Coronavirus pandemic: daily updated research and data.
Read more

What’s happening to life expectancy in Britain?

Our World in Data presents the empirical evidence on global development in entries dedicated to specific topics.

This blog post draws on data and research discussed in our entry on Life Expectancy.

Most of us want to live a good and fulfilling life; for many this also means living a long one. So when headlines state that life expectancy in Britain is falling, it’s important to pay attention.

Here’s a list of some pretty unnerving headlines covering this topic.

What’s behind these alarming titles? Let’s dig deeper to understand what the data is and isn’t telling us.

Is life expectancy in Britain really falling?

The stories behind these headlines are primarily based on a simple fact: the 2016 statistical bulletin from the UK’s Office for National Statistics’ (ONS) included estimates of projected life expectancy that were lower than the earlier projections published in 2014. To be specific, the new projections were one year lower in the ONS’s 2016 revision.

Does that mean that life expectancy is dramatically falling over time as the headlines suggest? Not really.

In the chart below we show the ONS’s life expectancy projections for males and females under the 2014 and 2016 revisions. We see that there is an unambiguously positive trend in all lines. Life expectancy was, and still is projected to increase substantially through the 21st century, both for men and women. The difference is that projections for both males and females, by 2041, were approximately one year lower in the ONS’s 2016 revision.

So the story is not really that life expectancy is falling over time; or even that life expectancy has stagnated. The story is that life expectancy will likely keep growing, but the ONS reduced the magnitude of predicted growth.

The ONS revision is less alarming than the headlines suggest, but it certainly deserves attention.

Where do ‘life expectancy projections’ come from?

The term “life expectancy” refers to the number of years a person can expect to live and it is based on an estimate of the average age that members of a particular population group will be when they die.

Estimating average age-at-death for a group of people would be a simple exercise if we had complete records and we were only interested in a group of individuals who have already died. But this is of course rarely the case – records are often incomplete and we are typically interested in making inferences about how long a group of people can expect to live in the future. So we need to make assumptions.

There are several different approaches. A common one consists in estimating the average length of life for a hypothetical cohort assumed to be exposed, from birth through death, to the mortality rates observed at a given year. This approach, known as ‘period life expectancy’, does not take into account changes in mortality rates over time – it is completely based on a static snapshot of observed mortality patterns across people of different ages in a specific year.

But since age-specific mortality patterns are not usually constant over time, it is common to add an extra layer of assumptions, in order to estimate period life expectancy projections. These are produced by first making assumptions about how the mortality patterns for people of different ages will evolve over time, and then estimating life expectancy for hypothetical future cohorts.

The life expectancy projections in the chart above are period life expectancy projections. They rest on strong assumptions about how mortality rates across different age groups will improve in the future.

Why did the ONS revise its life expectancy projections?

The assumptions behind the 2014 and 2016 period life expectancy projections are mostly identical except for one crucial detail: The 2016 revision made changes to the assumed mortality rate improvements for one specific group of individuals, the so-called ‘golden cohort’. Who are the people in the ‘golden cohort’?

When we look at rates of mortality improvements over the 20th century we find an interesting pattern: the cohort born between 1923 and 1938 (the so-called ‘golden cohort’) have shown consistently higher rates of mortality improvement relative to other cohorts.

In the ONS 2014 projections of life expectancy, it was assumed that this higher rate of improvement for the ‘golden cohort’ would continue for many years ahead. However, over the last few years this effect within the cohort has almost disappeared. Improvements in mortality rates have fallen more closely in line with other population groups, so in the 2016 ONS revision they therefore removed the special treatment for this group.

This raises two fundamental questions: what is special about the ‘golden cohort’, and why has this ‘golden effect’ disappeared in recent years?

The golden cohort effect

The success of the golden cohort in terms of mortality improvements remains to some degree an enigma. Many have attempted to explain it, but the evidence remains inconclusive.

