Delivering the annual canvass - England

As a minimum, each local data set should include:

• First name
• Surname
• Postal address

Data sets which also include the following fields may help to improve the quality of the match results:

• Middle name(s)
• UPRN data
• Date of birth 
• Previous names
• Age of data, for example when it was created or last updated

Local data sets do not have to include data for all properties on your register and could be more focussed on specific property types or groups of individuals.

For example, you may have data from social services or education departments which may only cover some of your electorate, but which may still be useful.

How do I ensure a data source is of high quality?

You should ensure that you only select data sources that are high quality. A high quality data set is one where the information within it is reliable, trusted, current and accurate. The Ministry of Housing, Communities and Local Government have a resource available on how to assess the quality of local data which can be found below:

You should also consider what data standards and good practices exist in relation to the local data sources you wish to use and whether the data holder follows them.

For example, the Department for Work and Pensions has set out detailed guidance on good practice for the processing and use of council tax benefit and housing benefit data, which includes guidance on the checking of evidence provided to local authorities and how to deal with fraud. An ERO for an area that delivers its benefits service to these standards should be able to be confident in using benefits data for local data matching.

When determining the quality of a local data source you should take into account when it was last updated and whether the source(s) of the data are reliant on information provided by individuals with no checks by the local authority on the accuracy of the information.

For example, applications for library membership may be based entirely on information provided by the service users and so you may conclude that, as a result, library data is not suitable for local data matching.

In addition to being able to trust that the local data source is of high quality. You also need to consider any cost implications associated with processing local data and, if there are any costs, whether they make the best use of the resources you have available to you.

For example:

• Will work need to be carried out manually by you and your staff to ensure the data is useable or will you need support for the data matching from another team, such as from IT, to manipulate the files? 
• Can the data file be managed by a partially automated process by loading it into your EMS system or similar?
• Will you need to purchase any new or additional software or licenses for any existing software?

You should test new local data sets before you use them. You should talk to your EMS provider about how your EMS system may help you to do this in practice. 

You should evaluate the effectiveness of your local data sets each year using the data accuracy scores provided by the Ministry of Housing, Communities and Local Government. These scores will provide an indication of how effective the different local data sources have been in matching electors. This will enable you to make more informed decisions about using local data sets in future canvasses.