight covers it: a real settlement whose central term — the transfer of Chandigarh — was never implemented, and whose signatory was assassinated within a month.
Assam, August 1985. It ended a six-year agitation over migration, provided for detection and deletion of illegal voters against a 1971 cut-off, and led directly to elections and a new state government. It has held imperfectly; the citizenship questions it addressed are still live forty years later.
Mizoram, June 1986. This one worked completely.
The Mizo insurgency had run for twenty years. The accord brought the insurgent leadership into electoral politics — its head became Chief Minister — granted statehood, and ended the conflict.
It has not restarted. Mizoram has been among India’s most peaceful states for four decades, and has among its highest literacy rates.
Set against everything else in this part, that is the single most successful thing this government did, and almost nobody outside the north-east mentions it.
7.5 — Schools, water, vaccines
The National Policy on Education of 1986 produced the Navodaya Vidyalayas — residential schools intended to take able children out of rural districts and educate them at a standard otherwise unavailable to them. There are now over six hundred of them.
A set of technology missions was established from 1986 on specific, measurable objectives: drinking water, immunisation, literacy, oilseeds, telecommunications and dairy.
The design is worth noticing because it explains section 7.8. Each mission had one objective, a named person responsible, and a deadline. None of them required a contested political judgment.
7.6 — Two responses to pressure
Two episodes belong here because they cut in opposite directions and both are informative.
Bhopal. On the night of 2 December 1984 — three weeks before the election — a gas leak at a pesticide plant in Bhopal killed thousands of people and injured many times that number. It remains the worst industrial disaster anywhere.
The response that belongs in this chapter is the Environment (Protection) Act of 1986, a wide framework statute giving the central government general powers to regulate emissions, discharges and hazardous processes. It is the foundation of Indian environmental regulation and it exists because of Bhopal.
The response that does not belong in a credit chapter is the settlement of the claims, which was reached in 1989 for a sum widely regarded as inadequate, and the fate of the criminal proceedings, which is a subject of its own.
The defamation bill. In 1988 the government introduced legislation that would have made it substantially easier to prosecute journalists and shifted the burden of proof against them. It passed the Lok Sabha.
The press protested nationally, in concert, and the government withdrew it.
That is worth recording precisely because of Chapter Eight. It is the one occasion in these five years when a decision taken quickly by a small circle met organised opposition from outside the structure and was reversed. The mechanism that was missing everywhere else existed here — not inside the government, but in a profession with the means to make itself heard — and it worked immediately.
7.7 — The amendment that was defeated
In 1989 the government introduced a constitutional amendment bill to establish elected local government — village and district councils with guaranteed elections, reserved seats, and a fixed share of funds.
It passed the Lok Sabha and was defeated in the Rajya Sabha, by a small margin, by an opposition that regarded it as an attempt to bypass state governments.
Three years later, in 1992, the Seventy-Third and Seventy-Fourth Amendments enacted substantially the same design under a successor Congress government. They are the constitutional basis of local government in India today.
So the credit for panchayati raj is customarily given to 1992, and the draft is from 1989.
7.8 — The ledger for this part
| Where | What was done |
|---|---|
| Mizoram | A twenty-year insurgency ended by accord in June 1986. Statehood granted, insurgent leadership brought into government. The peace has held for forty years. |
| Assam | The 1985 accord ended a six-year agitation and produced elections. The citizenship questions it addressed remain unresolved. |
| Punjab | The 1985 accord. Chandigarh never transferred; the elected government it produced dismissed in 1987. Part Ten. |
| Uttar Pradesh | Ayodhya, February 1986 and November 1989. Chapter Four. |
| Rural India generally | The first telephone exchange, the first public call office, and in six hundred districts the first residential school of its kind. |
| Tamil Nadu | The state from which the Sri Lanka deployment was staged and in which the Prime Minister was killed in 1991. |
A public call office was a booth with a telephone, a meter and a person to run it.
For a family with a son working in another state it was the difference between a letter that took a fortnight and a voice on the same evening. For a farmer it was the price at the mandi before loading the cart. For a village with one telephone it was the first time an ambulance could be called at all.
