Working with UK demographic data and Google Fusion Tables

There are some publicly available spatial, i.e. geographic, data on population available in the UK. One of the sources is the Office of National Statistics (ONS) which houses a lot of data and you should also look at Census Dissemination Unit although much of their data is restricted to academics (but look here for some public data).

One of the problems is that this data doesn’t use the normal postcodes or postcode sector, but Super Output Areas (SOA). As the link explains these were developed to provide areas that were more similar in population size and would not change over time. There are multiple levels to SOAs, but the most useful are Lower Layer SOAs (LSOA) and Middle Layer SOAs (MSOA) as data is available for these levels. However you need to do more work to view the data as standard tools like Google Maps don’t know about these shapes.

Click the image to go to the London LSOA Atlas

Example: London LSOA Atlas

Before I start explaining how to handle this data I suggest you visit the excellent London LSOA Atlas which provides lots of demographic data at the Lower Layer SOA level (There is also a London MSOA Atlas too, and that is quicker to load). This is a really excellent site which allows you to look at all sorts of demographic data around London. Want to know how many people aged 40 to 44 live in each area? – no problem, just click Mid-year Estimate 2010 on the left hand menu and select the 40-44 age band.You can also get access to the underlying data via the Get data link on the Atlas page (see top right menu). Well done London!

Unfortunately I need to bring you down to earth as if you want to go outside London I don’t know of an equivalent of London’s LSOA Atlas. There is data available and, according to the Census Dissemination Unit more will be released over time. But you have to do a lot more work to make this visible.

Accessing Lower and Middle Layer SOA data

There is demographic data at LSOA and MSOA level showing population separated into five age bands for male and female for 2010. This is available from Office of National Statistics here. These are fairly big Excel files, but in Zipped format.

The second thing you need if you want to view the data on a map is the shapes of the LSOA or MSOA. Now getting the shapes is slightly more difficult, but it is still publicly available. You need to contact the ONS Geography unit as explained on this link and they will send you, free of charge, a CD with the shapes on. They say it could take some time to reach you but we found it came quite quickly. UPDATE: These files are now available online here.

If you have a Geographic Information System like ARCgis or MapInfo then you can view it straight away. However if you want to show it in Google etc. then you need to convert them to the right format.

Example of LSOA data: Females aged 65+

How to convert these shapes and get the data into Google Fusion Tables (FT) is very complicated and is outside the scope of this blog (here is a link to description of a method of uploading shapefiles to FT, but we have NOT tried it and therefore cannot endorse it).

However we have converted the data as we need it for the predictive analysis we do for our clients and we thought you would like to see an example of what is possible.

The picture to the right is an example of  displaying demographic data via Google Fusion Tables by LSOA, in this case how many females aged 65 or over are in each area (red is higher).

To access the actual data via Google Fusion Tables then click this link. You can then click any one of the coloured shapes and get the specific details for that LSOA.


There is good spatial demographic out there and with the UK government opening up data sources there is likely to be more available in the future. The main problem is getting it into a form that can be easily uploaded to free mapping tools such as Google, Bing Maps or OpenStreetMaps. However those tools are also coming. Its not easy now, but maybe in a few years time this blog post would look very different. Happy mapping!

1 Comment

  1. […] to work with a range of location data, e.g. address, demographic data, sector or […]

Leave A Reply