PLR consultation 2022
The following was sent to the DCMS, in response to the open consultation Public Lending Right: Rate Per Loan 2021 to 2022 consultation.
Public Lending Right: Rate Per Loan 2021 to 2022 consultation
I support Public Lending Right (PLR), and paying authors and contributors for lending books in libraries. It’s important that this is maintained, and continually improved.
I object to the proposed rate per loan increase. This is primarily due to it being incorrect, but also due to it being unsustainable.
I will raise various issues, relating to the rate and it’s calculation. I would recommend rejecting the British Library board’s proposal and requesting they look at fixing the rate per loan calculation, to re-align it with accurate lending data.
An ever-increasing rate per loan
In 1982, at the beginning of the PLR scheme, the rate per loan was 0.5p. If the rate had remained stable, accounting for inflation, it would now be about 2p per loan.
It has steadily increased. This resulted in a peak of 11.26p last year, roughly 6 times what was initially proposed.
The Society of Authors say that for the sale of an £8.99 paperback, an author would typically receive 67p. Within PLR counting, ‘renewing’ a loan counts as an additional loan. As readers often renew a book many times, the public can pay a greater royalty for a single borrower reading a book, than the author would receive from the book being sold. An increasing rate per loan is hard to justify if nothing is done to address this issue.
Problem 1. An ever-increasing rate per loan is unsustainable and brings the validity of the overall fund into question
The proposed increase to 30.53p creates an extraordinarily high rate. It should be frozen, and any excess funds (plus additional funding if required) used to fix and improve the PLR process.
Proposal: Freeze the rate per loan and invest in improving the PLR process and promoting library lending
Invalid data used in calculations
It is understandable that the pandemic caused a reduction in loans, and triggered the proposed increase in rate per loan. This reduction is likely to have been more than estimated. Libraries were regularly ‘auto-renewing’ loans for users, leading to over-inflated lending figures, regardless of official rules as to which loans count towards PLR.
The rate of 30.53p is calculated from a year of loan sampling that ended in June 2022. That sample is ‘totalled-up’ using CIPFA total lending figures. However, those totals are taken from financial year 2020/2021. The CIPFA data is 15 months out of date.
This methodology relies on loans being similar between years. When the pandemic hit, this failed, along with the rate per loan calculation. The total lending figures are not just inaccurate, they bear no relation to the year they are used for. This cannot be seen as an awkward technicality, caused by exceptional circumstance. It is a process that was put to the test, failed, and needs to be fixed satisfactorily. There is no validity in adjusting the rate per loan based upon data that we know to be unrepresentative.
Further to this, the CIPFA returns from library services have been steadily declining year on year. For 2020/2021 they represented around two thirds of services. In some regions, such as Wales, CIPFA data hit a low of being less than a quarter of library services, and fewer than in the PLR sample. CIPFA data no longer comes close to estimating total lending, and cannot be used for that purpose.
Problem 2. The estimation of total loans, and rate per loan, is fragile, failed when put to the test, and is reliant on out-of-date, unreliable, and declining third-party data
PLR processes need to be self-sufficient, and must ‘sample’ all library services, all of the time. This is no longer a significant challenge. There are relatively few Library Management System vendors in operation in the UK, and they support PLR reporting.
Ensuring that all services always report data could result in less administration than keeping a rotating sample, which includes liasing with services when they are added and removed from the scheme, and associated work for those services.
Proposal: Make changes to the sampling policy so that all library services in the UK submit loans data at all times
Incorrect information presented to authors
Authors and contributors receive information on the loans of their books, in statements and online reports. These will be significantly incorrect.
Take this public tweet from an author expressing joy that their book was borrowed over 1,200 times, from last year’s statement:
Dead chuffed with my @PLR_UK statement - #SightUnseen was borrowed over 1200 times! That makes me happy 😊😊 Thanks to all the lovely #readers and fantastic #library peeps! ❤️
That kind of delight is one of the beautiful things about PLR: authors are overjoyed to see the number of times their works are borrowed in libraries. But this was inaccurate. The total loan estimations for last year statements were taken from 2019/2020 CIPFA data, which didn’t represent the drop in lending from the pandemic.
