Does anyone know how many people live in Britain?

Can Britain trust its economic statistics? The nation’s arbiters of numerical truth, the Office for National Statistics, yesterday released what on the face of it was good news for the Home Office and a vindication of the previous Conservative government’s policies to reduce worker visas and the number of dependants of migrants arriving in the UK. But in truth – and in the same data dump – the previous year’s figure had been revised up so much (by 307,000) that had it not been, the net migration figure published yesterday would have matched the previous record high.

These revisions matter. Douglas McWilliams, founder of the Centre for Economics and Business Research, calculates that the new figures mean the previous June population estimate changes by a large enough amount that the second quarter fall in GDP per capita may be twice as bad as officially estimated (from 0.3 per cent to 0.6 per cent). Changes in population impact how we view our economic history.

Of course, there’s always going to be a margin of error when it comes to migration – or any statistical estimates, for that matter. And huge underestimates in migration figures are nothing new. The campaign group ‘the3million’ operated on figures that suggested there were three million EU citizens living in Britain during Brexit. But in the end, the true figure turned out to be closer to six million.

During the pandemic, too, vaccine data suggested we didn’t know how many people lived in the country. Vaccination coverage rates in some areas reached more than 100 per cent. How? Because more people were vaccinated than existed in the official population figures used to calculate the rates: clear proof that the UK population has repeatedly been undercounted. Other records from the NHS have found that, as of September, there were 63 million patients registered to GPs in England – yet the official population is 57 million.

Jobs data – arguably far more important to Britain’s economic prospects – is potentially error-strewn, too. Swati Dhingra, a member of the Bank of England’s rate-setting committee, criticised the country’s labour market statistics this week, saying: ‘I grew up in India. There are a billion people there. We managed to get the labour force survey answered… I don’t find it particularly plausible that that’s hard to do.’

Her comments refer to problems with the ONS’s labour force survey, which has struggled, since the lockdowns, with poor response rates which render the data produced from the survey unreliable. ONS statisticians are working on a ‘transformed’ version of the crucial survey, but reports suggest it could take until 2027 for the improved version to come into effect. It could be that there may be more staff in the workforce than official figures suggest.

Some blame the ONS’s move to Wales in 2007 for the chaos. When it happened, nine in ten London-based staff chose to leave their jobs rather than move to Newport. A review by economist Sir Charles Bean found that ‘the loss of statistical expertise which resulted from the relocation decision has had a significant – though not necessarily permanent – detrimental effect on the capability of ONS and the quality of its output’. I think this was unfair then, and now: the ONS staff in Newport are some of the best and brightest working in their field (though I should declare that my first foray into serious employment was at the National Records of Scotland – the ONS’s tartan sister).

The issue, I think, is in large part due to changes in working practices brought about by the pandemic. The ONS stopped going door to door for many of its surveys, relying instead on online responses and telephone calls. In Scotland, this approach led to what many would call a completely failed Census. A year-long delay and an over-reliance on online, rather than in-person, methodologies led to Scotland’s census being released 444 days after England’s, and with a response rate of just 89 per cent – well short of the 94 and 97 per cent achieved in England and Wales respectively.

For policymakers, rate setters and businesses to make proper decisions to guide our country’s economy we need robust, reliable and trusted statistics. Investment may be necessary for the ONS to get boots back on the ground and return to more reliable door-knocking methods for surveys and data gathering. It won’t be as simple as that. But whatever the solution turns out to be, it is a problem that can no longer be ignored.