LSOAs, LEPs and lookups : A beginner’s guide to statistical geographies

 

If this blog post has caught your eye, then you probably have some interest in using place-based data – excellent! Us too! There is lots of jargon and overlap when it comes to understanding the underlying geographies that data is published at, as well as the different types of areas that are used day-to-day across different organisations. 

When we talk about place-based data here, we are referring to the sort of data that we include in our Local Insight platform. This ranges from the well known Census and Indices of Deprivation data to a wealth of housing, health, crime and education data, to more specific datasets such as the prevalence of loneliness or the amount of green space per local area. (You can take a look at selection of the data we hold on our public site).

This resource aims to untangle some of the jargon and introduces some of the intricacies of standard geographies. From LEPs to LSOAs, all will be revealed.

This guide is for you if you:

  • are fairly new to using small area data
  • want to feel better equipped to use data published at small area level
  • struggle to find data for the geographies you’re interested in
  • want a better understanding of how different geographic areas are constructed.

Statistical Geographies

The vast majority of place-based open data is published at at least one of the following geographies. Output Areas and Super Output Areas are standard areas that were primarily designed for the publication of the Census. They have been designed to be fairly homogenous in terms of population size, so that you can compare like-for-like when looking at changes over time & when comparing different areas and different datasets.

Output Areas (OAs): These are the smallest of the geographies that data is published at and have an average population of about 310 residents (the table below shows the upper and lower thresholds). Not very much data is published at this level, although Census outputs are.

Lower Layer Super Output Areas (LSOAs): LSOAs have an average population of 1500 people or 650 households. A lot more data is available directly at LSOA level, including the majority of the data included within our tool, Local Insight.

Middle Layer Super Output Areas (MSOAs): MSOAs have an average population of 7500 residents or 4000 households. There are some datasets out there that are published at MSOA level as the smallest geography – for example estimates on prevalence of different health conditions.

 

Area type Lower threshold Upper threshold
People Households People Households
Output Areas 100 40 625 250
Lower Layer Super Output Areas 1,000 400 3,000 1,200
Middle Layer Super Output Areas 5,000 2,000 15,000 6,000
Electoral wards/divisions 100 40 n/a n/a
Source: Office for National Statistics

The whole of England and Wales can be broken down into these constituent areas – or building blocks (Scotland and Northern Ireland are a different story). OAs nestle within the boundaries of LSOAs, LSOAs nestle within the boundaries of MSOAs and MSOAs nestle within the boundaries of Local Authorities.

NB: Each image is not to scale. These images show how OAs nestles within LSOAs and LSOAs within MSOAs.

These geographies are really useful in providing the structure for collecting, processing, storing and aggregating data, as well as being a great unit to show comparison. However, they do have one pretty big drawback and that is people do not tend to relate to them, at all. There are no names associated with LSOAs and they cut across neighbourhoods rather than aligning with real communities on the ground.

Wards

People generally tend to be more familiar with the term wards and will be more likely to identify which ward they are from, rather than which LSOA. Wards are a very useful unit for analysis precisely because of this.

However, datasets are often not published directly at ward level. Firstly, because ward boundaries change a lot and are therefore less likely to be consistent over time (not too mention the administrative headache it would be). Secondly, wards vary greatly in size (anything from 1000 to 30,000 people), and therefore it is difficult to compare different areas to each other.

So when using place-based data, you may need to source data for different types of areas depending on the questions you are trying to answer and the audiences you are engaging with (read to the end for resources that can help with this).

Other geographies

Below are some other types of geographies that you may come across on your travels. Again, the majority of place-based data isn’t published directly at these levels. However, it is possible (in most cases) to aggregate to these geographies. These geographies are less commonly used than wards, but useful to be aware of nevertheless. 

If there are any other types of areas you want included in this guide, please get in touch on support@ocsi.co.uk – we’d love to know what is helpful!

Parishes:

Parishes are the lowest tier of local government. Unlike Super Output Areas & Wards – not everywhere is covered by a parish. Approximately 35% of the English population live in a civil parish.

As with wards, parishes can vary greatly in size, with some of the smallest having fewer than 100 residents. This can make it quite tricky to get accurate data for these areas, as place-based open data doesn’t tend to be published at this level of granularity.

Local Enterprise Partnerships (LEPs):

LEPs are voluntary partnerships between local authorities and businesses, straddling multiple local authorities to help determine local economic priorities (such as approaches to economic growth and job creation).

There are currently 38 LEPs and these are not mutually exclusive. For example; Lewes District Council is a member of both the South East LEP and the Coast to Capital LEP.

For the most part, they are aligned with local authority boundaries, so it is fairly straightforward to aggregate data published at local authority level to these geographies. However, as with every good rule – there are exceptions. For example, in Hampshire there are a couple of districts that are split over different LEPs.

Parliamentary constituencies:

The whole of United Kingdom is broken down into 650 Parliamentary Constituencies. Each of these areas are represented by one Member of Parliament (MP). These boundaries largely use local government wards as the building blocks for proposed constituencies.

Again, these constituencies vary in size and population. New rules stipulate stricter limits on the size of the electorate in each constituency. The 2018 Boundary Review has recommended changes, which will see a reduction in total number of seats from 650 to 600. Each constituency must contain between 71,301 and 78,507 electors. For more detailed information on the proposed upcoming changes, take a look at the background information put together by Boundary Commission for England.

Resources

ONS Open Geography Portal

The ONS Open Geography Portal is a really handy resource for converting statistical geographies such as OAs and LSOAs into other more meaningful geographies such as wards.

If you want to know which underlying areas make up your wards, then the LSOA 2011 – Wards lookup or OA 2011 – Wards lookup is a good place to start. Using Output Area definitions may be more accurate

Step by step

  1. Navigate to the LSOA 2011 – Wards lookup (this can be found by clicking on lookups -> best fit lookups -> LSOA 2011 -> Wards)
  2. Select the version of the wards you want to download (if in doubt, choose the most recent)
  3. In the top right, click download – full spreadsheet
  4. Once you have opened the Excel spreadsheet, filter column G (this should be LAD18NM) for the local authority area you are interested in. It is useful to filter by local authority first as there are lots of wards across the country with the same name, e.g. there are more than 20 wards called ‘Castle’ across England
  5. Sort column D (WD18NM) in alphabetical order
  6. From here, it is easy to see which LSOA codes (column A – LSOA11CD) corresponds to each ward.

Highlighting the LSOAs that make up Brickhill Ward in Bedford. Please note that this is a best fit estimate taken from ONS lookup.

Local Insight

Once you know which underlying geographies make up your wards, you are one step closer to converting LSOA level data to ward level data. However, it can still be a fairly time-consuming & complex process to do so.

Our tool Local Insight does all of the hard work for you. All you need to do is set up your neighbourhoods and wards as custom areas (either by drawing on a map or selecting from a list of LSOAs) and then Local Insight instantly aggregates more than 900 local level datasets to these areas. We also update the data regularly, with any new or updated datasets added to the system within a working week of release.

This means you can get instant access to data for the areas you care about whether you are interested in the Indices of Deprivation, Universal Credit or local health data.

If this sounds like it could solve some of your data headaches, get in touch for a demo and free trial.