Corruption Is Gravity. Build a Rocket.

We have spent seventy-eight years trying to remove it. We never removed gravity either. We built rockets.

I. The most honest police officer in the country

He exists. I’m certain of it. Somewhere right now there is a sub-inspector who genuinely does not want a single rupee he has not earned.

Here is his week.

His station has a monthly number that travels upward. It is written down nowhere and known by everyone. His posting was decided by someone who paid for the right to decide it, and his transfer next year will be decided the same way. His constables earn less than the men they are collecting from. His budget for fuel runs out in the third week of every month, so the vehicle runs on money that came from somewhere.

If he refuses all of it, he does not clean up his station. He is moved somewhere he cannot affect anything, and someone else is in his chair by Friday.

He is not a bad man. He is a particle in a field.

This sounds like an excuse, which is why almost nobody says it out loud. It is not an excuse. It is a spec sheet.

And here is the proof that it is a field and not a character flaw. Take the same person and put him in a queue in Delhi: he finds the gap. Put him in a queue in Singapore: he stands in it. Same man, same morals, same mother. Different field.

We have spent three generations trying to fix the man. The man is fine. The field is the problem.

We keep trying to fix the man. The man is fine. The field is the problem.

II. Corruption is gravity

Not like gravity. I mean it behaves the same way, obeys the same shape of law, and once you see the correspondence you cannot unsee it.

Gravity is the weakest force in physics. Genuinely the weakest of the four, by an almost comic margin, roughly thirty-six orders of magnitude weaker than electromagnetism. A fridge magnet defeats the entire mass of planet Earth. You lift a paperclip and you beat a planet.

And yet gravity runs the universe. It decides where galaxies go, where stars form, whether a planet holds an atmosphere. Because it is weak, but it is everywhere, it never switches off, it only ever attracts, and it adds up.

One clerk asking for five hundred rupees is a fridge magnet. It is nothing. It is not worth a news cycle. Aggregate it across a billion people, every day, for eight decades, and it determines the trajectory of a civilisation.

The second correspondence is stronger than the first. Gravity works in gradients, not in absolutes. The constable orbits the SHO. The SHO orbits the commissioner. The commissioner orbits the minister. Earth does not obey the Sun because it agrees with the Sun. It obeys because of mass.

Everyone in the system is simultaneously being pulled and pulling. There is no vantage point from which anyone is simply an actor with free choice. The officer who takes is also the officer who gives. This is why the language of villains fails so completely. There are no villains in a gravitational field. There are only masses and distances.

III. The thing we want to delete is load-bearing

So remove it, you say.

You cannot. And not for the reason you assume.

You cannot remove gravity from the universe because gravity is what holds the universe together. Delete it and everything disperses. No galaxies, no stars, no planets, no chemistry, no you. The force that traps us is the same force that assembled us.

The political version is exactly as tight. The thing we call corruption and the thing we call ambition are the same substance at different concentrations. Self-interest is what makes a man wake at five, open a shop, take a loan, hire his cousin, build something. Any mechanism powerful enough to fully suppress the first will suppress the second, because at the level of the incentive there is no way to tell them apart.

Run the thought experiment honestly. Suppose we win completely. Tomorrow morning everyone becomes a monk, perfectly, incorruptibly selfless.

Nothing gets built. Nobody starts anything. No enterprise, no risk, no invention, no wealth. A very peaceful society, slowly starving.

Which means the honest formulation of our problem is not how do we remove this. It is:

The thing we want to delete is load-bearing. What do we build that works in spite of it?

IV. Both of our available answers are the same answer

This is where I lose interest in our politics entirely, and I think it is why a lot of capable people have quietly checked out.

The left’s answer to corruption is: add state. More regulation, more oversight, more agencies, more agencies watching the agencies. A vigilance commission. Then a Lokpal. Then Lokayuktas. Then a commission to review the commissions.

The right’s answer is: remove state. Deregulate, privatise, let the market discipline it. Which in practice does not remove mass from the system. It transfers it to whichever private bodies are already largest.

Both are the same move in different clothing: add weight to the vehicle and hope.

Consider what we have actually built. The CVC. The CBI. The ED. State anti-corruption bureaus. Lokayuktas. The Lokpal, demanded by a national movement, legislated in 2013, and left without a chairperson until 2019. Each was announced as the answer. Each is now itself a body whose independence, appointments and selective attention are argued about.

This is not a failure of implementation. It is what the mathematics predicted. Which brings us to the equation.

V. We never solved gravity

Read that sentence again, because everything turns on it.

Newton did not solve gravity. He described it. Einstein explained it more beautifully than anyone has explained anything, and removed precisely none of it. There is exactly as much gravity today as there was in 1600.

