Thursday, August 11, 2011

PL 40/11: Hard work for OCLC

Filed under: IFLA, statistics — Tags: — plinius @ 1:30 pm

OCLC is one of the pillars of the global library community.

The organization has recently started to compile library statistics at the global level. Trying to do something in this area is praiseworthy. Doing it well is almost impossible, however.  Several decades ago Unesco tried to collect global library statistics. They abandoned the attempt.  OCLC  may be able to devote more resources to this task. But producing meaningful and useful statistics by collecting data from above is much harder than people think.

I have three main reasons for that statement:

  1. Low demand. The actual demand, among librarians and library researchers, for comparative national statistics, is low. There are very few professional articles in this area, even in regions where good data are available – like Scandinavia and Finland.
  2. Scattered data. Keeping up to date with the scattered data that exist takes too much time and effort (relative to the demand) for even a rich organization.
  3. Weak systems. The systems that collect data at the national level are far too weak, in the great majority of the world’s two hundred countries ( Are there fifty good national systems for collecting, processing and publishing library statistics? I doubt it.

OCLC says: These statistics that represent the total global library universe include data, if available, for the total number of

  • libraries
  • librarians
  • volumes
  • expenditures
  • and users

for every country and territory in the world, broken down into the major library types:

  • academic
  • public
  • school
  • special
  • and national

The statistics also include available data for languages used, and the number of

  • library schools
  • publishers
  • and museums

The three regions represent those used by the OCLC Global and Regional Councils: the Americas (North and South America), EMEA (Europe, the Middle East, Africa, and India), and Asia Pacific (Asia, Australia and Oceania).

As soon as OCLC goes into details, problems crop up:

The staff of the OCLC Library extracted data from respected third-party sources, both electronic and print, that in their judgment are the most current and accurate sources to which they have access.

  1. For many countries, data were either unavailable (indicated in the charts as NA) or sporadic.
  2. Also, for a lot of the world, the data were not as current as the compilers would have liked.
  3. Many of the printed sources were several years old and many websites were suspect.
  4. It was felt, however, that a fairly recent figure was better than none at all, with a cut-off date of around 1980.
  5. The OCLC Library staff had to determine which of the sometimes several possible sources was the most reliable in terms of accuracy and currency.
  6. Seldom did the data from two or more different sources coincide for the same year.
  7. For the sake of uniformity, the total number of libraries represents administrative units and not service points, since not all sources report service points consistently.
  8. Some of the country entries did not specify which unit they were reporting, in which case it was assumed that it was administrative units.

The sources are the best OCLC has access to. That is a very weak guarantee.

Can researchers depend on these statistics? Or library advocates? From the description above, for instance (3), (6) and (8), I would not trust these data.

But let us look at how OCLC handles Norway. This is a country with excellent library statistics, comparatively  speaking. The results are not encouraging.

OCLC reports the following numbers on its web site:

Academic libraries (all data from 2009)

  • 322 libraries
  • 728 librarians
  • 171,373 users
  • 13,402,720 volumes
  • $ 166,061,000 in expenditures


  • The number of libraries (322) is taken from IFLA World Report. This number is far too high, since it includes special as well as academic libraries. The correct number is 145 (source).
  • The number of librarians in academic libraries (728) in 2009 is correct. The link provided by OCLC now takes us to the 2010 data, however. Such updating of links to the newest data often happens – and is very hard to notice from above. In 2010 the number of librarians was 775.
  • The number of full-time equivalents (FTEs) is actually a better measure than the number of persons, since many librarians work part-time, especially in  public and school libraries. The 728 academic librarians corresponded to 669 FTEs.
  • The numbers of users (171 tousand) actually refers to the number of borrowers in academic libraries in 2009. People may of course use the library actively without borrowing a single document.
  • The number of volumes (13.4 million) is correct (2009).
  • The expenses in 2009 were 965 million NOK. The corresponding dollar value will always fluctuate, of course.
  • During 2009 the dollars fell from 7.0 to 5.8 against the Norwegian crown. The total expenses measured in in dollars increased accordingly, from 138 million USD at the beginning to 166 million USD at the end of the year.
  • OCLC must have used a conversion rate from the late autumn. A mid-year rate, or an annual average, would have been better (such data are easy to find). The level of accuracy ($ nnn.061 million) is meaningless.

