The Reckoning

1The invoice: what the Green Revolution actually cost

Punjab was asked to feed India and it did. The arrangement worked so well that for thirty years nobody asked what it was being paid for with. The answer is water, and the Central Ground Water Board has now put a date on when it runs out.

The deal, stated plainly

From the mid-1960s, Punjab was given high-yield seed, subsidised fertiliser, a guaranteed price for wheat and rice, and eventually free electricity for tubewells. In return it grew the grain that ended Indian famine.

Part 12 covered the economics of that arrangement — the single buyer, the administered price, the 14 lakh free tubewells. This chapter covers the physical bill.

The numbers, and the date

The second projection matters more than the first, and it is not really about quantity.

Good-quality water in Punjab sits in the first aquifer, down to about 100 metres. Below 300 metres the water is heavily mineralised and contaminated — not fit for irrigation and not fit for drinking.

I want to mark the status of that projection honestly. It is a projection, made by extrapolating current extraction rates. Extraction rates can change, and Chapter 1's next section shows they have started to. A projection is not a prophecy. But it is the best estimate the responsible technical body has published, and nobody has produced a serious argument that it is too pessimistic.

The good news, given its full weight

Rule 5 of this work says nobody is only a hero or only a villain, and the same discipline applies to trends. So here is the recovery, stated at full strength rather than buried.

The extraction improvement is genuine and it deserves to be said loudly, because Punjab is almost never credited with anything.

But note what is driving it. The improvement comes substantially from expanded canal irrigation — that is, from surface water rather than groundwater. And Part 12 established where the surface water is controlled from: Bhakra in Bilaspur district and Pong in Kangra district, both in Himachal Pradesh.

So the aquifer recovers to the exact extent that Punjab increases its dependence on storage it does not own. That is not an argument against doing it. It is the correct thing to do. It is simply the trade that is actually being made, and it should be visible.

The stubble numbers, and why I do not fully believe them

This is a case study in how to read a government statistic, and it belongs here because the improvement is being claimed as a policy triumph.

I include this at length because it is the single best illustration in this work of Rule 2. A number went down. A policy was credited. And the mechanism that produced the number changed at the same time, in a way that would produce the same result with no change in behaviour at all.

The health question, marked carefully

The Malwa belt has been called Punjab's cancer belt for two decades, and a train from Bathinda to Bikaner carries patients to a cancer hospital in Rajasthan often enough to be called the cancer train. About a third of that train's roughly 300 passengers are travelling for that hospital.

The elevated incidence is real. The confident explanations are not. That distinction is the whole of what an honest reader can take from this file, and it is why I have not used the cancer figures as a headline anywhere in this work.

22025: the year the water came the other way

In August and September 2025 Punjab had its worst flood since 1988. Fifty-seven people died. Around 2,555 villages were affected. And the mechanism that made it worse is the same mechanism Part 12 identified as the state's fundamental dependency, running in the opposite direction.

What happened

The immediate cause was meteorological: back-to-back heavy rainfall in August and early September turned a rainfall deficit into a surplus very quickly.

The aggravating cause is the one that matters for this work. The rain filled the upstream reservoirs, which then had to make controlled releases, and those releases moved down into a Punjab already saturated.

The finding that closes the argument between Parts 12 and 13

Put the two chapters together and the shape of Punjab's actual position appears.

The same dependency, two directions

And this is going to recur. Whatever anyone believes about the causes, the pattern of very heavy rainfall arriving in compressed bursts is the pattern the region is being told to expect. A state with an emptying aquifer and an increasing flood exposure is not experiencing two problems. It is experiencing one problem — loss of control over the timing of water — expressed twice.

3The farm: an arithmetic nobody can escape

A Punjabi farmer is caught between two facts. The only crops with a guaranteed buyer are the two destroying his water. And his holding has shrunk by nearly a third in five years. This chapter is that arithmetic, with the numbers that are solid separated from the numbers that are not.

