Using Deprivation Indices in Library Campaigning a Guide from Friends of Gloucestershire
Using deprivation indices in library campaigning – a guide from Friends of Gloucestershire Libraries.
The indices of deprivation (IDs) can be a useful tool in library campaigning. If you fear that poorer areas will be hit hardest by your authority’s plans, or you suspect that your authority has not paid proper regard to poverty and deprivation in making their plans for the library service, the IDs can be an excellent way of illustrating or even proving this point.
At first glance the IDs look quite complicated, but please be assured that they are not. The following process is time consuming and perhaps a little bit hard to explain on paper, but really easy, and at the end can leave you with a document that will be very useful in illustrating your points about deprivation and your authority’s library strategy. NOTE: Unfortunately there is no standardised means by which a local authority presents its ID data. The examples used in this guide are from Gloucestershire, so my apologies if the steps do not make sense when applied to your own authority’s presentations of the data – however, the principles should be the same.
Before we start a little bit of useful background information:
What are the indices of deprivation?
The Indices of Deprivation (IDs) are measures of deprivation collected nationally and analysed and published every 3-4 years. Data is collected around 37 indicators, designed to highlight characteristics of deprivation such as crime, unemployment and poor access to healthcare and education. The measurements are applied to geographical areas called Lower Super Output Areas (LSOAs), which are smaller than electoral wards and usually have a population of between 1,000 – 3,000. There are just over 34,000 LSOAs in England. The most recent indices date from 2010 (published March 2011), but importantly are based on data collected in 2008, so do not show the effect of the recent economic downturn on deprivation levels.
Due to the relatively small unit of analysis (LSOAs) in comparison to other national statistics, the IDs provide a usefully localised picture of deprivation, and can highlight pockets of deprivation within more affluent areas. The IDs are intended for use by government, to inform the targeting of resources and policy to improve life in disadvantaged communities.
How is the data organised?
In most presentations and analyses of ID data, the 37 indicators are collated into 7 key ‘domain indices’. These are:
1. income
2. employment
3. health and disability
4. education, skills and training
5. barriers to housing and services
6. living environment
7. crime
These 7 domain indices are then combined to produce the Indices of Multiple Deprivation (IMDs) – a ‘headline measure’ commonly referred to in government reports, and used to provide a general indication of the deprivation level in an area. Two supplementary indices, deprivation affecting children and deprivation affecting older people are also sometimes included within this overall measure.
Each LSOA receives a ‘score’ for each of the 7 domain indices listed above (or 9 if the 2 supplementary indices are included), as well as an ‘overall score’ using the IMDs. These scores are then ranked, with 1 = most deprived.
Accessing the IDs.
It is much better to access ID data from local authorities rather than from central government as it will be ready-sorted into the LSOAs that are of relevance to your campaign. All local authorities should make this data publicly viewable, and you should be able to find it by searching online for ‘indices of deprivation’ plus your authority’s name.
When you find the right webpage (example) you should have the option to download a spreadsheet of data which will look something like Figure 1.
The first four columns refer to the LSOA (which are identified first by a number code, then a name), next (third column) comes the ward within which the LSOA is situated, and (fourth column) the local authority area.
The columns where colour has been added (this might not have been done in all cases – this is the Gloucestershire data and other authorities may present their data differently) are the indices themselves. Here, the first of these columns refers to the IMDs, the next seven are the individual domain indices listed above, and the remainder are supplementary indices (including the 2 mentioned above – deprivation affecting children and older people).
Figure 1 – Screen-grab of partial indices of deprivation data for Gloucestershire
Due to the small geographical size of LSOAs each authority contains hundred, and the spreadsheet will consequently have many rows. For instance, there are 367 LSOAs in Gloucestershire and as many rows in this spreadsheet.
You will also notice that in the Gloucestershire example there are several sheets incorporated into this spreadsheet. We will look more at these during the analysis, but briefly, they show the county and national pictures of deprivation rankings, both in terms of actual rankings and quintiles (groupings of 20%).
Analysing the data
At this point it is useful to have beside you a list of the library branches you are concerned about and, if you’re not sure already, a map showing their location or a list of addresses.
Indices of multiple deprivation – the county picture.
First we are going to look at rankings within the authority area (in this example – the county of Gloucestershire), so we are working from the first spreadsheet sheet shown in Figure 1 – ‘COUNTY Ranks’. We are going to look for the top 10% and the top 10-20% of deprivation within the county based on the IMDs (the headline measure combining the 7 key domain indices).
The statisticians at Gloucestershire County Council have helpfully highlighted the relevant cells in red (top 10%) and yellow (top 10-20%) so we can either scan down the list and pick out the right colours or, more easily, add a ‘filter’ to the column heading ‘IMD county rank’ and select ‘sort ascending’ (remember 1 = most deprived). In the Gloucestershire example, this obviously groups the red and then the yellow cells together at the tops of the list, as those LSOAs appearing in the top 10% and top 10-20% of deprivation will have the lowest IMD rankings and so worst deprivation.
If your authority has not colour coded your data in this way, the same outcome can be achieved by looking at how many LSOAS there are in the county (look how many populated rows there are, then subtract the number used at the top of the spreadsheet for header rows). Then, follow the same ‘sort ascending procedure’ and count down and highlight 10% and then 20% of these rows. (E.g. in an authority with 300 LSOAs you would include the first 30 (top 10%) then the next 30 (top 10-20%).