None of that appears in a growth statistic and all of it happened between 1985 and 1995.
Set against Chapters Three to Six, it is worth noticing which of the two lists the people affected would have chosen.
7.9 — Why these worked
Compare the list in this chapter with the decisions in Chapters Three to Six and one difference does all the work.
Nothing in this chapter required a contested judgment about what would happen.
Building a rural exchange requires money, engineering and somebody competent left alone. Lowering the voting age requires a decision about principle. Establishing a school system requires an organisation. None of them turns on a prediction that could be wrong.
Every failure in Chapters Three to Six turned on exactly such a prediction. That a community would accept a reversal without the other community reacting. That unlocking a shrine would defuse a mobilisation rather than start one. That an armed group would disarm because the agency that armed it said so.
Which is the front matter’s hidden assumption arriving with evidence. A majority of 414 buys unlimited capacity to execute. It buys nothing at all for predicting, and predicting requires the thing this government did not have: somebody with standing whose objection had to be answered.
The items in this chapter are unusually easy to verify because most of them still exist and can be counted.
Telephone connection figures, exchange numbers and public call office counts were published annually. Software export figures are published. The Sixty-First Amendment is in the constitution. The three accords are published texts with dates and signatories. The Navodaya schools have a register. The Seventy-Third and Seventy-Fourth Amendments can be compared clause by clause with the 1989 bill.
That is why this chapter states outcomes rather than intentions: the intentions are the same in every government’s programme, and the difference here is that the things were built.
What it cannot show: how much of it would have happened anyway. Telecommunications and computing were transforming everywhere in the 1980s, and India was not the only country to liberalise them. What is defensible is narrower — that these decisions were taken here, at that time, against a permission system that Part Four showed defeating almost everything else.
That distinction is the most useful thing in this part, and it generalises past this government. Ask of any proposal whether it requires building or predicting. A government with a large majority and a weak process is reliable at the first and dangerous at the second, and India has had several of them.
What was built and still works: rural telephone exchanges and the public call office; a computer and software policy that founded an industry; the vote at eighteen; the Navodaya schools; the technology missions; and an environment statute after Bhopal.
The Mizoram Accord of June 1986 ended a twenty-year insurgency and the peace has held for forty years. It is the most successful thing this government did and almost nobody mentions it.
Panchayati raj was drafted in 1989, defeated in the Rajya Sabha, and enacted in 1992 on substantially the same design. Nothing in this chapter required a prediction that could be wrong — which is exactly what everything in Chapters Three to Six did.
8The Pattern
Part Eight found a party converted into an instrument of one office. This chapter is the test of that finding, because the office changed hands and the person could hardly have been more different.
8.1 — The question
Part Eight’s Chapter Two established that the 1969 split transferred control of candidate selection, and that after 1971 every Congress member of parliament owed his seat to one office. Its conclusion was that the party’s capacity to test or remove a leader had gone from unusable to unimaginable.
That finding has two possible readings and they lead to entirely different places.
The biographical reading: a leader of exceptional determination took a party apart and ran it personally. On this reading the problem departs with the person.
The structural reading: an arrangement was created in which the party had no mechanism for testing a proposal, and the arrangement would produce the same effects with anybody in the office.
December 1984 is the experiment. The office passed to a forty-year-old with no political history, no faction, no record in the machine, and a stated intention to clean it up. If the biographical reading is right, something should change.
8.2 — What did change
A good deal, and it belongs on the record.
The Emergency’s instruments were not used. No national emergency was proclaimed, no press censorship was imposed, and detention without trial was not deployed against political opponents on any comparable scale.
He said publicly, at the party’s centenary in December 1985, that the organisation was in the hands of brokers of power who dispensed patronage — which is not a thing his predecessor ever said about a party she had built.
And in 1989 he lost an election and left office, which is the second peaceful transfer of power in this series and is not a small thing.
8.3 — What did not
The decision-making did not change at all.
Look at the four decisions in Chapters Three to Six. Every one was taken quickly, by a very small circle, without any institutional body testing the proposal, and with the objection arriving afterwards from outside.