The debt

The landholding line is the one that does not get enough attention. A 31% fall in average holding size in five years is not a slow structural drift. It is a collapse in the viability of the unit of production, driven by inheritance division and by distress sale.

And the interest spread tells you the rest. When a household has exhausted institutional credit, it moves to the arhtiya and the private lender at 17% to 21%. A farm generating thin margins on an administered price cannot service that. This is the mechanism, and it is well documented.

The suicides, and the honest position on the data

This is a place where I am going to be more careful than most writing on the subject, because the numbers are used as weapons by everybody and the underlying data is weak.

The trap, in one paragraph

Here is why "just grow something else" has failed for forty years, stated as an arithmetic problem rather than a moral one.

Every element of that trap was created by a policy that was, in its time, correct. India needed grain and Punjab produced it. Nobody set out to build this. It is what a successful policy looks like sixty years later when nobody redesigned it.

4Drugs: what the evidence actually supports

You have heard that most of Punjab's young men are addicts. That claim is not supported by any evidence I can find. What the evidence does support is a serious, concentrated opioid problem with hundreds of thousands of dependent users — which is bad enough, and is being obscured by the exaggeration.

The claim in circulation

The widely repeated version says that some very large share of Punjabi youth — the figure most often quoted is around 70% — are drug addicts. It appears in films, in political speeches from every party, and constantly online.

I have not been able to find a survey that supports it. What I have found is that the number appears to have escaped from a small, non-representative study and then been repeated until it became a fact about Punjab that everybody knows.

What the actual surveys say

Those four rows are not measuring the same thing and it is essential to say so.

Why the exaggeration is not harmless

I could have left this chapter as a correction and moved on. I am not going to, because the inflation of this particular figure has three specific consequences and they all point the same way.

And there is a fourth consequence, which is the one that should worry anybody who actually cares about the users. If the problem is believed to be universal, it is also believed to be hopeless. Nobody funds treatment capacity for a lost cause. A precise number of 232,000 is a number you can build clinics for.

5The emptying — and the door that just shut

For twenty years Punjab exported its surplus young people and imported their money. Everyone called it prosperity. Between 2024 and 2026 the receiving countries changed the rules, and the arrangement Punjab's economy quietly depended on has stopped working.

The scale of what was happening

Read the last row twice. In one year, Punjabi families sent abroad, in tuition and living costs to a single country, a sum in the region of three and a half billion dollars.

That money left Punjab. What came back was remittances — and India's remittance inflows reached a record $110.47 billion in 2025-26, the first time the figure has crossed a hundred billion, with Punjab among the states most dependent on that flow.

The reversal

Some of this is redirection rather than disappearance — the agencies are switching destinations, not shutting entirely. I do not claim migration has stopped. But a fall from 51.6% to 8.1% of a single destination's intake in two years is not a redirection. It is the closing of the specific channel Punjab had built its aspirations around.

Why this is the most under-discussed fact in Punjab

So Punjab now faces a question it has not had to answer since the 1980s: what does it do with its own young people?

The state has an economy that does not generate enough graduate employment, an agriculture sector that cannot absorb more labour on shrinking holdings, an industrial base that has been relocating for two decades, and — as of 2024 — a substantially narrowed exit.

Part 11 asked what keeps a separatist movement alive in a state that will not vote for it. Part 12 asked whether the exit it proposes would work. This chapter is about a different exit, the one hundreds of thousands of families were actually using, and it is the one that just narrowed.

6The economy that stopped leading

Punjab was the richest state in India. It is now somewhere around tenth. This chapter is about how that happened, and about a statistic that gives two opposite answers depending on which honest method you use.

The slide

Nothing dramatic caused that. No single disaster. It is what happens when a state grows more slowly than the country for twenty-five consecutive years.

The statistic that answers itself twice

Now the case study in reading numbers, and it is a good one because both figures are correct.

Is Punjab above or below the national average?

