jonnob/lsoa_foi_template_maker

Creates a form that can be used to anonymise address data at source allowing fine grained FOI requests which do not breach the data protection act

Sucessfully make sensitive geospatial FOI's using pre-anonymisation

Requesting geospatial data throught the UK FOI system can often result in rejection due to risk of breach of the data protection act.

In order to avoid such issues, the code in this repository creates an excel file that takes postcode data from the target departments database and returns the LSOA code.- This allows the target department to delete the postocde data leaving just the LSOA and preventing the disclosure of sensitive information, or the identification of individuals.

The goal of this repo is to encourage government transparency and access to government data whilst protecting the privacy of individuals.

This approach as been successfully applied to over 120 FOI's with zero rejections (Although there was a cartain amount of back and fourth in some cases!).

The approach uses the fact that all government departments use relational databases and can extract data in CSV or excel format.

Instructions

• Download the ONSPD zip file and extract to the data folder in the repo
• If using docker build the image
• Run the R script or docker image
  • If you are not using the dockerfile make sure you have installed the required packages
• The resulting files will be in the folder ./data/Regions. For ease of use the templates are organised by Region of England and Wales

Notes

If you want to save space navigate to the "Data" folder within the "ONSPD" folder and delete the "multi_csv" folder and the ".txt" file. only the ".csv" is used. This will save approx 2GB of space.

Help with Java jdk and jni.h

Installing Java dependencies for R can be a horrible experience, and beyond the scope of this readme. You will have to work it out yourself if you end up with the dreaded "jni.h: No such file or directory" message. This stackoverflow question may help. It may be easier to use the Docker image.

Data

The process only uses a single dataset, ONSPD a database of postcodes and a lot of useful related data. The ONSPD can be found at this link ONSPD. The link was the most recent at the time of creating the repo. An update for census 2021 areas may be appropriate soon though.

Making the FOI request

Making an FOI request like this often causes the receiving party concern that completing the request will breach the data protection act. It won't. Over the course of sending requests using this template I also made a request template, that reduced the amount of time between sending the request and getting a positive response. I often include a copy of the completed template, ideally a neighbour as this makes it less likely the recieving party will claim it breaches the data protection act. Most initial rejections are because the reciever is scared of doing something wrong, only a handful of teams are deliberately difficult. With that in mind the letter primarily tries to reassure the reciever. Sometimes the recieving party does not have the technical skill to do things like 'paste as values' or copy down formulas', in these cases they are often grateful for a quick phone call just to walk them through it.

The letter is for my original use case of council tax. However both the template and letter can be adapted as necessary.

See below

To Whom it may Concern

Please provide me with the list of all exemptions and discounts for domestic council tax in so and so local authority at LSOA level (not including postcodes) using the template provided.

Please read the instructions in the yellow box within the template for how to complete the template.

I have attached an example from some other local authority prefably one close by.

This information has been successfully provided by over 120 local authorities all with different IT systems, including all the local authorities in London,Birmingham, Manchester and Cumbria.

This request as been specifically designed so that it does not breach the data protection act. Please read the instructions carefully and check the attached completed example before responding that it does.

Please note you do not need to do any geographical matching the template does that for you.

If you have any queries please contact me for clarification.

Regards your name etc

Citing

Bourne, J. Empty homes: mapping the extent and value of low-use domestic property in England and Wales. Palgrave Commun 5, 9 (2019). https://doi.org/10.1057/s41599-019-0216-y