Shah Bano. Announced in January 1986, legislated by May. The warning came from a retired judge, by letter, after the Bill was introduced.
Ayodhya. Whatever was decided and by whom, no account of it involves a cabinet discussion, and the account from inside the Prime Minister’s Office describes a Prime Minister asking afterwards who had done it.
Bofors. The government’s response ran through a Joint Parliamentary Committee with a government majority, which settled nothing and was boycotted.
Sri Lanka. Signed on 29 July, deployment beginning on 30 July, on an assessment from an agency with the least capacity to be objective about it.
And one of the two most consequential things the government did — the Tenth Schedule — extended the structure to every party in India by statute. After 1985 no member of any Indian legislature could vote against a party direction without losing his seat. What Part Eight described as a Congress condition became a national one, permanently, enacted by a Prime Minister who had criticised the party’s brokers three months earlier.
8.4 — What that establishes
The experiment gives a clear result.
The person changed completely. The instruments of coercion were not used. The rhetoric was the opposite. And the failure mode was identical: decisions taken fast, by few, untested, with the professional or judicial objection arriving too late and from outside the structure.
Part Eight’s finding was structural. It survives the person, which means it was never a fact about Indira Gandhi, and the biographical reading of these two decades — which is the reading almost all Indian political argument uses — is wrong.
Put the two governments side by side as a controlled experiment.
Same office. Same party. Same constitution. Same civil service.
Different person in almost every respect: age, temperament, experience, political history, stated intentions, and willingness to use coercive instruments.
Same failure shape: a decision formed by a handful of people, adopted at speed, with no body able to test it, and the objection arriving afterwards from a judge, a general or a foreign government.
When you change every variable except one and the outcome does not change, you have identified the variable.
Which is not how this period is discussed by anybody.
Everybody who argues about this period of Indian politics assumes that the problem was the person.
The hostile version: a dynasty produced leaders unequal to the office. The defensive version: capable leaders were undone by circumstance and by their opponents. Both are arguments about individuals, and both take for granted that the way to explain a government is to explain who ran it.
Set the two governments beside each other. One leader was fifty-one, had spent twenty years inside the machine, had split her party to control it, and used every instrument in Part Nine. The other was forty, had been a pilot four years earlier, had no faction, and used none of them.
The decision failures are the same shape in both. A proposal formed in a small circle. No body with standing to test it. Execution at speed. The objection arriving afterwards, from a judge, a general or a foreign government.
When two people that different produce identically shaped failures in the same office, the office is the variable.
And the mechanism is not mysterious. Part Three found the party losing its internal check by 1954. Part Eight found candidate selection concentrated in one office by 1971. This part found that check extended to every party in India by statute in 1985. At no point in that sequence did anybody build a replacement — a cabinet system that genuinely deliberates, a parliamentary committee structure with teeth, a civil service able to record a formal objection and survive it.
Which means the honest finding of this whole series so far is not about a family. It is that India built a system in which the quality of government depends entirely on the judgment of whoever holds one office, and then stopped building anything that could compensate when that judgment was wrong.
That is a much less satisfying conclusion than a dynasty. It is also the only one that survives putting two completely different people in the same chair and watching the same thing happen.
A claim of that kind should say how much weight two cases can carry.
The comparison in this chapter is between two governments this series has already documented in detail, and it therefore rests on Parts Eight and Nine as much as on this one.
What makes it usable as evidence is that the two are unusually well matched. Same office, same party, same constitutional framework, same senior civil service, and only nine years apart. Comparisons across countries or across decades are contaminated by everything that differs; this one is contaminated by very little except the person, which is exactly the variable being tested.
What it cannot show: whether a third person would have done differently. Two cases are not a series, and this chapter’s claim is stated as a finding about two governments rather than as a law about the office.
The stronger version of the test would run the comparison across every Indian Prime Minister since 1947, scoring the decision process rather than the decisions. Nobody has done it, and it is the most useful piece of work anybody could do on the material in this series.
