Britain's record net migration rose to 906,000 in 2023 and it then declined to 431,000 in 2024, according to migrationwatchuk.org.... or did it...?!?
In "Making policy in a statistical vacuum", Michael Simmons draws attention to the core weakness of Britain's policymaking: reliance on unreliable data.
Referring to a 2014 ONS admission that it had overshot migrants due to gaps in airport surveys at places like Luton, Simmons underlines how net migration figures were quietly revised by 166,000 to 906,000. These were figures no minister could reasonably have relied on.
As Full Fact observes, migration statistics are always subject to revision and revisions, sometimes major, occur regularly as new data emerge. Far from being evidence of failure, these revisions are an inevitable feature of evolving datasets.
The problem we have is that if evisions fundamentally rewrite previous data, as they have done in the past, then the public and policymakers have no idea what the actual migration figures are?!?
Improving Migration Statistics...?
Efforts are being made to address these statistical shortcomings.
The Office for Statistical Regulation (OSR) completed an overall review in 2022 (a review of revisions!) , urging the ONS to publish more details about its methodology, publish the size of uncertainty, and broaden user consultations, and the ONS has essentially complied with all of this.
Somewhat ironically the ONS' efforts to get better may be creating more uncertainty, as new methodologies get bedded in, but make it difficult to compare with previous data.
An independent review of the ONS in 2025 identified systemic flaws due to planning and budgeting shortcomings which were preventing consistent performance.
In short: what a mess!
And there's fall out....!
Shoddy data generates political distrust. If people learn about "statistical black holes," runaway revisions, or farcical registration irregularities, faith in both data and decision-making disintegrates, and this is exactly what Simmons warned against.
To regain credibility, the ONS must at the very least be clear about how much unxertainty there is within the data.
I mean if we must make policy based on uncertainty, at least let's be clear about this!
Immigration is tricky to measure, let's just be honest about it!