At IFLA, on Thursday August 12, we’ll do a full day workshop on statistics for advocacy.
Thomas Hennen – Director of Waukesha County Federated Library System in Wisconsin – constructed a widely used index of library quality for the US about ten years ago. In 2008 Keith Curry Lance and Ray Lyons designed a competing index.
Norway has 430 public library systems – one for each municipality. If I create a single library quality index for Norway – and group the libraries by quartiles –
- about one hundred libraries will be very satisfied – “among the top 25 percent”
- about one hundred will be rather satisfied – “above the middle”
- about one hundred will be rather dissatisfied satisfied – “below the middle”
- about one hundred will be quite unhappy – “close to the bottom”
Interenational universities – who compete aginst each other in the global market – must accept such quality ratings. But internationally only the top hundred or the top thousand universities count. Libraries are ranked locally, in national or regional settings. Here, even low ranks are painfully visible.
The answer must be: differentiate.
Libraries need to focus more on empirical data. Great values, good intentions and stories of individual benefit will not be enough. The political demand for accountability – documenting what you do – is growing. Give us numbers!
But we can differentiate much more. Let libraries compete in many sports: visits, loans, children’s services. Big libraries work under different conditions from small ones. For each particular sport we may divide the libraries into leagues based on size.
Below I illustrate – using Norwegian data.
The five leagues I have chosen include
- municipalities with more than fifty thousand inhabitants: 14 library systems
- twenty to fifty: 35 libraries
- ten to twenty: 57 libraries
- five to ten: 88 libraries
- less than five: 234 libraries
The leagues represent the following shares of the population (2009):
- 37 percent of the population live in municipalities with more than fifty thousand inhabitants:
- 22 percent in municipalities with twenty to fifty
- 17 percent in municipalities with ten to twenty
- 13 percent in municipalities with five to ten
- 12 percent in municipalities with less than five
In each league libraries may compete in selected “sports”. The choice of indicators requires a process that involves the library community. The extensive consultation process behind BIX may serve as an example.
I do not see this as a purely political process, however. Many technical aspects and arguments need to be brought into the discussion. But a straight technocratic approach, where statisticians decide and librarians follow, is unlikely to work. Decision-oriented statistical systems must combine technical quality and social legitimacy. Ulla Wimmer has made the same points with regard to the German index BIX.
Below I have selected twelve indicators to illustrate possible choices:
- Visits per capita
- Loans per capita
- Activity level (= loans + visits per capita)
- Loans of children’s books per child
- Loans of fiction books per adult
- Loans of non-fiction books per adult
- Turnover of non-fiction books
- Loans of non-book media per capita
- Operational costs (=salaries + media expenses per capita)
- Productivity (=activity/operational costs)
- Reference questions per 100 visits
- Active users per 100 capita
At the moment I can only calculate values for the first ten. Data for (11) and (12) are collected by The Norwegian Archive, Library and Museum Authority (ABM-utvikling), but are not published in a reusable digital format.
In the Metropolitan League (more than fifty thousand), the 2009 results were:
Visits per capita
Loans per capita
Loans of children’s books per child
Loans of fiction books per adult
Loans of non-fiction books per adult
Turnover of non-fiction books
Loans of non-book media per capita
Operational costs per capita
Asker is clearly the outstanding library in this league, ranking first on five different categories. But money explains part of their (well-deserved) success. The most efficient libraries in this category are Trondheim, Tromsø and Bergen. We also note that as many as eight libraries:
can boast of a top three position – on at least one of the ten indicators.
- Public library ratings. Wikipedia.
- Ray Lyons. Public Libraries Statistical reports
- Frequently Asked Questions on the HAPLR Index
- BIX-Information in English
- PL 52/10: Library ratings. Two US quality indexes compete for attention