Which is the finding this series has been assembling for eleven parts without intending to, and it will be tested again in every part that follows.
What changed: the Emergency’s instruments were not used, the rhetoric about the party was the opposite, and in 1989 he lost an election and left.
What did not: every decision in Chapters Three to Six was taken fast, by a small circle, untested, with the objection arriving afterwards from outside. And the Tenth Schedule extended the structure to every party in India by statute.
Two people that different producing identically shaped failures means the office is the variable. Part Eight’s finding was structural and it survives the person.
9What Would Have Prevented It
Four moments, all of them cheap. And a defence that is stronger here than in most parts of this series.
9.1 — Four arguments for the defence
These are the strongest available. Two are conceded outright.
One. The position he inherited was close to impossible. Sworn in on the evening of his mother’s assassination, at forty, with four years in politics, in the week Part Eleven describes. No preparation, no faction of his own, and an election six weeks later. Very few people would have done better and most would have done worse.
Two. Chapter Seven’s list is real. Conceded in full and given a chapter rather than a paragraph. Telephones, computers, the vote at eighteen, three accords of which one ended a war, the Navodaya schools, an environment statute, and the panchayati raj design that was enacted three years later.
Three. He accepted defeat and left. In 1989, having lost, he handed over power. Part Nine said that of 1977 and it says it again: the transfer of power after a defeat is the hardest thing a political system does, and it has now happened twice in this series under governments that had every instrument available to prevent it.
Four. He was right about panchayati raj and lost the vote. The amendment was defeated by the opposition in 1989 and enacted on his design in 1992. A government cannot be indicted for a reform its opponents blocked.
9.2 — Four moments
Part One requires an alternative that was available at the time. All four here are, and all four are cheap.
April 1985 to May 1986: do nothing about Shah Bano. The judgment stood. The agitation was serious and would have been survived — the government had 414 seats and four years left. The Supreme Court’s own decisions of 2001 and 2024 establish that the position was legally sustainable, because it is the position now. Cost: a year of political difficulty.
February 1986: not facilitate the unlocking, or reverse it. The High Court directed on 3 February that the property not be altered. A central government that regarded the order as dangerous had instruments — and did not use them. Cost: an agitation from the other direction.
November 1989: not permit the shilanyas, and not launch the campaign at Faizabad. These were the Prime Minister’s own decisions, are disputed by nobody, and required nothing but declining. Cost: nothing measurable.
July 1987: obtain the disarmament before deploying. The accord could have provided for the force to enter after verified surrender of weapons rather than the day after signature. Cost: a slower agreement.
None of the four required a concession to anybody, a constitutional amendment, or any risk to the government’s majority.
9.3 — What I would have to retract
Three things would break this part.
If the Prime Minister’s Office record for February 1986 showed that the unlocking was ordered from Delhi, Chapter Four’s careful grading becomes over-cautious and the chapter should be considerably harder.
If it showed the opposite — that the government attempted to reverse it and was prevented — then Chapter Four’s charge reduces to the 1989 decisions alone.
If a military appreciation prepared before 29 July 1987 were released showing that the army was consulted and considered the deployment sound, Chapter Six’s central claim fails and the third instance of the pattern disappears.
All three depend on files the Indian government holds. That is the seventh part in a row ending on that sentence.
The final question, and the defence’s best ground.
He arrived with no preparation into the worst possible week, held an office his party had made unaccountable long before he entered it, and inherited a set of conflicts — Punjab, Assam, Sri Lanka, Ayodhya — that had been accumulating for decades. He settled three of them by agreement within two years, which nobody else managed before or since. Judging a forty-year-old with four years in politics against an ideal decision-maker, on questions that had defeated his predecessors, is not analysis.
Nobody is asking for an ideal decision-maker. Section 9.2 lists four occasions on which the correct course was to decline to act — not to legislate, not to facilitate, not to campaign, not to deploy before a condition was met. Declining does not require experience, judgment or courage. It requires somebody in the room whose objection has to be answered, and Chapter Eight established that no such person existed. That is the charge, and it is against the arrangement as much as the man.