National libraries

  • 1 library
  • 178 librarians
  • 3,793 users
  • 3,116,781 volumes
  • $ 62,182,500 in expenditures


  • The number of librarians (178) is correct. It corresponds to 123 FTEs.
  • But the link leads us to 2010, where the numbers are 143 and 136.5 respectively. The number of librarians went down by twenty-five percent. The number of FTEs went up by ten percent.
  • It seems as if many part-time positions disappeared or were converted into full-time employment. I do not know what happened inside the organization. But from a statistical point of view the professional resources of the National Library did not decline by twenty-five, but increased by ten percent from 2009 to 2010.
  • The number of registered borrowers (3,793) and volumes (3,116 thousand) is correct.
  • The trailing digits (… 781) are unnecessary.
  • The expenses in USD depend on the rate of exchange.

Public libraries

  • 841 libraries (2010)
  • 930 librarians (2004)
  • 1,314,521 users (2001)
  • 16,468,702 volumes (2008)
  • $ 134,336,000 in expenditures (2004)


  • The number of libraries is taken from IFLA World Report for 2010. It does not show the number of administrative units in 2010 (about 430) but the number of branches around 2004-2005. (See link, p. 44)
  • The number of librarians is taken from a summary in English (probably the most recent one available) of the annual statistical report from 2004 (!).
  • Data for every year up to 2010 are of course available, but not in English.
  • Data on users are from the old LibEcon study, which was published in 2001. The data themselves must be older.
  • I will not try to trace LibEcon’s source. The number has no practical use.
  • A series of recent surveys show that approximately half the population use the public library. The current number of users must be well above two million.
  • The current number (2009) of borrowers can also be found,  in a recently published spreadsheet from the National Library.
  • The number of volumes (16.5 million) given for 2008 is taken from the CD-rom version of Saur’s World guide to libraries (2011). The correct number for 2008, in the web version of the statistical yearbook for 2008, is 19.5 million books.
  • Information is available for 2009 and 2010.
  • The expenditure ($ 134 million) comes from the English language summary mentioned above. The published number for 2009, is 1,319 million NOK or approximately USD 206 million, based on the mid-year conversion rate.

School libraries

  • 246 librarians (2004)
  • 2,783 libraries (2009)
  • 486,294 users (1987)
  • 8,350,884  volumes (2001)
  • $ 16,490,103  in expenditures (2001)


  • The number of qualified librarians only refer to the upper level (secondary schools). The current number (2009) is 277.
  • The number of libraries, on the other hand, refer to all types of schools. The number of elementary and middle schools is five or six times as high as the number of secondary schools (10th-12th grade).
  • The number of users again refers to all types of schools. Referring to the ancient World Encyclopedia of Library and Information Services (ALA, 1993) as a statistical source – for the number of Norwegian school children in 2009 – is not a practice I would recommend.
  • All students are allowed, and even supposed, to use their school libraries. The number of users equals the number of students. Educational statistics are always available from Statistics Norway.
  • The number of volumes is taken from LibEcon. It only refers to the lower levels of schooling (grades 1-9) – around 2000 (source).
  • These libraries have currently about ten million books (but lots of them are outdated …).
  • In 2008 the secondary schools had about three million volumes (source).


The library statistics that OCLC offer for Norway can not be used for serious work. The quality is far too uneven. OCLC is facing the same problem that Unesco faced. Countries with weak library systems have hardly any statistics at all. There may be some numbers floating around, but they do not reflect the situation on the ground. Countries with stronger library systems, have better statistics, but they do not publish them in a form that is convenient for outsiders.

Concepts, definitions and procedures often differ from country to country – and sometimes within countries as well. To collect and interpret, say, French or German or Polish library statistics, you need access to their statistical sources, an understanding of their social systems (libraries, schools, universties, volunteers) and, of course, a smattering of the languages. Countries with good library systems have good statistics, and put the data on the web, but you still have to handle the local languages.

OCLC  have tried to use secondary sources like:

  • IFLA World Report
  • The LibEcon report
  • World Encyclopedia of Library and Information Services

That does not work. All compilers face the same problems. Quality can only be realized country by country, by people who are willing to spend a fair amount of time monitoring what goes on. Even for little Norway, that takes an effort.

The next step: processing the data so that others can use them, takes even more time. But that is the only way, I think, to produce workable statistics in the global library field.


1 Comment »

  1. […] PL 40/11: Hard work for OCLC. With Norway as a case we show how hard it is to get reliable and current statistics from many countries. […]

    Pingback by PL 43/11 « Plinius — Tuesday, August 16, 2011 @ 3:51 pm

RSS feed for comments on this post. TrackBack URI

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

Blog at

%d bloggers like this: