There is some degree of confusion over what people mean by "postcode" when applied to modelling mortality in the United Kingdom. There are varying ways of using postcodes, depending how much of the full postcode is actually used. To illustrate, we will use the postcode of our office here in Edinburgh, which is EH11 2AS.

The EH part of our postcode is called the postcode region, of which there are 124 in the UK. This if often what people think of when they talk about postcodes and mortality, as the UK has large differences in mortality and longevity around the regions. However, with an estimated population of 60.6 million in mid-2006 according to the Office for National Statistics, there are hundreds of thousands of people in each postcode region. As you might imagine, grouping people arbitrarily by region is a fairly crude way of assessing their mortality characteristics. In fact, it can be downright misleading: the mortality characteristics of a lawyer in Aberdeen have more in common with another lawyer in Yorkshire, than with manual labourers in Aberdeenshire, for example.

Some researchers have used the postcode district (EH11) or the postcode sector (EH11 2), but the best way of using a postcode is to use it in its entirety, i.e. EH11 2AS. There are around 1.7 million postcodes in the UK, of which around 1.5 million are residential addresses. Straight away we can see the power of using the full postcode, since it covers on average 40 people. A postcode will usually cover a single street, and to walk down a street in the UK will give you a very good idea of the sort of person who lives there. Thus, the full postcode can give a much more detailed insight into someone's mortality characteristics than a fragment of the postcode can. A group of forty people in a street will be more homogeneous than a group of several hundred thousand.

Of course, it is unwieldy to work with postcodes directly in mortality analysis as there are 1.7 million of them. Instead we use a so-called geodemographic profiler to simplify postcodes into a geodemographic type. The purpose of a geodemographic type is simply to group similar people of similar lifestyles, regardless of what their actual geographic location is. For example, the Mosaic classification from Experian assigns the same type code A02 to postcodes as diverse as AB10 6HX (in Aberdeenshire) and YO10 4BG (in Yorkshire). In all there are 61 Mosaic type codes, which is obviously a far more manageable number than 1.7 million postcodes. The Mosaic type codes are also fewer in number than the 124 postcode regions, but their best feature is that they are homogeneous: in our earlier example, the similar well-to-do lawyers now find themselves in the same Mosaic type code, AO2, rather than in the separate AB and YO regions. As a result, mortality analyses using geodemographic profilers are far more powerful than simple geographic regions.




Find by key-word


This blog brings together two pieces of work.  The first ... Read more
In a previous post we considered the advantages of repurposing ... Read more
A fundamental assumption underlying most modern presentations of mortality modelling ... Read more
Stephen Richards
Stephen Richards is the Managing Director of Longevitas
Geodemographics in Longevitas

Longevitas users can control the geodemographic profiler used in the Deduplication tab in the Configuration area. The Upload Processing section contains a drop-down option list for available profilers. Options for UK data include Mosaic, Acorn, P2, Health Acorn, FSS, CAMEO and Personicx.

A variety of other options exists for territories outside the UK, such as the USA, Canada and the Netherlands. Note that each profiler requires a separate licence from the owner: Experian for Mosaic and FSS, CACI for Acorn and Health Acorn, Beacon Dodsworth for P2, Eurodirect for CAMEO and Acxiom for Personicx.