DWP figures on benefit claimant deaths: what they do and don't tell us
The Department for Work and Pensions (DWP) has today published mortality statistics for out of work benefits claimants and disability and sickness related benefits.
The Information Commissioner ruled in April that the DWP had to publish the number of Incapacity Benefit and ESA claimants who died from Nov 2011 until May 2014 by 'category'—those in the assessment phase, who were found fit for work, put in the work related activity group, put in the support group, and who had an appeal pending.
Today's figures show the number of people who died while on the benefits or up to a year after coming off them. They also show the age-standardised mortality rates (ASMRs), which are a way of comparing the chance of a claimant dying with the chance of someone in the general population dying.
It's good to see that the releases include a methodology section explaining the calculations, plus data tables allowing people to scrutinise the analysis more closely.
We should be careful in interpreting these figures
The figures show some interesting results. In 2013, about 140 out of every 100,000 Jobseeker's Allowance (JSA) claimants died compared with 240 people in the general population. What we shouldn't infer from this is that claiming JSA lowers your chance of dying.
Similarly, the mortality rate for people claiming Employment and Support Allowance (ESA) who are put in the work related activity group is more than twice as high as the mortality rate for the general population, at 530 per 100,000. This doesn't mean that being placed in the work related activity group causes higher mortality; another explanation is that the people put in this group are more likely to have conditions which raise their mortality rates (allowing them to claim ESA) but which aren't severe enough to qualify them for the support group.
Or it could be a bit of both—we simply don't know from these figures.
How age-standardised mortality rates work
A population's mortality rate is the number of deaths that occur per 100,000 people over a year.
If we want to compare mortality rates between two groups of people, it makes sense to adjust for differences in the age and gender of people in the two groups. If we don't do this, then differences in mortality might just be driven by (for example) one population being older than the other.
Age-standardised mortality rates for 15-64 year olds allow for a better comparison between the general population and benefits claimants.
They work by taking the mortality rates for benefits claimants (split by age group and gender), and then calculating what the overall mortality rate would be if these rates were applied to the appropriate groups in an average population.
The figures present ASMRs for both the general population and benefits claimants, comparing them to an average European population.
Update 25 Sep 2015
The line reading "Or it could be a bit of both—we simply don't know" has been changed to read "Or it could be a bit of both—we simply don't know from these figures" to emphasise that we were referring to the limitations of these figures.