And yet there are human beings in orbit above your head right now.

We did not remove the constraint. We built machines that work in spite of it.

And here is the part most people get wrong, which turns out to matter enormously for the politics: a spacecraft in orbit has not escaped gravity. At the space station’s altitude, roughly 400 kilometres up, gravity is still about ninety percent as strong as it is where you are sitting.

Orbit is not escape. Orbit is falling, and moving sideways fast enough that you keep missing the ground.

That is the actual target, and it is a much more honest one than any manifesto currently offers. Not a country without corruption. A country moving sideways fast enough that corruption never lands.

And note the economics, because they match what we observe about institutional change. Reaching orbit is brutally, almost obscenely expensive. Staying there is nearly free. The cost is entirely front-loaded into the mechanism. Once the mechanism exists, the marginal cost of the next million honest transactions is approximately zero.

VI. The rocket equation as political theory

One equation governs all spaceflight. Tsiolkovsky published it in 1903, and it is the most politically useful thing I have ever read.

Δv = vₑ · ln(m₀ / m_f)

Two things determine how far you get. How much fuel you carry, and how fast you throw it out the back.

And the cruelty of it is this: the fuel term is a logarithm. The exhaust velocity term is linear.

Doubling your fuel barely helps. To go meaningfully further on fuel alone you need absurdly, ruinously more mass, and that mass needs structure to hold it, which is more mass, which needs more fuel. This is why single-stage-to-orbit vehicles keep failing on paper. It is a treadmill with logarithmic returns.

But improve the engine, throw the same mass out faster, and you get a straight, honest, linear return.

Now translate.

Fuel mass is every anti-corruption commission, every new statute, every additional layer of audit, every “stricter punishment,” every oversight body. It is enforcement. It scales by adding people who must themselves be honest, supervised by people who must themselves be honest. Logarithmic returns. We have been strapping on tanks since 1947 and going roughly nowhere, which is exactly what the equation predicts.

Exhaust velocity is a mechanism that makes one unit of honest effort go dramatically further. It does not ask anyone to be better. It changes what one honest action is worth.

The distinction is sharp enough to be operational: does this proposal require more good people, or does it make each good person count for more?

Almost everything in Indian public debate is fuel.

VII. We have already built rockets

This is not theory. We have done it before, more than once, and never once by giving a speech about morality.

The telephone. For most of independent India, getting a line meant a waiting list measured in years and a payment to somebody willing to move your name up it. Nobody solved that by disciplining the linemen or auditing the exchange. A different technology arrived, the copper stopped mattering, and an entire economy of small extortion evaporated with it. Note what did not happen: nobody’s character improved.

The reservation counter. Booking a train seat once meant a clerk with a ledger and a tout who knew which clerk. Computerising the reservation did not reform either of them. It deleted the ledger, and with it the discretion the whole arrangement fed on. That is the purest expression of the principle available: do not police the choke point, remove it.

RTI, 2005. It did not make officials honest. It made opacity expensive. One citizen, one application, one statutory clock. Exhaust velocity in legislative form.

Boring. Mechanical. Unromantic. Worth more, collectively, than every speech given in the same period.

And now the caveats, because a mechanism-first politics that cannot audit its own mechanisms is just a different flavour of faith. Digitising a public service reliably produces exclusion errors: people who are entitled and are denied because the system does not recognise them, in some cases catastrophically. Digitisation also moves rent rather than ending it, from the counter to the operator, from the file to the database — automated booking now takes scarce railway inventory in seconds, which is the tout problem rewritten in software. RTI has been weakened by vacancies, delays and amendments to the commissioners’ terms. And telecom, having dissolved one choke point, has concentrated into a handful of private networks, which is a mass problem forming in real time.

None of that refutes the argument. It is the argument. Rockets fail. The response to a failed launch is not to declare spaceflight impossible; it is to read the telemetry. What is unforgivable is refusing to look at it because the vehicle carries your name.

VIII. Staging: the law that kills revolutions

One more piece of rocketry, and it is the one that will make people angriest.

No rocket in history has reached orbit in one piece.

You build an enormous, powerful first stage. It does the hardest work, lifting everything through the thickest atmosphere. And then, while it is still perfectly functional, you throw it away. Because the machine that got you off the ground becomes the single heaviest thing holding you down.

Every reform movement dies at precisely this step.

The movement that overthrows the system becomes the system. The freedom fighters become the establishment. The anti-corruption agitation becomes a party, the party becomes an incumbent, the incumbent becomes a mass that other people must now orbit. This is not one country’s failing and it is not a story about anyone’s bad faith. It is the observed default across every country and century you care to check. The people are usually sincere. The physics is indifferent to sincerity.

They kept stage one bolted on. Nobody ever wants to jettison the thing that worked.

