Dennis

It is both.

Focus on Health January 2006 summary
Describes the health of people living in the UK across five key dimensions: health status, risk factors, ill-health, preventive, curative and long-term care services and mortality. Emphasis is placed on trends over time.

Over the last 25 years, improvements in survival have resulted in more people living longer and an increasing proportion of deaths occurring in older ages. However, there remain substantial social and geographical variations in health status, with people who are disadvantaged in terms of their educational, employment and socio-economic background having higher rates of reported poor health and limitations in daily activities. On average, the population in England had better reported health than the other countries in the UK.
http://www.statistics.go...h2005/health_summary.pdf

I was sufficiently intrigued to spend a bit of time digging around in the data. Let me preface my response by saying that I'm not a statistician, and a fuller investigation of the issue would probably take more time and subtler methods.

Having said that the trends are fairly clear. Using <a href="http://www.statistics.gov.uk/downloads/theme_population/deathsby-single-year-of-age-1961-2009.xls">data from the Office of National Statistics</a> (<a href="http://www.statistics.gov.uk/statbase/Product.asp?vlnk=14409">from this page</a>) which covers death rates in England and Wales, I've calculated the average age of death in each year for a) everyone b) over 5s (i.e. excluding child mortality) and c) everyone over 60:

<img src="http://www.citywire.co.uk/BinaryLoader.aspx?ContentID=451266&DeliveryTargetCode=System__Resources__Image" />

As you say, the decrease in child mortality has had a big impact. What I didn't realise was quite how big an impact - these next two charts show child deaths since 1961, and the dramatic fall in stillbirths as a percentage of total births:

<img src="http://www.citywire.co.uk/BinaryLoader.aspx?ContentID=451270&DeliveryTargetCode=System__Resources__Image" />

<img src="http://www.citywire.co.uk/BinaryLoader.aspx?ContentID=451268&DeliveryTargetCode=System__Resources__Image" />

Which is unambiguously wonderful news. We can also see that the most common age of death (child mortality excluded - for several years in the 60s, the most common age of death for men was zero) has risen steadily. The scale doesn't do it justice but to put it in context, the most common age of death has risen from 78 in 1998 to 88 in 2008!

<img src="http://www.citywire.co.uk/BinaryLoader.aspx?ContentID=451272&DeliveryTargetCode=System__Resources__Image" />

Another way of looking at it is to see how the distribution of age deaths has changed over the years. It's quite hard to show using a static image but here are three snapshots - 1961, 1985 and 2008:

<img src="http://www.citywire.co.uk/BinaryLoader.aspx?ContentID=451274&DeliveryTargetCode=System__Resources__Image" />
<img src="http://www.citywire.co.uk/BinaryLoader.aspx?ContentID=451276&DeliveryTargetCode=System__Resources__Image" />
<img src="http://www.citywire.co.uk/BinaryLoader.aspx?ContentID=451277&DeliveryTargetCode=System__Resources__Image" />

The shape of these curves seems a little surprising but I'm getting out of my depth here and won't try to explain them!

Of course, this kind of data can hide an awful lot - regional variations, socioeconomic variations and so on. So it needn't be in conflict with anecdotal evidence. But the overall picture, I'd say, is one of good news. Unless you're a Malthusian!

There is no doubt that mortality has declined, and that the cohort (to use the technical expression) born between 1925 and 1945 is showing a particularly strong increase in longevity. It is of course too early to draw useful conclusions as to the longevity in later life of cohorts born after 1945.

I have a tendency to be suspicious of the current received wisdom. Now that awareness of substantial longevity improvements has become widespread, could it be that future trends will disappoint? Lifestyle factors, and conceivably environmental ones could be pointing that way, even though medical science is continuing to come up with potential gains.

Thanks Jeremy!

William - very true. Increased wealth could be subject to diminishing returns, especially if a side-effect of that increased wealth is an increased incidence of obesity, heart disease etc. But I hope received wisdom turns out to be correct on this one.

Shameless self-promotion alert: A while back I interviewed Aubrey de Grey, a scientist who believes the first human who will live to 1,000 has already been born - sooner or later medical technology will advance to the point that life expectancy for most people will increase faster than their age. Quite a fun concept to play around with, whether or not you believe in the science.

You can watch/read it here: http://citywire.co.uk/mo...s-of-immortality/a399808

Rich - very well presented data.

To me the last three graphs are the most telling in the context of annuities as clearly the annuity providers should only be interested in the age of death of people who actually have annuities, ie the 60+ population.

The fact that the data is in the "bell curve" shape (albiet skewed) suggests that it is based on good random sampling of data. I would, however, like to see the same graphs for the 60 to 100 age range with the average age at death for men and the average death for women superimposed on the same time-line. In theory it should show that women will outlive the male population by some 10 years.