4 minute read

Many public libraries authorities in the UK submit CIPFA returns each year, answering a set of pre-defined performance questions. These include counts of loans, visits, service points, expenditure, and more. They’re sent to the Chartered Institute of Public Finance and Accounting (CIPFA), compiled, and released as annual library reports. These are available to subscribers, or for £475+VAT.

The data collected by each authority is rarely made publicly available, and there are no examples of real open data. There are a number of historic Freedom of Information requests for them on the What do they know site.

When the data is compiled it is no longer public. As CIPFA pointed out on Twitter:

The majority of the work to collect the data is paid for by the public, and undertaken by each authority. As authorities subscribe to CIPFA in order to view the aggregated data, this is then paid for again by the public. Then the DCMS pay CIPFA around £42,000 in order to release a set of PDFs comparing each authority.

This may have been an efficient method of compiling statistics a number of years ago, but there are public data portals that can hold such data and make it freely available. As well as being able to aggregate, combine, and compare the data for different authorities - see data.gov.uk and LG Inform Plus.

The concept of an annual survey is becoming outdated. There are local government systems that set up automated data exports to a schedule. These could be as often as is appropriate - every day, every month, or even every 5 minutes. Not all data collected for the CIPFA annual returns will support automated exports, but much of it could be done at far more valuable intervals.

Why make data open?

If such data is made open, what is to stop people using it ‘badly’, or releasing analysis that is flawed?

There are already many arguments against treating the public as untrustworthy data consumers, but such concerns rarely become an issue. A bigger danger is with closed-data, only available to select local authority leaders. This can be used for poor policy-making, without the public having the opportunity of checking the data themselves.

An example of this is highlighted in a Brent budget report in December 2014. This used CIPFA data to suggest reducing spend on stock in libraries:

Benchmarking with other London boroughs shows that Brent spends a relatively [high] proportion of money on library books, audio visual material and online resources in comparison (comparing number of libraries and population). Reducing the stockfund by £100,000 to £450,000 would see Brent with just above average stock spending in comparison to other London boroughs. For the stock spending per 1000 population’ indicator the reduction would see Brent move from 3rd to 2nd quartile in the CIPFA tables for London boroughs.

Brent Budget consultation document

There is an argument that expenditure data is useful, as one of many measures when analysing aspects of a service, and when merging this with other data (such as usage) to identify links. In this case it is taken as a single indicator to inform policy, and the data isn’t detailed enough to make such decisions. It is also very selective. In the instance of Brent, overall expenditure was significantly lower than average. It was only in the category for materials that the spend was above average. Any further reduction would then reduce an already below-average spend.

It leads to questioning how many authorities use CIPFA data in this way. If above average authorities in specific areas consider this a reason to get themselves in line with the average, those averages will consistently drop. This makes any downward trend a self-fulfilling one. This CIPFA management information service then serves as a race to the bottom in library investment.

Library usage and expenditure data is complex, and will always remain so. Limited and aggregated annual statistics will only serve to inform poor policy decisions, causing a negative effect on performance. It is often the case that some data is better than nothing - with poor data that can’t be scrutinised it can be far worse.

With detailed open data the public can oversee decisions made by their local authority, and analyse the data themselves to provide important local (and national) insights. That is their right, in the same way that they have a right to a library service.