Which yields the most uncomfortable design requirement in this entire essay: anything worth building must be engineered to discard its own founders. Not as a virtue. As a structural component, specified before launch, with a date on it, because it will never be voluntarily added later by the people it applies to.

Including us. Especially us.

The machine that got you off the ground becomes the heaviest thing holding you down.

IX. Where this argument is wrong

Let me attack my own metaphor, because I would rather point at the hole than have you find it.

Gravity is passive. It does not adapt. It does not notice your rocket, study it, and grow toward it. Corruption does. It is not a field so much as a predator field, with something like an immune response, and it learns faster than legislatures do.

The historical record is not encouraging. The free press was a rocket; it was bought. The joint-stock company was a rocket; it produced the East India Company. The internet was a rocket; five firms own the launchpad. Crypto was going to be the rocket, and produced new kings in under a decade.

And the worst version of the failure, which applies to me and to anyone reading this with ambitions: build a large enough rocket and it becomes its own gravity well. Every successful escape vehicle eventually becomes the thing the next generation needs to escape from. That is not a risk of this project. It is the base rate.

I do not have a clean answer to this. What I have is a design posture: assume capture. Assume that the thing we build will be corrupted, and that we will be the ones who corrupt it, and that we will have excellent reasons at the time. Then build the staging in before anybody needs it, while it is still cheap and nobody’s interests are attached.

If we ever stop saying this section out loud, stop reading us.

X. Democratically Elected System Hackers

Which is why the thing we are starting is called what it is.

Desh, because it is the country. And because it stands for Democratically Elected System Hackers.

Two words in that need defending.

Hacker. The word has been ruined, so let me be precise. A hacker is not a criminal. A hacker is someone who reads a system, works out what it actually does as opposed to what it claims to do, locates the gap between the specification and the reality, and reroutes it. Hackers do not argue with systems and they do not protest them. They read the manual more carefully than the people who wrote it.

That is a stance almost entirely absent from public life, which is saturated with people who have opinions about outcomes and nearly empty of people who have read the mechanism.

Democratically elected. This is not decoration and it is not a hedge. We go through the front door. Every attempt in history to fix a country from outside its own consent has ended somewhere ugly, and it has ended there reliably enough to count as a law. Legitimacy is not a constraint on the rocket. Legitimacy is the launchpad. There is no other surface stable enough to launch from.

Not left. Not right. Forward.

Left and right are directions on a flat map. We are not arguing about which way to walk across the plain. We are pointing out that the interesting direction is up, and that it has been available the entire time.

XI. What actually counts as a system hack

A metaphor that cannot be turned into a test is just a nice sentence. So here is the test. Every proposal we publish, and every proposal you send us, has to survive five questions.

1. Does it raise exhaust velocity, or add fuel mass? Does it require more honest people, or does it make each honest act count for more? If the answer is “more inspectors,” it is fuel. Reject it.

2. Does it remove the choke point, or police it? Corruption needs a bottleneck where someone can stand. Guarding the bottleneck creates a new position worth capturing. Deleting the bottleneck ends the game. Almost everything currently proposed guards it.

3. Does it survive bad people? Assume everyone operating the mechanism is entirely self-interested. If it still produces the right outcome, it is a machine. If it only works when staffed by good people, it is a prayer. We only publish machines.

4. Are the economics orbital? Expensive to build, near-free to maintain. If it needs permanent political will to keep functioning, it will die the first time attention moves, and attention always moves.

5. Does it have a staging plan? What is the mechanism by which this thing is dismantled or handed off when it becomes the heaviest mass in the room? If there is no answer, it is a future problem wearing the costume of a solution.

Five tests. Most of what is currently proposed fails at the first one.

What this is

This is not a place for opinions. There is no shortage of opinions.

Every piece we publish takes one mechanism, one specific, boring, load-bearing component of how a country actually runs, and opens it up like engineers. What is it supposed to do. What does it actually do. Where is the gap. What would a better engine look like. And then the five tests, in public, including on our own ideas.

Next: sortition. Selecting people for public office by lot. It sounds insane. Athens ran on it for the better part of two centuries, filling most of its offices and its entire Council of Five Hundred by lottery, using a stone machine built for the purpose. It is the most rocket-shaped idea in political science, because it does not fight corruption’s strength at all. It makes the targeting problem intractable. You cannot build a gradient toward someone who cannot be identified in advance and who rotates out before a gradient can form.

If you build things, code, companies, policy, infrastructure, anything with a spec, this is for you. We do not need more people who are angry. We have those in abundance, and anger is fuel mass. We need people who read specs.

Tell us where this is wrong. That is not a courtesy. It is the method.

Gravity is still there.

Let’s build the rocket.

desh>_