Where the industry went

Punjab's industrial thinning is often blamed on militancy, and Part 10 documented the capital flight of that decade honestly. But the larger movement happened afterwards, in peacetime, for reasons that had nothing to do with terrorism.

Note the connection between the last two rows and Chapter 3. The free power that keeps the tubewells running is cross-subsidised, and part of what pays for it is a higher industrial tariff. The policy protecting farm incomes is one of the policies pushing out the factories that would employ farm children.

That is not an argument for abolishing the subsidy tomorrow. Chapter 9 explains why that is politically impossible and probably socially catastrophic in the short run. It is an argument for seeing that the pieces are connected, which is the thing that thirteen parts of this work have been for.

7A state that ran out of parties

The Shiromani Akali Dal carried Sikh political demands for a century — the morchas, the Punjabi Suba, the Anandpur Sahib Resolution. In 2022 it won three seats out of 117. Nothing in Punjab's recent politics makes sense until you understand that a hundred-year-old institution vacated the field.

Thirty years of alternation

Punjab throws out incumbents. That is worth stating plainly, because it is the answer to anyone claiming Punjab's elections do not reflect Punjabi opinion. Every government since 1997 that sought a third consecutive mandate has been removed, and in 2022 the removal was total.

The collapse, and what filled it

The Akali Dal's fall has specific causes and they are not mysterious: the 2015 sacrilege incidents and the police firing that followed, prolonged family control of the party, the alliance decisions, and the eventual break with the BJP in September 2020 over the farm laws — which left it without its urban Hindu partner and without a farm constituency that trusted it.

Part 11 established what happened next, and this is the join between the two parts.

Where it stands now

Two structural observations, neither of which favours any party.

First, Punjab now has competitive multi-party politics with no party that speaks for the panthic constituency. That constituency has therefore been voting for independents. Independents cannot govern, cannot legislate a programme, and cannot be held accountable at the next election in the way a party can.

Second, none of the four parties has a published answer to Chapter 1. Not one of them has a water plan on the scale the CGWB projection implies. The 2027 election will be fought, on present evidence, on welfare transfers, corruption and law and order — while the aquifer runs to a schedule none of the manifestos mentions.

8The status quo, audited

Part 12 ran six tests on the proposal to leave. This chapter runs the same six tests on the arrangement Punjab is in. Reading them side by side is the point of writing both.

The same six questions

The two audits, side by side

Reading Part 12 and Part 13 together

That is Punjab's actual position, and I want to state what it does and does not amount to.

The sentence this chapter exists for

Fifty-five years of Punjabi political argument have been conducted almost entirely about the first question — whether Punjab should be inside India — while the answer to the second question, whether there will be usable water in Punjab, was being decided by tubewells nobody was counting.

9What would actually change it

This is the only chapter in thirteen parts where I move from what the evidence shows to what I think should follow, and I want that clearly marked. It is short, it is the weakest chapter in this work, and you should treat it as an argument rather than a finding.

Four things the evidence points toward

One. The water has to be priced, or at least measured.

Free electricity for 14 lakh tubewells is the mechanism producing 156% extraction. As long as pumping is free at the margin, no farmer has a reason to pump less, and Chapter 3 showed that the rational individual choice is the destructive collective one.

I am not going to pretend this is easy or that I would vote for it. Removing the subsidy outright would devastate small farmers already carrying debt at 17–21%, and no Punjab government could do it and survive an election — which is precisely why none has. But schemes that pay farmers for the power they do not use, and metering that measures without immediately charging, have been tried elsewhere in India and are at least a starting point. The honest position is that this is the highest-leverage change available and the least politically achievable, and that everything else is smaller.

Two. The guarantee has to cover something other than wheat and paddy.

Every diversification campaign for forty years has failed for the reason in Chapter 3: a household servicing debt cannot trade a certainty for a market. Diversification is not a persuasion problem. Until the assured purchase extends to maize, pulses, oilseeds or horticulture at a scale that makes switching safe, the wheat–paddy cycle will continue, because it is rational.

