The Chain of Proof: Pakistan's Price Data, Blockchain Attestation, and the Lesson of One Wrong Label

Two kinds of sentences sit side by side on the same page of the weekly bulletin on Pakistan's Sensitive Price Indicator. One carries a number — for the week ended September 24, the index rose 11.92 percent year on year, after 8.35, 8.62 and 10.64 percent in the three preceding weeks. The other carries a cause — a prolonged US-Iran conflict, elevated Brent crude prices, uncertainty around the Strait of Hormuz.

The first sentence has the Pakistan Bureau of Statistics behind it. The second has nothing. Every other price in the document — fuel, food, electricity — carries an attribution, while the causal sentences sit next to a blank source field. An official agency's arithmetic and an anonymous claim, printed together, weighted almost the same.

There is an old habit at my desk. Before I write a number, I ask who touched it, when, and whether it has changed since. In 2026, as Dhaka's new sports media was inflating, I watched 23 full match tapes and built a log where minute, event and the touch behind the event sat in separate columns. The habit has survived. A blank source field stops my hand — and that stop is why this piece exists.

Let me define the term first, because the story hides inside the word. The Sensitive Price Indicator is a weekly Pakistani measure: how much the prices of a fixed basket of essential goods moved from one week to the next. It is narrower than the monthly Consumer Price Index but faster, arriving weekly, and is read as an early inflation signal.

The contents of the basket are the real news. Liquefied petroleum gas, first-quarter electricity charges, diesel, petrol, onions, wheat flour, chilli, mutton, beef, fresh milk, plain bread, eggs, garlic, tomatoes, potatoes. Fuel, transport and the kitchen sit on one list.

Reading the average alone misleads. The bulletin itself notes that the burden falls unevenly across income groups. For poorer households the basket is weighted more heavily toward food, so the same index is much heavier to carry. The structure of the index matters more here than the trend.

Now to the part that forced this piece to be written. The confusion is about where the document actually stands. Its content says macroeconomics — indices, fuel, food, oil markets. Yet the imprint of the editorial pipeline it came from says football. There is no player in it, no club, no match, no tactical system. The tag is wrong in several places.

Then came a further request: write a blockchain news article based on this file. Three worlds — inflation, football, blockchain — now stand around a single sheet of paper. That mismatch is the most honest story here, because it is the modern face of journalism's oldest problem: information and evidence are not the same thing.

A chain does not make information true; it makes information's history readable.

The core mechanism of blockchain is simple: a ledger to which entries can be added, from which they cannot be deleted, each entry cryptographically bound to the one before it. That binding is the only real asset. It does not verify whether a statement is true, but it leaves the answer to who wrote it, when, and whether it changed since, open to everyone.

Imagine the Pakistan Bureau of Statistics publishing not the whole data block on-chain, but a single hash each week. Every release would get a unique cryptographic fingerprint. If someone later altered the number, cut a paragraph, or let a pipeline apply a wrong tag, the hash would fail to match. It would be caught — publicly, not factionally.

Attestation does not let the source field stay empty.

The practical fix is cheap: writing a 32-byte hash on-chain costs cents to a few dollars, and a statistics office does not need to run a validator. The imprint remains whether or not anyone asks for it.

A chain also records time, and that is where my real interest lies. Take the anonymous sentence about oil prices. If a daily crude price feed and the weekly index hash both sat on-chain, a reader could see in one place when oil moved and when the basket moved. Whether the sequence holds is the actual test.

This raises the oracle question. Bringing outside data on-chain requires an intermediary, and what that intermediary says is what gets written. A centralized feed speaks for one party; a decentralized network reconciles several independent sources into one value. But be clear: an oracle proves what a source said, not whether the source is right.

My habit of more than thirty years says that if the temporal order could be checked, half the causal sentences in economic journalism would die on their own. Oil rose, then the index rose — that is one claim. Oil rose in the same week the index rose, while supply contracts run for six months — that is another claim, carrying an entirely different meaning.

We need to go down to the roots of local supply, because index numbers do not float in the sky. Wheat flour, onions, chilli, potatoes arrive through trucks, weighbridges, cold storage and ports. From Karachi port to mill, mill to wholesale, wholesale to retail — if every step of that root system holds a record that cannot be altered, the story of prices stops being a guess.

This is where blockchain's real work is happening, not in stock-market excitement but in paperwork: electronic bills of lading, trade finance documents, port-to-port shipment records. Chattogram and Karachi both share the same problem — a gap between documents and cargo. Not a game; dirty work. Real change happens there.

Oil prices do not only hit the basket. When demand shifts in Gulf construction and services, remittance flows to Pakistan and Bangladesh shift with it. The same oil price reaches the household plate and the migrant family's kitchen at once. If cross-border payment flows are recorded on a chain, that link stops being an assumption and becomes a record.

Still, limits must be respected. Bangladesh Bank has repeatedly warned against crypto transactions, and Pakistani regulators have taken hard positions at times. The realistic form is therefore not retail tokens but bank-run permissioned systems, where the chain works as nothing more than a ledger of accounts.

One more illusion to clear up about tokenization: turning wheat into a token does not move the wheat an inch. In a supply chain, the value of the chain lies in the documents, not the cargo. The truck stuck at a border is not fixed by cryptography.

Costs matter too. Writing tokens consumes electricity, and that electricity charge sits in the index basket itself. Amid Pakistan's load-shedding, Dhaka's heat and the heavy monsoon clouds, a government office that tries to run a node will stall on capacity before it stalls on money. The smaller path is more promising: a small imprint, or a permissioned private ledger.

The labelling question deserves separate handling, because the error is reproducible. Once a pipeline tags a commercial report as football, the same error returns in the next batch, and each contaminated batch yields a fabricated fragment of analysis. If a chain held an audit trail of label changes — who changed it, when, under what rule — the error's birthplace would become public, and the contamination would stop before spreading downstream.

Now the most important question. The obvious reading is that putting everything on-chain restores trust. The evidence points the other way. The failure we observed was not tampering. Nobody altered the index number. There were two weaknesses: an unsourced causal sentence and a wrong tag. Hash a wrong tag and it becomes a permanent, cryptographically protected wrong tag.

A chain proves the envelope was never opened; it has nothing to say about whether the letter is true. The simplest deception happens outside the chain, in the notebook of a vendor walking a retail market whose name we do not know. An empty stadium still has a pulse in the grass if you listen; a blank source field is not silent either — it says nobody stands behind this claim.

One more thing to hold in mind: a ledger nobody reads does not exist. If an office publishes a hash but no journalist, researcher or official ever checks it, the whole exercise becomes decoration. Technology only creates the space; the checking has to be done by people, and wanting to do that checking is the real measure.

If someone had added a single hash to next week's release, with the daily crude price feed attached, the first beneficiary would not be an investor. A wholesaler in Rawalpindi, or a bus stand in Chattogram, might have grasped where the Strait of Hormuz sentence actually came from — or that it came from nowhere.

Fourteen seconds is long enough for a whole career to change boots; four weeks are long enough for an index to change direction. In evidence, these small intervals of time are what matter. Before believing information, you need to know where it was and who touched it. A technology that keeps only the record of that touch is the technology we actually need.