The Brexit vote, declining health and immigration

The Brexit vote, declining health and immigration

People voted Leave most often in those parts of England which had the worse health trends, which saw the greatest rises in mortality rates in the two years after the vote, and to which the least immigrants had come in the year before the vote.


In local authority areas where immigration to England from outside the UK was highest in the year to June 2016, the fewest British people voted Leave in June 2016 and the subsequent rise in mortality rate from Q1 2016 to Q1 2016 was lowest. It was in areas where there were already the most social and economic problems, and the fewest immigrants, that most people voted for “anything but this”. Most of these Leave areas were in the South of England. Most people who voted Leave were social class A, B or C1 – but they still lived in areas of the south that, like most of the UK, were not fairing well by 2016.


The local authority districts of England and boroughs of London can be ordered from the lows of South Staffordshire (0.11%) and Staffordshire Moorlands (0.13%) up to highs of Westminster (5.9%) and the City of London (8.5%). They are ordered here by the share of the population living in each area in June 2016 who had arrived from abroad within the twelve months to June 2016.

Almost no-one came from aboard to live in those two districts of Staffordshire between June 2015 and June 2016. In contrast, one in every seventeen residents in Westminster at the time of the EU referendum was a very recent arrival, as were more than one in twelve of the residents of the City of London, many of them living in the Barbican towers.

All of England can be divided into five groups of districts by the international in-migration rate of June 2016, from the districts with the least migrants (0.3% on average) to those with the most (still just 2.7% on average). Each group is defined to have almost exactly the same population in June 2016 as ever other of the five groups.


Immigration rate in 2016, five types of English district


You might have thought that this rapid in-flux of new people would have annoyed those who had a right to vote in the EU referendum, which given how the referendum was organised, did not included the international migrants themselves.

In fact, the very opposite occurred.

In Westminster and the City of London less than a third of those who voted in the referendum voted to Leave the EU. In contrast, in both of those two Staffordshire districts with the lowest international immigration rate in all of England, some 65% of all those who voted to leave, almost two thirds!

Nationally, across the five groups of English districts, the fewer migrants there were, the more people voted to leave of all those who voted. In fact, it was only where there were the most migrants that a majority vote to Remain:


2016 Leave vote by local authority type (authorities grouped by immigration rate 2015-2016), England


This is, however, far from the end of the story of a very interesting anomaly. On the very same day as the EU Referendum was announced, the official mid-year population estimates of the UK were published for the year ending June 2015. Those estimates revealed the most remarkable rise in mortality in that year. It was little commented upon at the time, as the result of the EU referendum was such a huge shock, but among the rise in deaths was one anomaly. There was no rise in deaths rates for people aged 25-29, the age group of highest net in-migration to the UK at that time.

The ‘health migrant effect’ has been known about for a long time, but we didn’t realise just how healthy these migrants were and how unhealthy those of us who were not migrants were (on average) back then. This was despite increasing numbers of people saying when surveyed that they felt more ill, because those reports were ignored. Then more people in the UK began to die earlier than before.

Between the first quarter of 2016 and the first quarter of 2018 the numbers dying across England rose by an average of 15.8%, but by more in areas with fewer in-migrants:


The rise in mortality from Q1 2016 to Q1 2018 by local authority type (grouped by Immigration rate 2015-2016), England


Life expectancy peaked across the UK in 2014 and has been lower ever since, falling most years. The Office for National Statistics revealed an analysis that demonstrated this on August 7th 2018. One of the key graphs from the ONS analysis is shown next and it reveals yet another anomaly that has yet to receive much attention. It is perhaps the pattern for women that is most revealing:


Rise in life expectancy per year in weeks, 2010-2016 or nearest period and six years earlier

Average annual increase in period life expectancy at birth, selected countries, females 2004-2016


Women: average annual increase in period life expectancy at birth, selected countries. Office for National Statistics analysis of Human Mortality Database
See here.

All the reports on the international comparison of trends in life expectancy concentrated on the fact that women in the UK had seen the lowest rate of life expected improvement from 2010 to 2016 of the twenty affluent states being compared – which itself actually disguises the overall fall since 2014 as there were still some improvements in the early 2010 – 2012 period.

What no one commented on was the huge rise in women’s life expectancy in the UK in the six years prior to 2010: the 2004-2010 period. And, given what we now know about the relationship between migration in the UK and mortality this is, perhaps, important.

In May 2004 the EU welcomed in ten new members states, almost all central and Eastern European. Only the UK, Sweden and Ireland at first allowed migration from these new member states and the vast majority of those who came, came to the UK.

In hindsight it would appear that the women who came may well have been some of the most energetic and healthy, young, and well-educated. This will not be the whole story; but the picture changed abruptly once migration to a wider range of Western European countries was permitted.

So what about the men? Well, again a very similar picture is seen. Rapidly improving life expectancy among men in the UK prior to 2010 (only Portugal sees a greater improvement from a lower base), and then sudden stalling after 2010, with only the USA seeing even less improvement.

Of course, the UK experienced huge cuts to vital public services after 2010, especially to those services necessary for sustaining life amongst people who were most frail and most in need, especially the elderly with Alzheimer’s and dementia. The cuts to local authority budgets and the consequent repeated decimation, year on year of adult social service visits, meals on wheels services coupled with the stalling in the needed rise in health care funding and so much else very neatly matches up to the rising death rates among, at first the most elderly women living on their own and then not quite so elderly women and the most elderly men, and then more and more of the population – almost all originally UK born. It was not the in-migrants who were dying in rising numbers. The migrants were young and healthy and disproportionately worked as health carers.


Rise in life expectancy per year in weeks, 2010-2016 or nearest period and six years earlier

Average annual increase in period life expectancy at birth, selected countries, males, 2004-2016


Infant mortality rates also rose in the UK as midwifery services were constrained. The greatest rises were seen around the time of birth when access to a safe delivery services is key. The rises were statically significant for the most recent two years and not reported from elsewhere in Europe. In some of the best organised countries of Europe, infants are half as likely to die in their first year of life as those born in the UK. Figures released by ONS in late October 2018 confirmed the rise in infant mortality is still continuing in England. This is not happening elsewhere in Europe.

Of course, premature babies are born all across Europe. More that are sick survive in most other European countries and stillbirth rates are higher in the UK than elsewhere in most of Europe. But could there be a relationship between changes in migration patterns and the recent rise in infant mortality in the UK? Here is one potential clue, up to a third of midwives working in some hospitals are EU but not UK citizens. Once all of Europe opened up to migration from the East, attracting midwives from Eastern Europe to the UK became harder.

It will be many years before we understand the full story of went wrong in the UK to result in life expectancy falling significantly for the first time from 2014 onwards and infant mortality independently also rising. Those figures from ONS released in late October 2018 confirmed that age-sex standardised mortality rates for both men and women across all of the UK remained, by 2017, below the levels achieved in 2014. Mortality rates have been getting worse since 2014 for everyone in the Uk. We also know that 2018 has been even worse (so far). I will end with an anecdote.

On Friday September 21st 2018 it was revealed by the BBC that all but one of the many maternity units in Oxfordshire had shut their doors the night before due to lack of midwifery support. That same day the local Oxford Mail newspaper revealed that expectant mothers scheduled to give birth at the Cotswold Birth Centre in Chipping Norton should check Facebook posts to see if that centre, the one near them, was open when they went into labour.


For a PDF of this article and the original October 2018 publication it appeared in click here.