Ministers are busy debating whether to delay youth minimum wage increases, but there's a bigger problem hiding in their own data: Britain's workforce appears to have disappeared.

The Office for National Statistics' labour market overview shows a figure of 4.2 billion in 2021, down from 12.5 billion in 2020. That's not a typo. The government's own numbers suggest two-thirds of whatever they were measuring simply vanished during the pandemic year.

This isn't about unemployment rising or wages falling. This is about the basic counting going wrong. The dataset, which covers a decade of British labour market data, shows steady growth from 2017 to 2020, then this dramatic collapse that makes no economic sense.

Think about what this means. While politicians argue over whether 18-year-olds deserve a few extra pounds per hour, the statistical foundation for those debates might be fundamentally broken. How do you set minimum wage policy when you can't reliably count the market you're regulating?

The timing couldn't be worse. As inflation shows signs of cooling thanks to lower fuel and airfare costs, workers need accurate data to understand whether their wages are keeping pace. But if the government's labour statistics are this unreliable, those calculations become meaningless.

This isn't just a technical problem for statisticians. Real policies affecting millions of workers get built on this data. Employment support, training programmes, regional development funding. They all depend on having reliable numbers about who's working where and for how much.

The ONS hasn't explained what happened to cause this massive drop. Was it a methodology change? A data processing error? A pandemic-related adjustment that went wrong? Without answers, every labour market statistic from 2021 onwards becomes questionable.

Meanwhile, workers like the nurse who recently discovered she'd overpaid nearly £3,000 in tax are being told to check their payslips more carefully. Perhaps it's time for the government to check its own numbers with the same scrutiny.

You can't run an economy on broken statistics. And you certainly can't make informed policy about wages and employment when your basic counting appears to be wrong by billions.