Migration and that North/South death rate difference

We've another of those terrifying facts detailing the gross inequality of the UK:

The effect of this economic dereliction is far deeper than simply a clunky rail service, however. This week, researchers from the University of Manchester and the University of York warned that the rate of premature death in people under 45 was falling in the south, but stagnating in the north. In 2015, the number of premature deaths of people aged 35 to 44 was 50% higher in the north than the south. Since 2008, the regional death gap has widened alarmingly, bucking a decades-long trend. Life expectancy is already lower in the north; now, if you don’t live in the south, your chances of dying young have increased.

As with all the other studies and claims about regional inequality the bit that no one is taking account of is migration. Bournemouth, for example, regularly appears near the top of longevity lists. That it is, to some extent, a retirement town, where people move in their 60s, is the explanation. People who are already in their 60s quite naturally have a longer expected lifespan than those who have already died at this age.

Similarly with the Appallachia story in the US. The region is depopulating. Those who leave are the young and educated, exactly those we expect to have the longer lifespans. We do not claim this is all of it, only that it is some of it - even if overall rates of drug death, suicide, alcoholism, don't change at all, if that portion of the population least prone to them leaves then the recorded rate in an area will rise.

British internal migration is not so extreme perhaps but it does happen rather more than in most other European countries. And as the ONS tells us the internal migrants are more likely to be North to South, young, healthy and male. Each of those things, men more than women, the young more than the old, healthy more than unhealthy.

At which point we do again insist that this is some part of what is causing that divergence of death rates. And also we insist that we're not going to pay much attention to figures presented which don't at least attempt to quantify the extent to which the recorded numbers are simply a product of such migration patterns.