What would settle it: nothing. It is a question about how much of government is structure and how much is judgment, and this series has now offered eleven parts of evidence towards structure.
Why people care so much: because the biographical answer is available to everybody and requires nothing, and the structural answer implicates every party that has held the office since and has changed none of it.
That is the case against this part, and the second and third defences I think simply win. What none of the four reaches is section 9.2, because every alternative there consisted of not doing something.
Four defences: an impossible inheritance; a real list of things built; he lost in 1989 and left; and he was right about panchayati raj and was outvoted. The second and third are conceded outright.
Four moments, all cheap, all requiring only that a government decline to act: Shah Bano, February 1986, November 1989, and deploying before disarmament.
Declining does not require experience or courage. It requires somebody in the room whose objection has to be answered — and Chapter Eight established that no such person existed.
10An Honest List of What We Do Not Know
One thing in this part is asserted by almost every account of the period and is not established. This chapter is mostly about why I have refused to assert it.
10.1 — Genuinely unknown
Five things in this part cannot be established, and the first is the one that would change it most.
Who decided the unlocking at Ayodhya on 1 February 1986. Chapter Four sets out both accounts. One says it was arranged from Delhi as a counterweight to the Shah Bano reversal. The other, from an official inside the Prime Minister’s Office, says the Prime Minister was not informed and was asking afterwards who had done it.
Why it is unknown: the Prime Minister’s Office record and the Uttar Pradesh government’s file are both closed. This is the single most consequential unresolved question in the part.
Whether a military appreciation was prepared before the Sri Lanka accord. Chapter Six’s claim is that the army was committed before the premise of the agreement had been tested.
Why it is unknown: if such a document exists it has not been released, and whether one was prepared at all is itself unestablished. That absence is suggestive and it is not proof.
Whether any Indian politician received money in the Bofors matter.
Why it is unknown: thirty-eight years, four governments, Swiss documents, a charge sheet and multiple High Court judgments have produced no such finding. The one occasion on which a court considered the charge against the Prime Minister, it quashed it. He died in 1991 and cannot answer it.
What the political calculation behind the Shah Bano reversal actually was. Every account describes a fear of losing Muslim support. None of them is a record of a discussion.
Why it is unknown: cabinet papers closed.
How much of Chapter Seven would have happened anyway. Telecommunications and computing were transforming in every comparable country in the 1980s.
Why it is unknown: it is a counterfactual. What is defensible is narrower and is what the chapter claims: that these decisions were taken here, then, against a permission system that Part Four showed defeating almost everything else.
10.2 — Solid
The result. 414 seats of 514 on about 49 per cent — the largest majority any Indian party has held.
The Tenth Schedule. Its text, its two limbs, and the fact that the second disqualifies a member for voting against a party direction on any subject. Constitutional text.
The Shah Bano sequence. The judgment of 23 April 1985, the announcement of 15 January 1986, the Bill of 25 February, the Krishna Iyer letter of 28 February, the Act of May 1986, and the judgments of 2001 and July 2024 which restored the position.
The Ayodhya chronology. The application of 25 January 1986, the order of 1 February, the locks broken within about forty minutes with the state broadcaster present, the officials’ testimony that no disorder was anticipated, the High Court’s direction of 3 February, and the events of 14 and 15 February.
The 1989 decisions. The shilanyas in November and the campaign launch at Faizabad on 3 November. Not disputed by anybody.
The Bofors record. The contract and its bar on commissions; the broadcast of 16 April 1987; the Swiss documents of January 1997; the quashing of charges against Rajiv Gandhi in February 2004; the quashing of the remaining charges in May 2005; the closure report accepted in March 2011; and that no conviction has ever been obtained.
Sri Lanka. The accord of 29 July 1987, the deployment from 30 July, Operation Pawan from October, the withdrawal completed in March 1990, and over 1,100 Indian dead.
Everything in Chapter Seven. Most of it still exists and can be counted.
10.3 — What this part refuses to assert
Two things, and they are refused for the same reason.
That the Prime Minister ordered the unlocking. It is the standard claim, made in most accounts, and it is not established. An official who worked in his office says the opposite.
