Deaths statistics in England and Wales

I found some interesting data on the ONS website recently. It related to the number of registered deaths by year by age of the person in England and Wales. A bit morbid maybe but interesting none the less.

Above is an animated gif of the number of deaths by year of age as a proportion of the total number of deaths in that year. The gif was created in GIMP, following their tutorial, although I did find it a lot easier to just paste the charts as new layers. I've decided that I don't really like animated gifs. You can't stop them, and when you want to study a particular point in detail it's gone before you can.

Anyway, there are 3 interesting points to this:

1. Average age - The movement of the peak to the right.
2. Infant mortality - The drop in the percentage of the number of deaths of people aged zero.
3. The war - The dip in the percentage of the number of deaths of people aged 80 around the year 2000.

Average Age

You can clearly see that the peak of the chart moves to the right significantly from 1963 to 2010. In 1963 the peak was around age 78, in 2010 it had moved to 86. Below is a chart showing the increase in the mean over time (rather than the mode given before).

1963 looks like an anomaly but a steady and declining increase can be seen from then. Modelling the line shows a peak at around 80. Splitting this line by gender (below) gives the often cited difference in average life expectancy for men and women. 

 The gap shows no sign of closing:

A different scale shows this more clearly:

My final chart in this section just shows the total number of deaths registered by year and the proportion that this is of the total population. As you can see this is steadily decreasing over time. Whether this is due to better health care or increasing migration of younger people (and hence less likely to die) is unknown.

Infant Mortality

It's impossible to tell from just this data whether infant mortality is decreasing. I'd need birth statistics to be sure. All things being equal though, the decrease in the proportion of the population under one dying:

is due to the decrease in infant mortality rather than a decrease in the birth rate.

The War

World war 2 that is. I assumed that that was the reason for the dip in the number of deaths registered for 80 year olds when looking at the initial chart. After all, if people had died during the war there would be less to die later in life.

The following chart suggests a different reason though.

The dip is the same for men and women and occurs during the ages 81 and 82. 2000-81/82 = 1918 and 1919. I initially assumed this meant a significant proportion of people born during those years would have died during WW2.

Given the dip is relatively similar for men and women and is very sharp (i.e. only covers a couple of years) it must be that the birth rate immediately after WW1 was unusually low. One can think of reasons why this might be so but without the data it's hard to tell or be sure.

There is a similar dip at 54 which corresponds to 1946. This is more evidence suggesting this reason is true - that after both wars the birth rate reduced significantly.

As always, comments are welcome.