If John Healey wants to succeed as Chancellor, he needs to get a grip on the Office for National Statistics (ONS).
I’ve written about the troubles and tribulations of our hapless Newport-based statistics office many times now, both in this newsletter and the magazine, and the problems keep coming. Its most high-profile failure of recent years has been the delivery of the Labour Force Survey (LFS). That’s the main source of our knowledge about Britain’s jobs market: unemployment, inactivity, redundancies etc etc.
During the pandemic, response rates to the survey collapsed and the integrity of the data it produced was diminished. That set off a chain of errors. As I’ve noted before, a decade ago, more than half of households approached took part; during lockdown that fell to 27 per cent and it has crept back to only around a third since. The sample has shrunk so badly that figures for one entire industry had to be suppressed after the number of respondents fell to just five people.
Since then, the ONS has done lots of work to improve the survey. But, yet again, this week’s job market figures have been hit by errors. In a quality update released with Tuesday’s labour market data, the ONS admitted an ‘operational error’ that ‘should have been detected earlier’ led to its telephone interviewing operation being understaffed for six weeks. That meant contact rates dropped and the missing data had to be filled in with guesswork. The impact, the ONS insists, is ‘minimal’ – mainly affecting the data on hours worked, which drive productivity figures.
Now, each individual error, mishap and mistake may be ‘minimal’, but the greater impact is not. The net effect is that we have been left without the tools to properly measure the health of our economy. That mattered for the political fortunes of the last Labour administration and will dictate the fate of this one too.
Take employment: the accepted picture is that Rachel Reeves’s early interventions destroyed jobs specifically among the young. That’s what our most reliable payroll data – HMRC’s Real Time Information – shows anyway, with 163,000 payrolled jobs disappearing since Labour came to power. But compare it with the LFS – as the below graph shows – and jobs are up by nearly 780,000. Which is it?
There’s plenty of other evidence, not least the evidence of our eyes and ears, that the HMRC payroll data tells the true picture. However, if I were the Chancellor, I’d be hammering home the inconsistency of this data, not least to take the sting out of Opposition attacks.
Private data suggests the last government might have had a brighter economic story to tell too. Look at Companies House incorporations of new start-ups and it would suggest that entrepreneurship in Britain is essentially dead. But compare that with data from Stripe – a payment processor – on UK sign-ups to its platform and it would suggest we’re actually going through a bit of an entrepreneurial boom. Which would make sense given how easy AI tools now make it to start and run a business.
But the government seems disinterested in getting to the truth of this matter. Which is puzzling because there are clear incentives for them to rebuild our statistical infrastructure. Worst case: the negative picture of the British business environment is true and ministers can try to undo the damage with clear sight of the results. But best case? They might have a politically game-changing, positive picture to paint of Labour’s rule over Britain's economy. Either way, fly blind and you can't complain when you hit the mountain.
Comments