I found the easiest way to relate this to the libraries was to copy and paste the columns showing LSOA number, name, and IMD ranking to a new sheet while this filter was on. I did this for the LSOAs within the top 10%, and then for those within the top 10-20%, Giving me this (Figure 2):
Figure 2: Screen-grab of top 10% and top 10-20% most deprived LSOAs within Gloucestershire based on IMDs (partial)
You can then go through your lists, highlighting any where you know the library serving that LSOA faces cuts or closure (Figure 3):
Figure 3: Screen-grab of top 10% and top 10-20% most deprived LSOAs within Gloucestershire based on IMDs with LSOAs affected by library cuts highlighted.
Note: There may be a smarter/quicker way of doing this without all the copying and pasting – if anyone finds it please let me know!
Although most LSOAs are helpfully named so as to make it obvious where they are (name may correspond to ward name or recognised neighbourhood) others do not. The electoral ward may be helpful here, otherwise you can input the LSOA number (first column) here, and the LSOA will be shown on a map (select ‘Ordance Survey overlay’ to see village and town names).
Domain indices – the county picture
This process can then be repeated for each of the individual domain indices. If you like you can do all 7, but for my Gloucestershire report, as time was limited, I took 4 of the 7 key domain indices that appeared particularly pertinent to library use:
1. Income
2. Employment
3. Education, training and skills
4. Barriers to housing and services (includes transport links as indicator)
As well as the two supplementary indicators - deprivation affecting children and deprivation affecting older people.
Note: The indice ‘Barriers to housing and services’ was included as Gloucestershire is a large rural county, and poor public transport links have been identified by library users as a major barrier to many accessing the ‘new’ library service planned by GCC.
For each indice, the same process as described above was repeated, but this time working from the column headed with the relevant indices name. e.g. ‘Income’ or ‘Employment’. Although this is quite time consuming, it is worth doing as even if an area does not appear in the top 10 or 20% of depirvation using the IMDs (which remember are a collation of all 7 key domain indices), it may appear in the top deprivation rankings for one or more of the individuals domain indices.
A good example from the Gloucestershire data is the rankings for the domain indice ‘Barriers to housing and services’. None of the LSOAs that appear in the top 10 and 20% of deprivation for this indice feature in the top 10 or 20% for the other indices or the IMDs. In contrast to the other rankings, which are dominated by LSOAs within the county’s major urban areas of Gloucester and Cheltenham, the LSOAs which rank most derived under this indice are dominated by rural areas of the Forest of Dean, Cotswolds and Stroud and Tewkesbury district. This allows us to make a useful argument about the impact that the closure of small rural branch libraries and the withdrawal of the mobile service will have on these kinds of areas.
The national picture
Some authorities also provide rankings based on the national indices (but not all). If your authority does, they will probably be included in the spreadsheet as a separate ‘sheet’ (as in the Gloucestershire example). You can switch to this sheet and repeat the processes outlined above to create a picture of where the LSOAs in your area affected by library cuts sit within the national picture of deprivation.
The child well-being index
Another dataset which may be useful – especially if you wish to illustrate the impact of library cuts and closures on children and young people – is the Child Well-Being Index. (CWI)
Like the IDs, this is a small area index (using LSOAs), and is designed to measure deprivation exclusively as it impacts upon children. The CWI is a relatively new dataset (first published in 2009 using 2007 data) so may not be available from all local authorities yet. Like the IDs it is a ranking system applied to LSOAs, and follows similar methodology, comprising seven domain indices which can be analysed either independently, or in conjunction to arrive at an overall ranking. They are:
1. Material Well-being
2. Health and Disability
3. Education
4. Crime
5. Housing
6. Environment
7. Children in Need
If the CWI data is available from your local authority, you will again probably be able to download a table which looks something like Figure 4:
Figure 4: Screen-grab of child well-being index data for Gloucestershire
This is quite similar to the IDs spreadsheet shown earlier, except the individual domain indices are shown on different sheets instead of within the same one. It is important to notice that in the CWI rankings, 1 = best (or least deprived) - the opposite to the IDs. I don’t know if this is typical or a peculiarity of the Gloucestershire data presentation, but always check the key or explanatory notes of any spreadsheet to avoid mishaps such as claiming the 10% least deprived areas are the most deprived!
As I was short of time when conducting my ID analysis for Gloucestershire, I only looked at the overall CWI ranking, but you could also look at the individual indices as with the IDs. Again the Gloucestershire statisticians had made things easy by including the column ‘percent national’ which allowed me to identify the LSOAs within the top 10% and 10-20% very easily.
To figure out the county rankings I simply sorted the data using the column ‘Rank of CWI’ - this time choosing ‘sort descending’ (with this presentation of the CWI, 1 = best/least deprived remember). I then copied and pasted the first 37 (10% of the LSOAs in Gloucestershire) into a new spreadsheet, followed by the next 37 to give me a list of the LSOAs ranked worst/most deprived in Gloucestershire using the CWI. This is exactly the same process as used for the IDs and detailed above, and was followed by the exact same process of highlighting those LSOAs where I knew the local library was affected by planned cuts or closure.
Writing up your data.
You should now have lots of excel spreadsheets (or sheets on the same spreadsheet depending on how you like to work). Each sheet should be titled with the name of the indice it refers to (IDM, name of the individual domain indice, CWI etc…) and whether the ranking you have copied and pasted refers to the national or county/local authority-wide picture.