The loans of that book may have been 500 or less. Still something to be delighted about, but providing inaccurate information is doing authors a disservice.
Problem 3. The statements on lending are significantly incorrect, and misleading to authors and contributors
It is reasonable for authors to expect accuracy, as well as payment, in their statements. Maintaining a system that we know gives increasingly invalid data could be considered to be negligence in fulfilling the PLR function.
The reporting of loans data should be increased in frequency to be on a nightly basis rather than every other month, and this additional detail should be part of reports and online statements. Authors (and publishers) should know, for example, which days of the week their books are most commonly borrowed.
To aid in this change, extra automation should be added so that reporting is automatic from library system to a central data repository. Although this would need development work, primarily from LMS vendors, there are other examples of this kind of reporting in library services, and no technical barrier to it being achieved.
Proposal: Enhance the sampling to increase the regularity of reporting and ensure data transfers are automated
Restricted access to PLR data
Despite problems, PLR sampling is still the most significant dataset of library lending across the UK. It includes not just statistics (e.g. counts of books lent) but individual titles and the month they were lent. It is a dataset that can inform on lending trends across seasons, genres, child and adult fiction, reference, etc.
During the pandemic, library organisations needed to understand the extent to which library lending had been affected. As services reopened, work was also required to analyse how lending recovery compared with other public trends, such as high street footfall.
Libraries Connected ran data collection exercises to try and gain this insight. This involved liaising with library services and collating this data on a month-by-month basis. Although impressive, engagement with this was patchy, as well as time-consuming.
In combined staff effort, this would have amounted to many thousands of pounds spent compiling that data. All the while, PLR sampling was ongoing, both for physical and e-lending. But the data is not made widely available, despite being invaluable.
Just last month (October 2022) Libraries Hacked received a query from a researcher, asking about availability of lending data, specifically for library services in Scotland.
“I am trying to put together an example of how to use data from different sources to produce dashboards and understand trends over time. I was hoping to use Scotland’s public library borrowing data to do this, however it doesn’t appear to be easily available
[…]
The other source I found was the UK-wide Public Lending Right (PLR). Their latest data (which I think is from their sample set of libraries) is from 2020/21 but it only provides a pdf of the headlines not the underlying data. Is there a reason that this data is not openly available do you know?”
Query received to [email protected]
Such queries aren’t uncommon. My answer is that unfortunately PLR data is one of the only combined sources on library lending, but that it is unavailable to the public, with no valid reason for it not being openly available.
There is no doubt that this situation actively harms library services across the UK. Publishing data on library lending is not just a ‘nice to have’, it is essential for 21st century library services to operate effectively within a data-rich society and economy. Collecting data is complex and time-consuming. The few sources already in place need to be fully utilised, improved, and made available for wide re-use. This would also help the library sector comply with local and central government open data policy.
The goal should be to promote library lending to higher levels, as well as publishing data for the library sector, publishers, and the public, to gain insight into library lending.
Problem 4. Essential library lending data is unavailable to the wider sector and public, which leaves library services unable to operate effectively, unable to comply with data policies, and hinders sector innovation
With additional improvements in coverage and frequency, PLR loans data would be transformative for public libraries. To have significant impact it would need to be widely available on an open licence. It should be able to be used by library services, by the public, and for commercial applications, without restriction.
A key Government Design Principle is ‘Make things open it makes them better’. Libraries are the originator of this principle through open knowledge and information. It is long overdue for modern library services to follow this with data publishing.
I firmly believe that the key to tackling declining library lending (and the increasing rate per loan), is to increase the scope for library lending data analysis. It is not a ‘silver bullet’, and will not in itself increase loans, but is an essential first step for any positive action. Without access to good data, there are no actions that can be taken that are effective or data-informed.
Proposal: Make extended PLR data widely available for download, openly licensed, and available as soon as possible after it is submitted by library service systems.