This article is part of The Statistician’s ongoing commitment to data-driven journalism. The numbers behind the headlines are rarely as simple as they appear — and rarely as complicated as those in power would have you believe. Statistics are not neutral. They are chosen, framed, and published by institutions with interests. Our job is to read them critically, place them in context, and ask the questions that the raw figures cannot answer on their own.

The story behind this piece begins, as most stories at The Statistician do, with a number that seemed too clean. A figure cited in a policy document, a statistic repeated across news cycles without examination, a data point that had become so familiar it had stopped being interrogated. We pulled the original source. We read the methodology. We spoke to the researchers who produced it and to the critics who challenged it. What we found was more complex — and more important — than the headline suggested.

Data journalism is not simply the act of including numbers in a piece of writing. It is the discipline of understanding where numbers come from, who produces them, what they measure, and crucially, what they leave out. Employment figures that exclude discouraged workers. Poverty lines drawn at levels that bear no relationship to actual cost of living. Crime statistics that reflect policing decisions as much as criminal behaviour. Growth metrics that count the production of weapons and the clean-up of oil spills as contributions to national prosperity. The statistic is never the whole story. It is the beginning of one.

At The Statistician, we believe that an informed public requires not just access to data, but the tools to read it. That means publishing our methodology alongside our conclusions. It means acknowledging uncertainty where it exists rather than projecting false confidence. It means returning to stories when new data emerges, even when that data complicates an earlier narrative. And it means treating our readers as people capable of handling complexity — because the alternative, the simplified story, the clean number, the confident headline, is not journalism. It is noise dressed as signal.

This article will be updated as new data becomes available. All sources are linked in the footnotes. The methodology used to analyse the primary dataset is available in full on request.