That he took money over Bofors. It is the most repeated allegation in Indian public life and no court has found it in thirty-eight years, including a court that specifically considered it.
Both refusals cost this part something. The first removes its most quotable line and the second removes the sentence everybody expects a book like this to contain.
The reason is the one Part Eleven’s Chapter One gave about November 1984, and it applies here or it applies nowhere. A series that will repeat an unproved allegation against one man cannot demand proof when the allegation runs the other way. Part Eleven refused to name individuals against whom no body has found, in a chapter about the killing of two thousand seven hundred people. The same discipline has to survive contact with a subject where the popular version is more convenient.
If either file is opened and shows what most accounts assert, this part will have been too careful, and I would rather be wrong in that direction.
Genuinely unknown: who ordered the unlocking, whether a military appreciation preceded the Sri Lanka accord, whether any Indian politician received Bofors money, what the Shah Bano calculation was, and how much of Chapter Seven would have happened anyway.
Solid: the result, the Tenth Schedule, the whole Shah Bano sequence, the Ayodhya chronology, the 1989 decisions, the Bofors court record, the Sri Lanka deployment, and Chapter Seven’s list.
This part refuses to assert that the Prime Minister ordered the unlocking or that he took money. A series that repeats an unproved allegation against one man cannot demand proof when it runs the other way.
Sources & further reading — Part 12
Timeline, 1984–1991
Dates only. The arguments are in the chapters.
| Date | What happened |
|---|---|
| 31 Oct 1984 | Indira Gandhi assassinated. Rajiv Gandhi sworn in that evening. Part Eleven covers the week that followed. |
| 2–3 Dec 1984 | The Bhopal gas disaster, days before the election. It is the direct cause of the Environment (Protection) Act of 1986. |
| Dec 1984 | General election. Congress wins 414 of 514 seats on about 49 per cent of the vote. |
| 1984 | The Centre for Development of Telematics is established. The new computer policy cuts duties and recognises software as an industry. |
| 23 Apr 1985 | Shah Bano. A five-judge bench holds that section 125 of the Code of Criminal Procedure applies to a divorced Muslim woman. |
| Jul 1985 | The Punjab Accord. Part Ten’s Chapter Eight. |
| Aug 1985 | The Assam Accord ends a six-year agitation. |
| 1985 | The Fifty-Second Amendment inserts the Tenth Schedule. A member is disqualified for leaving his party or for voting against a party direction. |
| Dec 1985 | At the party centenary in Bombay he describes the organisation as being in the hands of brokers of power. |
| 15 Jan 1986 | He announces the intention to legislate to nullify the Shah Bano judgment. |
| 25 Jan 1986 | A lawyer applies to the Faizabad court for the gates of the disputed structure at Ayodhya to be unlocked. |
| 1 Feb 1986 | The locks are ordered removed and are broken within about forty minutes, with Doordarshan present and broadcasting. |
| 3 Feb 1986 | The High Court directs that the nature of the property not be altered. |
| 14–15 Feb 1986 | Protests and rioting in several states. The Babri Masjid Action Committee is formed. |
| 25 Feb 1986 | The Muslim Women (Protection of Rights on Divorce) Bill is introduced in the Lok Sabha. |
| 28 Feb 1986 | V.R. Krishna Iyer, a retired Supreme Court judge, writes to the Prime Minister protesting against the Bill and warning of the consequences. |
| Mar 1986 | The Bofors contract is signed — 410 howitzers, about ₹1,437 crore, with an express bar on agents and commissions. |
| May 1986 | The Bill becomes law. Maintenance is restricted to the iddat period. |
| 1986 | The National Policy on Education. The Environment (Protection) Act. The technology missions begin. |
| 30 Jun 1986 | The Mizoram Accord. A twenty-year insurgency ends. The peace has held. |
| 12 Apr 1987 | V.P. Singh resigns as Defence Minister, over a separate submarine commissions inquiry. |
| 16 Apr 1987 | Swedish radio broadcasts the allegation that Bofors paid bribes. |
| 29 Jul 1987 | The Indo-Sri Lanka Accord is signed. The Indian Peace Keeping Force begins deploying the next day. |
