Just to remind, we do not measure life expectancy by place of birth

Knowing more about the world around us is a good thing. Often enough we cannot know the thing we wish to know so we use a proxy. This is fine, as long as we understand that we are using a proxy and that such proxies have their limitations. Use the proxy to try to tell us something the proxy cannot tell us and we’ll get into all sorts of trouble.

For example, we do not measure lifespan by place of birth. We measure lifespan by place of death. This should be obvious enough for that future is unknowable and yet we do want to construct some sort of idea of how long this glorious life will be.

This is, sadly, important:

Girls born in the poorest areas of England will have almost 20 fewer years of good health compared with those in the wealthiest, according to figures that also reveal overall life expectancy in the most deprived areas has dropped significantly.

Female healthy life expectancy at birth in the most deprived areas was 19.3 years less than in the least deprived areas in 2018 to 2020, according to the Office for National Statistics (ONS). For males it was 18.6 years less.

All of that is abject nonsense.

“It also highlights that in the most deprived areas, people are living more of their life in ill health,” he said. “Girls born in the poorest areas of England live 19 fewer years in good health than those born in the wealthiest. A staggering difference in life chances.”

That is not true.

Earlier ONS figures are here.

Deprivation deciles are based on the Index of Multiple Deprivation

And:

The English Indices of Deprivation measure relative levels of deprivation in 32,844 small areas or neighbourhoods, called Lower-layer Super Output Areas, in England

So, to recap. We measure where someone dies and at what age. There is no cross-collation or checking of where someone was born in the system. Absolutely none, we have checked this with a coroner’s office, the information simply is not available (the long form certificate necessary to be able to check place of birth against place and age of death is simply not public information).

If we were to be measuring at the level of a country, or possibly even a county, then the proxy - age and place of death - would be useful if not perfect in determining life expectancy at birth. Not perfect because some do change country (it is what, 14% of the UK that is currently foreign born?) and more change county.

But when we start measuring at this level of detail - those Lower-layer Super Output Areas are around 1500 people each, much smaller than the XXX part of a postcode and a small handful of them at the XXX XX_ level - then we are using a very, very, imperfect proxy.

For the percentage of the population that dies in the same or adjacent postcode as their place of birth is around and about zero.

We simply do not measure, in any useful manner at all, life expectancy by place of birth. Therefore we cannot, except at the grossest - ie national, roughly and maybe - level make any useful conclusions about life expectancy by detailed place of birth. Which also does mean that we can make no policies which will change this as we’ve no damn clue what it is in the first place and couldn’t measure any change that we did manage to make either.

Failing to understand that an imperfect proxy is being used means that we’re getting into all sorts of trouble here. So, we should stop doing this, right?