Post is pinned.Post has attachment

Errors have been found in the 2011 OAC lookup files provided by ONS (including the postcode directories).

In total, 6,416 Output Areas (2.8% of all Output Areas) have been assigned to the wrong Subgroup code and name in the ONS lookup files (Note: These errors only impact ONS outputs and not those provided via the GitHub page or via

For more detail about these errors and for corrected ONS lookup files please visit:

Please note: The postcode directories provided by ONS (such as the ONSPD and NSPL) will contain the correct codes from their August 2017 releases onward.

Opportunity to feed into plans for the 2011 Area Classification for Super Output Areas...

ONS is progressing the production of a 2011 Area Classification for Super Output Areas, which we are looking to publish this Spring. This classification for the UK will cover Lower Layer Super Output Areas in England and Wales, Data Zones in Scotland, and Super Output Areas in Northern Ireland. This will complete the set of Area Classifications to be produced covering different geographies following the 2011 Census.

We are currently considering whether to produce and publish this classification for either (1) Supergroups and Groups only, or (2) Supergroups, Groups and Subgroups.

In coming to a decision over whether to produce a subgroup cluster or not, it would be very helpful for us to know from users with an interest in this classification, whether a subgroup cluster is likely to be of interest, or whether your interest is likely to be more at Supergroup and/or Group level.

If we are to produce the classification to include subgroup clusters (in addition to Supergroups and Groups), then we need to be satisfied that this will meet a user requirement.

It would be appreciated if you could please provide comments on any user needs for Supergroup, Group or Subgroup clusters for the 2011 Area Classification for Super Output Areas by email to

Thank you.

Post has attachment

Hello, commendable on adopting an open approach, I am just curious, is the methodology transferable to census datasets outside the UK. If so what things should one keep in mind.

Hello, my query concerns the TOAC. I have tried to get some information on how this classification was derived. An author of this paper: Singleton, A., Pavlis, M. and Longley, P.A., 2016. The stability of geodemographic cluster assignments over an intercensal period. Journal of Geographical Systems, 18(2), pp.97-123. has confirmed that the TOAC derivation is not covered there. Does anyone know of a paper or report that describes the construction of the TOAC? Thanks.

Hi, thanks for all your hard work. I've been trying to understand the project by reading the R code. I've downloaded the 4.4Gb file, but when I unzip it (and the files inside), the R code doesn't seem to correspond to the (unpacked) directory structure. Have I overlooked something obvious, like I'm supposed to download a separate file. Thanks so much, your work is inspiring!

Hi first of all these Classifications are very interesting and thanks for making them accessible.
My questions: is there any document explaining what each super group and sub_group mean (I mean brief explanations)?

Hello friends,

In all versions of the ONS Postcode Directory, I'm finding 2d2 / 2d3 subgroups switched in the oac11 field, as well as 3d1 (listed as 3D2) / 3d2 (3D3) / 3d3 (3D1).

For example, in National_Statistics_Postcode_Lookup_Latest_Centroids.csv from May 2017, EC2Y 8DR / E00000001 is classified as 2D2. Should be 2d3.

Do you agree? That'd be 100,000+ postcodes with the wrong OAC.

Hi... is there a way I can upload postcodes in bulk to find the OAC classification?
Thanks very much in advance.

Hello, I am interested in finding or deriving an open classification of Westminster Parliamentary Constituencies (WPC). So far I have been unable to find one. What I was thinking of doing was using a ONS lookup of COA to WPC to find out what types of COA make up each WPC. Sometimes this is easy, for example in the Barking WPC , 80% of COAs are Multicultural Metropolitans whilst the remaining 20% are Ethnicity Central. Sometimes this is not so easy, eg Arundal and South Downs is 37% Rural Residents; 25% Urbanites and 25% Sub-urbanites, plus some others. I guess I am asking, is there a way to infer a classification for a larger population geography based on a classification of its smaller geography constituent areas? I note that ONS publish both a COA and a local authority classification but I think these are largely independent. Thanks.

Wait while more posts are being loaded