Wilkinson and Pickett are, yes, still wrong


This would simply be laughable if it weren't for the fact that it's so dangerous. Richard Wilkinson and Kate Pickett are, once again, telling us that it's all inequality that fuels our woes. And yet they've entirely misunderstood the statistic that they're using to prove that this is so. And to add to the embarrassement Wilkinson at least is supposed to be a demographer, meaning that he really is supposed to know how badly he's cocking things up here. Yes, we're being rather fierce here but rightly so:

New statistics from the ONS have revealed that women in the most deprived areas of England can expect to have 19 fewer years of healthy life than those in the most advantaged areas. For men, the figure isn’t much better, with a gap of over 18 years. To put that into perspective, those born in the poorest parts of England can now expect to live the same, or fewer, healthy years as someone born in war-torn Liberia, Ethiopia or Rwanda. And a third of people in England won’t reach 60 in good health.

Those statistics are here. And the first part of the paragraph is correct. Ages at death in poorer areas, health before death in poorer areas, are lower/worse than they are in richer areas.

But this has absolutely nothing at all to do with life expectancy at time of birth. Simply because no one at all is even attempting to measure life expectancy at birth. What people are measuring is age at death, health before death, in certain areas.

The difference is, and you may have noticed this, people actually move around during their lives. Further, it is not (necessarily) true that income inequality leads to health inequality. For it is also true, as we've pointed out many a time before, that health inequality can and will lead to income inequality. That ghastly disease that cripples someone in their 40s is going to have an impact on their income during the remainder of their life. We cannot therefore look at income inequality and claim that it causes health inequality. Simply because there are two processes at work.

Further, we cannot look at lifespans in an area and insist that these reflect the life chances of those born in that area. Take, as an example, a retirement town like Bournemouth (say, any other will do). People often retire there at, say, 65. Can we then look at average lifespan in Bournemouth and correlate it to that of someone born in Bournemouth? No, of course we can't: for the average lifespan in Bournemouth is going to be boosted by including large numbers of people who only impact the numbers after they've survived to age 65. And, obviously, those rich enough to be able to move for their retirement.

This migration over lifetimes will lead to selection: the richer will go to richer areas (if nothing else on the grounds that they can afford the property prices) and the poorer will go to poorer areas. At least part of what is being measured is therefore the effects of this selection, not the life chances of those born in these areas. As such we simply cannot accept the conclusions they are making from this data.

And as up at the top, Wilkinson at least is supposed to be a demographer and he really is supposed to know all of this. It astonishes that they keep pushing this obviously incorrect line.

These statistics are compiled on the basis of LSOAs, lower super output areas. There's some 32,844 of these in England. That is, each LSOA is a unit of roughly 1,500 people give or take a bit. So, how many people die in the same 1,500 people strong grouping that they are born in? The geographical area inhabited by that same 1,500 people? Moving three streets over on marriage would take you out of such a small area.

Quite, somewhere between not very many and none these days. These statistics are simply valueless in trying to prove what Wilkinson and Pickett want to torture them into showing.