This is a decision for the Government of India rather than for Punjab, which is itself worth noticing.

Three. The 1984 file should be closed, and the argument for closing it is not sentimental.

Part 11 found that the diaspora movement's most effective material is not its own programme, which persuades almost nobody, but the documented conduct of the Indian state between 1984 and 1995. Part 11 also found that the misuse of the word "Khalistani" against ordinary Punjabis supplies live evidence for the movement's core proposition.

So the case for completing the 1984 prosecutions, and for treating Sikh political dissent as ordinary political dissent, is not that it would be gracious. It is that these are the two things keeping a movement alive that cannot win an election in Punjab. A state that wants fewer separatists has a direct, calculable interest in both.

Four. The demand that would actually change Punjab's position is about water, not borders.

Part 12's finding was that a separate Punjab would have less water security than Punjab has now, because the storage is in Himachal Pradesh either way and secession removes every legal mechanism for arguing about it.

The inverse of that finding is the constructive one. Punjab's water position is contestable inside the federation — in the Supreme Court, in the finance commissions, in the Bhakra Beas Management Board, at the ballot box, in a national politics where Punjab has thirteen parliamentary seats and a great deal of moral standing it does not currently spend on this.

What I am not saying

I am not saying any of this will happen, or that Punjab's politics is currently capable of it. Chapter 7 found that none of the four parties has a water plan and that the 2027 election is unlikely to be fought on one.

Nor am I saying these four items are sufficient. They are the four the evidence in this work points to most directly. Somebody with a better grasp of agricultural economics, or water engineering, or Indian federal finance, would produce a better list, and I would rather read theirs than defend mine.

10The final ledger: thirteen parts, one table

This is what the whole work was for. Every major claim tested across thirteen parts, in one place, tagged. Each row is a summary of a verdict reached in the part named — the reasoning and the sources are there, not here.

How to read the tags

The ledger

The four sentences that carry the whole work

11What I got wrong, and how to check me

Part 1 promised that I would show you the range, include what damages my own case, and refuse to write propaganda in either direction. This chapter is where I account for how well I kept that, and where I tell you the weakest points so you can attack them properly.

Where this work is weakest

What I changed my mind about

Three things, and I record them because Part 1 said I started this to find out rather than to prove.

I expected the ISI to be more central than it is. I began Part 10 assuming Pakistan's role would explain a great deal about the militancy, because that is the account I grew up hearing. The documentary record supports arming, training and sanctuary — and it does not support causation. Every cause was domestic. That finding changed how I wrote Parts 10 and 11 both.

I expected the status quo to be more comfortable than it is. I have said this in the author's note and I will repeat it here. I intended Part 13 to be a chapter about difficulties. The CGWB projection turned it into something else, and the argument between Part 12 and Part 13 is not the one I thought I would be writing.

And I did not expect Part 11's central finding. I assumed the diaspora movement was sustained by nostalgia and foreign money. The evidence says it is sustained by an Indian grievance file that has never been closed, and by a word used against Sikhs who want nothing to do with it. The most effective recruiter for Sikh separatism is not in Lahore or Surrey.

How to check me

Sources are listed at the end of every part. That is not decoration. Here is the practical method.

The last thing

Part 1 opened by saying that everybody in Punjab has an opinion about Punjab and almost nobody arguing has read anything — and that I included myself in that.

Thirteen parts later I want to be careful about what I claim to have earned. I have read a great deal more than I had. I have found that most of the confident things I was told growing up were wrong in one direction or the other, and that the correction almost never lands where either side would like.

The Indian state did things in Punjab it has never accounted for, and its own commissions and courts say so. The movement that claimed to answer that did things it has never accounted for either, and killed a great many Punjabis while claiming to speak for them. And underneath both arguments, quietly, for sixty years, the water has been going.

If this work is useful for anything, I would like it to be this: that it becomes slightly harder, for anyone who has read it, to say a simple thing about Punjab.