What’s difficult to explain is not that those born in the golden cohort have done significantly better than their older cohort – we might expect this to be the case thanks to improvements in healthcare, prosperity, and medical advances. What is difficult to explain is that those in the golden cohort show consistently higher improvements than those born after.

This puzzling pattern is not exclusive to Great Britain: It has also been documented in France, Austria, Germany, Italy, Japan, Netherlands and Switzerland.

Despite the fact that the evidence is inconclusive, we do have some hints about factors that might make the golden cohort special. These factors include a special childhood in terms of health and nutrition, as well as exceptional behaviour in terms of habits, such as smoking.

This last factor is particularly compelling, since cigarette smoking peaked and then declined remarkably in the second half of the 20th century. Since the decline in smoking translates into lower mortality rates, and these improvements are often greatest for people who are around at the peak of the epidemic, it is natural to expect larger improvements for the golden cohort.

The end of the golden era

This leaves us with the question of why the golden-cohort mortality improvements have changed in recent years. Again, there are many potential explanations and the evidence is inconclusive.

First, it could be that the relative advantage of the golden cohort was bound to end. Indeed, if a dominant driver of the golden cohort’s mortality improvements is the dramatic reduction in smoking, we would expect that this effect would eventually disappear as changes in lung cancer mortality begin to stall.

And second, it could be that there have been significant changes in health inputs, such as changes in flu or disease outbreaks, lifestyle factors such as diet, or healthcare and treatment.

What has changed in elements of health and care in recent years?

In 2015 we saw a notable increase in mortality rates and as a result, a temporary drop in life expectancy. The majority of this excess mortality occurred in the first three months of 2015—notably in January—from a particularly fatal influenza season predominated by the influenza A(H3N2) virus. It’s known that mortality rates, particularly in the elderly, are higher during seasons of influenza. Indeed, Britain was not alone in experiencing a particularly fatal winter. The trends were widespread across Europe. In 2015, female life expectancy at birth fell in 23 of the 28 EU countries, alongside drops in male life expectancy in 16. The effect was even more dramatic in older demographics: female and male life expectancy aged 65+ fell in 25 and 21 EU countries, respectively.

Additionally, as widely acknowledged, in the period 2014-2016 there were also changes in public spending, both medical and wider services. There are important links between cuts to public spending and negative impacts on health outcomes, so even if particularly fatal influenza periods are to blame for mortality hikes in recent years, inadequate healthcare support can only serve to exacerbate this issue.

We must also be wary of the geographical and distributional aspects that national life expectancy figures hide. Whilst the average life expectancy in Britain continues to rise, particular areas have been badly hit in terms of economic decline as well as health outcomes. Blackpool has the lowest life expectancy in England and there are signs that this is falling. The drivers of this health crisis are likely to be complex: Blackpool has one of the highest rates of obesity, smoking, liver disease and antidepressant subscription rates in the country; it has been hit by austerity but its economic and health decline likely run deeper than this. Understanding these complexities deserve our attention.

In conclusion: What’s going on with life expectancy in Britain?

According to the figures by the ONS, life expectancy in Britain will likely continue rising throughout the 21st century. However, it remains true that mortality rate improvements for a very specific cohort of people born between 1923 and 1938, have slowed down in recent years. With the available evidence, it’s not possible for anyone to say why health improvements for the golden cohort have slowed. This is something that deserves more attention and research.

The media coverage on this was weak – headlines were misleading and arguments were often not supported by the evidence. In particular, several articles over-attributed the role that austerity may have played in explaining the ONS’s technical revisions of life expectancy. It is correct to be worried about the impact of public funding cuts on health and wellbeing. If trends in recent years are the result of multiple compounding factors, austerity cuts can only serve to exacerbate this issue.

But it’s important to be accurate and fair when presenting the evidence: there is not enough empirical support, at least yet, for the claim that the last couple of years of austerity in Britain have translated into large and widespread reductions in life expectancy with knock-on effects for generations to follow.