Libraries Hacked

Promoting open data and digital prototypes in libraries

3 minute read

On 11th December 2017 CIPFA released their annual library statistics for financial year 2016/2017. Headline figures include a drop in budgets by £66m, and 105 library closures.

On the same day, CILIP called for a sector-led Open Data working group:

CILIP will be calling on ACE to engage their Research and Evidence Team to support the proposed Open Data Working Group. At the same time, CILIP is calling on Department of Culture, Media and Sport to support the sector-led data initiative and divert existing resources away from the CIPFA service - with the ultimate aim of a regular open data supply through the data.gov portal.

Current public libraries statistics

How does the current process work, and why should it be diverted away from CIPFA?

  1. A CIPFA working group meet a couple of times a year to ensure survey questions are still relevant.
  2. After each financial year, library services are sent a spreadsheet by CIPFA and asked to complete it
  3. Services complete this between April and July
  4. These are submitted to CIPFA in July
  5. In December the data is ready, alongside reports and data tools, available only to CIPFA subscribers.

This aggregated data is owned by CIPFA, at the point it is released in December. The returns each authority complete are publicly-owned, but are rarely made openly available.

Subscription fees to CIPFA are unaffordable even for some library services. A single report can normally be purchased for around £450, though publishing research from this data would need to be approved by CIPFA. Having public data held and controlled by a private organisation will only lead to services giving up on any coordinated data work.

There are at least two stages at which the public pay for this. Firstly the data collection itself - hundreds of library services dedicating many hours each year to complete the stats. And then the subscriptions that local authorities pay to view the data.

In 2016, it was highlighted on Twitter that the DCMS paid CIPFA £48,122 for comparative profiles - summary PDFs that include graphs for each service. These have been sponsored since 2012. That is another significant amount paid by the public to CIPFA for data they’re not allowed to access.

The above isn’t compatible with central and local government policy for open data. Even if it were satisfactory for the library sector, it remains poor value for the public. These issues are described by Nick Poole in CILIP’s press release:

“Currently, data that is paid for by the taxpayer is only available behind a considerable paywall. This means that contrary to Government commitments on transparency and open data, the sector, the public and interested parties cannot access vital information about the UK’s school, prison, health and public libraries”

An alternative

There are established portals that local government should be be publishing to, and are designed to hold data on public services. See data.gov.uk or even LG Inform Plus, which provides data aggregation services. From there it would be possible for anyone, including CIPFA, to extract and aggregate the data as necessary.

A new process could look like:

  1. A public library working group meet regularly to discuss data collection standards
  2. Library services maintain a core dataset of library data. Initially this could be co-ordinated to provide annual data (like CIPFA), but some data, such as usage, could be maintained at more frequent intervals.
  3. The data would be maintained on data.gov.uk

CIPFA could still add value to this process, if they choose to. If service managers wish to read PDFs of how their library service compares to others, any organisation, including CIPFA, can offer this. However, CIPFA would not own the data, and they would be in competition with other data specialists.

Ideally, insight from this data would be achievable in the public sector through investment in data skills. For example, being able to calculate how many borrowers there are per 1000 population, and compare this to neighbouring services. This will be making use of other available open data, as CIPFA currently do.

The Libraries Taskforce have been working on definitions for a core dataset for public libraries. Although still relatively early days, it’s important that library services get involved and take ownership of this process, to collaborate on what data they currently collect internally, and what they could make open. Everyone (aside from CIPFA) shares the same frustrations with the current situation, it’s down to library services and leadership organisations to do something about it.

And publish CIPFA returns as open data, don’t send them to CIPFA.