On my previous post I talked about the Index of Multiple Deprivation (IMD) in a London only setting. A common trait with the maps I created and those you can find elsewhere is the use of LSOA boundaries that reflect the geographical reality of the lay of the land. When concentrating on London in particular this does have an impact on how you perceive deprivation. This is in part down to how Lower Layer Super Output Areas (LSOAs) are constructed. An LSOA by design has to contain at least 1,000 residents and 400 households, with a national average of around 1,500 residents. London of course does not just consist of the denser populated core, but also more rural areas on the outskirts. As the methodology dictates that each LSOA has to have at least 1,000 people in it, the geographical extent of LSOAs tends to be larger in these more rural areas. This means visually they will be more dominant than central areas and can give a “false” impression as to how much of London’s population live in either more or less deprived areas.
In an attempt to try and address this visualisation problem I have used the Cartogram Geoprocessing Tool (Version 2) in ArcGIS 10 to create new maps of the IMD in London in 2007 and 2010. I created normal IMD maps for both years using the standard LSOA boundary data. Then for each year using the cartogram tool, each LSOA was rescaled by their estimated population in 2007 and 2009 respectively using mid-year population estimates available from the Office for National Statistics. The results of this can be seen on the “scrubber” images below for 2007 and 2010.
Move your mouse over the picture, to swipe between the 2007 London IMD standard and cartogram maps.
Show dividing line?
Move your mouse over the picture, to swipe between the 2010 London IMD standard and cartogram maps.
Show dividing line?
It becomes apparent from the cartograms that a greater number of London’s population live in deprived areas when compared to the standard maps. Conversely areas on the outskirts of London that are less deprived, appear noticeably smaller on the cartograms, reflecting the smaller number of people who live there when compared to central areas. In this respect the cartograms give a much better visual indication of how much of the London population live in more deprived areas, at the cost of geographical accuracy. I find it interesting that while the shape of London is obviously morphed by the cartograms, it has not been morphed out of recognition. This is most likely down to the fact the building blocks of LSOAs are based upon population. If the same exercise was performed at Ward level for example, using Ward population estimates that London would look dramatically different.