The ONS scandal that has made your bills higher than they should be

Is there a bigger numerical oopsie than getting inflation wrong? Given how important this data point is, and how profoundly it affects all of us, it is hard to find one.

Yet this is what happened in April, also known as “misery month” because of the slew of tax and price rises that took effect and delivered a vicious kick in the guts to the average consumer.

These price rises were always expected to send the Consumer Prices Index (CPI) northwards, but not to the extent that the official data showed. The Office for National Statistics (ONS) put CPI inflation at 3.5 per cent, well ahead of the consensus City forecast (3.3 per cent).

The trouble is, it got its sums wrong. It has now emerged that the actual number was 3.4 per cent. The cause? Dodgy data on vehicle excise duty – car tax to you and me – submitted by the Department for Transport (DfT) and used in the calculation.

An error of 10 basis points mightn’t seem like such a big deal. It doesn’t look like a big deal on paper – but it really is. That number feeds through to businesses' pricing decisions and to wage settlements. It influences a dizzying array of contracts, too. The older Retail Prices Index (RPI) for the month was also wrong, coming in at 4.5 per cent when it should have been 4.4 per cent, meaning mobile phone contracts with built-in rises linked to RPI increased by more than they should have.

The biggest problem, however, is the key role CPI inflation plays in the formulation of monetary policy. The Bank of England’s target is set at two per cent. How are the rate-setters on its Monetary Policy Committee (MPC) expected to set interest rates to hit that if the numbers are constantly moving? Needless to say, getting interest rate policy wrong can choke off growth, leading to higher unemployment, and even recession. So this is no small SNAFU. It matters. It is, in fact, a scandal.

Wait a minute, I hear you say, isn’t the problem here with the DoT? Surely that’s where the red faces should be. Well, yes, but here’s the thing: the pointy heads who pore over this stuff when it comes out were ringing alarm bells over the DfT data before the ONS admitted to the error. If outside commentators could smell a rat then you would expect the same to be true of the ONS. At the very least, you would expect someone there to pick up the phone and ask the DoT to double-check its figures. The ONS admits as much – albeit en passant – when it says in its statement that it is “reviewing our quality assurance processes for external data sources in light of this issue.”

The Office for Statistics Regulation was clearly on the right track when, in a recent interim report on the economic stats put out by the ONS, it highlighted that staff had said “early warnings of emerging problems” were not always welcomed. This begs the question that, if someone did see something off, were they listened to?

The report was prompted by signs of institutional weakness. The unreliability of the labour force survey, closely watched by both the MPC and the government, has been a bone of contention for some time. Errors in trade figures were also identified from January 2023 to December 2024.

Andrew Bailey, the governor of the Bank of England, was asked about the impact of poor ONS data on interest rate policy at a hearing of the Treasury Committee on Tuesday. “It does have a bearing on it,” he said. “We certainly spend more time on it and that’s obviously what we should do, given the uncertainty.”

That might sound bland but central bankers choose their words carefully and with good reason: they move markets. It’s probably overegging it a bit to say that a comma in the wrong place in one of Mr Bailey’s speeches could create a crisis, but only a bit. Participants pore over his missives for clues on future interest rate policy. Billions of pounds can be moved if they sense a hardening or softening of the MPC’s stance. Behind the scenes, I suspect Bailey has been a good deal more blunt.

Despite the error, the ONS said it would not change the inflation figure “in line with policy” in this area. That policy is partly informed by the sheer number of contracts that depend on the CPI and/or the RPI. Things would get very complicated if money was moving back and forth every five minutes as a result of the inflation data being subject to constant revision. So there will be no mobile phone refunds. However, this also means the records will forever be wrong. Tables and graphs and suchlike will require an asterisk from hereon out.

This is an embarrassment when good data has never been so important. The ONS simply has to get its house in order. ONS chief Sir Ian Diamond resigned last month on health grounds. Stopping the rot must be at the top of his successor’s inbox.