| Oct 1987 | Operation Pawan. The force sent to oversee disarmament is fighting a war. |
| Dec 1985 | India is a founding member of SAARC at its first summit in Dhaka. |
| Aug–Sep 1988 | A defamation bill that would have made prosecution of journalists substantially easier passes the Lok Sabha and is withdrawn after a national press protest. |
| 1988 | The Sixty-First Amendment lowers the voting age from twenty-one to eighteen. |
| 1989 | The constitutional amendment bill on panchayati raj passes the Lok Sabha and is defeated in the Rajya Sabha. |
| 3 Nov 1989 | The election campaign is launched at Faizabad with a promise of Ram Rajya. |
| 9 Nov 1989 | The shilanyas is performed adjacent to the disputed structure. |
| Dec 1989 | Congress loses the general election. Power is transferred. |
| Mar 1990 | The Indian Peace Keeping Force completes its withdrawal. Over 1,100 Indian soldiers have been killed. |
| 21 May 1991 | Rajiv Gandhi is assassinated by an LTTE suicide bomber in Tamil Nadu. |
| 1992 | The Seventy-Third and Seventy-Fourth Amendments enact panchayati raj on substantially the 1989 design. |
| 2004, 2005, 2011 | The Bofors charges are quashed in stages and the case is closed. No conviction is ever obtained. |
| 2001, 2024 | The Supreme Court reads down the 1986 Act, and then holds that section 125 applies to all women irrespective of religion. |
Part Thirteen starts here. It leaves the chronology and takes the parts of India this series has kept promising to return to — the north-east, the tribal belt, and the places where the Indian state has been least visible and most armed.
Glossary
Every hard word used in this part, in plain English.
| Word | What it means |
|---|---|
| Defection | Leaving the party on whose ticket you were elected. Endemic in Indian state politics from the 1960s and the problem the Tenth Schedule was written for. |
| Iddat | The waiting period, roughly three months, that follows a divorce under Muslim personal law. The 1986 Act restricted maintenance to it. |
| IPKF | The Indian Peace Keeping Force, deployed in Sri Lanka from July 1987 to March 1990 under the accord. Over 1,100 of its soldiers were killed. |
| Mandate | The authority a government claims from an election to do particular things. An election produces a number of seats and not a list, so a large majority is a large permission rather than a large mandate. |
| Section 125 CrPC | A religion-neutral provision of Indian criminal procedure under which any person unable to maintain themselves may claim maintenance. Its purpose is to prevent destitution, not to regulate family law. |
| Shilanyas | A foundation-stone ceremony. Performed adjacent to the disputed structure at Ayodhya in November 1989 with the government’s permission. |
| Tenth Schedule | Inserted by the Fifty-Second Amendment in 1985. Disqualifies a legislator who leaves his party or who votes against a party direction on any subject. |
| Whip | A party’s direction to its legislators on how to vote. In most parliamentary systems defying one ends a career; in India since 1985 it ends a membership. |
What Comes Next
Part Thirteen — The Other India
Part Thirteen leaves the chronology for the places this series has kept promising to return to. It asks:
- The Armed Forces Special Powers Act was passed in 1958 as a temporary measure for one region. Where is it in force now, for how long has it been, and what has been established about its use?
- Part Five’s ledger recorded the north-eastern states as absorbing ten million refugees twice and receiving central attention only as a security frontier. What has been built there, measured against what has been spent?
- The Fifth and Sixth Schedules of the constitution protect tribal land. How much has been transferred anyway, under what authority, and to whom?
- Displacement by dams, mines and industry has a number attached to it. Where does the number come from and what does the evidence support?
- Part Two’s counting rule has been applied four times. What happens when it is applied to the people this part is about?
- Every part of this series has found instruments used earlier and harder in places where the state’s presence was in question. Does that pattern hold here, and what does it look like when it does?
Questions, not topics. If any of them is answered badly, you will be able to tell.