Requesting library data via Freedom of Information

Regular readers may know I’ve previously written about requesting data from multiple public libraries using Freedom of Information (FOI). These have been genuine requests - to understand and compare datasets. However, FOI isn’t an efficient way to obtain data. It is slow, and often requires follow-up requests to clarify or correct the data provided.

For libraries it’s also not ideal. There isn’t much time to respond to requests, and these have to be handled as they come in. It would be better for a library service to have regular data processes in place, where they know what data they will be publishing and when. FOI is also a major source of data breaches - the Information Commissioner’s Office (ICO) has just released more guidance on disclosing documents to the public, due to data breaches involving personal data in FOI responses.

Those breaches were not caused by too much open data or transparency - they were from not having clear data sharing processes, causing a lack of data literacy when responding to FOIs. Some of the most high profile data breaches are situations where the requested data should have already been openly available, the public body then made a mess when responding to a data request.

The ideal situation for library data would be for services to adopt an open by default policy for data, and publish in standardised formats under an open licence. The primary purpose would be to serve the public, but it could benefit any organisation requiring data from libraries, including sector bodies and other library services.

For example, the task of administering Public Lending Right payments to authors should include the British Library querying openly published data from library management systems. Instead, there is a separate data collection process arranged by the British Library, available to no-one else. It’s such wasted effort that could achieve so much more.

There are other data processes like CollectionHQ gathering data for their collection analysis, Nielsen for their LibScan service, CIPFA, the Reading Agency, etc. And for plenty of them, the library service don’t have oversight of the process or really know what they’re providing, let alone being able to make use of it themselves.

The same goes for library locations. If library services published branch and mobile locations in a standardised format, there would be no need for the LibraryOn project to maintain that data themselves and accept emails asking them to update it. It would also be easily fed into Google and other map services.

Open by default would significantly reduce data efforts but should also increase data literacy in libraries, with data becoming a clearly defined public task of public libraries.

Essentially the principle could be: open data publishing by libraries for any purpose NOT data collection by third parties for specific purposes.

Anyway, that ideal is a long way off! But we have Freedom of Information. For all its faults, it is a good direct link between the public and library services.

After a short pilot, data requests have gone out to all library services in the UK. So what data is being requested? Fairly basic usage data for the previous financial year (2024/25):

1. A count of loans, per library branch, per month
2. The top one hundred titles loaned, with a count of each, for the year
3. A count of reservations, per library branch, per month
4. A count of active members, per home branch, for the year
5. A count of visits, per library branch, and per month.

Some goals and principles of these data requests:

• Collect a small standardised set of basic but useful data from each library service in the UK
• Analyse each one to understand differences between services, and to identify trends
• Request data that is easily available within library services to limit the amount of work required
• Ensure the data is openly licensed and made widely available so that it can be used by anyone

This is not an attempt to replicate CIPFA, or the work that has been done by Libraries Connected, or the current work by Arts Council England (ACE). Alongside great sector support work, it is also important to have a public-first data strategy for library data. How that happens in future remains to be seen, but it should be on the radar of all services.

To some extent it is (rightly) easier for a member of the public to get data than it is for a sector support body. Those organisations can’t make demands of services, or insist they send data. In many cases they may be restricted by policies in the requests they can make of public services. Despite that it’s hard enough for the public as well, and FOI requests are not exactly easy going. We’ll see how it goes!

So, we should be thankful for Freedom of Information, and the opportunities and rights it gives. But library services could also be working towards a future where data is published carefully, and made widely open. For the public benefit and also their own.