CampaignLab / data-pipeline

Scripts and schemas that aim to make data from the inventory easier to analyse
8 stars 8 forks source link

Ward Demographics #29

Open edpenington opened 5 years ago

edpenington commented 5 years ago

Hi all,

As a bit of a follow-up from the day at the start of December, I've been doing some work on ward populations etc. I've attached a couple of files, one showing how I've processed this stuff and one showing the final csv form, which will hopefully be useful.

In terms of analysis, there's actually quite a lot you can do with the data the ONS has been keeping. For one, you can work out how many people there are in an electorate, and use some of the broad-brush age groupings that pollsters use to see how many people you expect to vote for who (e.g. 18-24yr olds, 65+). What's more promising is that you can pull apart the demographics of different areas in a very detailed way by looking at the shape and distribution of ages and genders in the different wards. For example, you can use age profiles to identify: wards with large student populations; wards that contain a boarding school; wards where a lot of 18 year olds go off to university. Using more involved techniques, you can also proxy for wealth, ethnicity, education etc. etc.. Many of these things are publicly available at the ward level directly, but haven't been updated since the 2011 census. Since age statistics are updated more regularly, we can get more accurate and up-to-date pictures of the demographics of different electoral areas, and where might have changed in a big way.

The analysis part of this is currently quite rushed, but I will hopefully have the chance to show how useful it can be in predicting voting and vote shares before long. In the meantime, there are some bits in the attached where I compare age profiles to "Output Area Classifications" from the 2011, which should give a picture of how effectively it is picking up variation.

edpenington commented 5 years ago

ok never mind the file is way too big! i will find a way around that :)

jm2dev commented 5 years ago

Hi,

do you have a link to those data sources?

Thanks.

edpenington commented 5 years ago

Hi Jose,

Apologies I am travelling and unavailable right now but the source is the ONS Small Area Population Estimates (SAPE). There might be more detail on my branch of the data pipeline

Best

Ed

On Sat, 23 Feb 2019, 15:57 Jose Miguel, notifications@github.com wrote:

Hi,

do you have a link to those data sources?

Thanks.

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/CampaignLab/data-pipeline/issues/29#issuecomment-466664171, or mute the thread https://github.com/notifications/unsubscribe-auth/AMnMD7PjEp1gVGOaQJIqc-tP85KHDsk-ks5vQWTYgaJpZM4ZUyCQ .