I’m intrigued, but not convinced, by Jonathan Hume’s analysis of the Work Capability Assessment in the new edition of Radical Statistics. He identifies a significant bias against claimants in poorer and less healthy areas. “Counter-intuitively”, he writes,
healthier areas were finding claimants fit for work less frequently and placing claimants into the Support Group more frequently. This is the opposite of what would be expected of an accurate test of disability.
I’m not sure that his expectations are right. The numbers he’s been processing are not about what happens to people with disabilities, but what happens after claims for Employment and Support Allowance, which is a long-term incapacity benefit. In areas of higher employment, there should be fewer people of working age with disabilities (because people with disabilities tend to suffer lower long term incomes and can’t afford to live in the more salubrious areas), but the disabilities which do lead to ESA claims will tend to be more serious, because people with lesser disabilities have more opportunities for employment than they do in poorer areas.
What he does establish, however, is that there are unexplained variations between areas, and clear biases in the outcomes of WCAs – for example, that more people are placed in the Support Group when more WCAs are done. The WCA was developed on the basis of a population-based assessment, and if it’s producing inconsistent results at the level of the population, that does raise questions about its validity.