Code Number: 035-120-E
Division Number: VI
Professional Group: Statistics
Joint Meeting with: -
Meeting Number: 120
Simultaneous Interpretation:   No

Canada's National Core Library Statistics Program:
First Steps in Developing Key Performance Indicators

Abstract

Paper

Introduction

The National Core Library Statistics Program was established to coordinate, collect, analyze, and report library statistics in Canada. It is a joint effort of the National Library of Canada and an advisory committee comprised of representatives of various library associations and agencies.

The first year of operation was 1995-96, during which 1994 statistics were collected. In 1997-98, 1995 statistics were similarly processed, and in 1998-99 the third year of the Program was completed covering 1996 statistics. Plans are underway to continue the Program for 1997.

The types of libraries currently included in the Program fall into three broadly defined sectors: public, academic, and special. School libraries have not been represented in the Program so far, although public libraries housed in schools are included.

In order to identify the effectiveness of libraries as Canada's primary access agency, the Program collects library statistics in 24 categories broadly grouped as follows:

• general information (institutional characteristics) - 6 questions
• services (information and circulation transactions) - 2 questions
• staff - 5 questions (4 questions in 1994 and 1995)
• collections - 3 questions
• interlibrary loan - 4 questions (8 questions in 1994 and 1995)
• expenditures - 4 questions.

Two versions of the database containing these statistics are available for public use. One version contains the 1996 data and the other contains the combined 1994-1995-1996 data for those libraries participating in all three years of the Program. Both versions are in the file format SPSS 8.0 for Windows. For reasons of confidentiality and anonymity, libraries are not identified by name in either of the publicly-available databases, but there is provision for a numeric library code that permits individual libraries to contact the National Library to verify their own institutional data.

The 1996 database reports information for 1,643 libraries in Canada. By type of library, the database represents

• 998 public libraries and 6 provincial and territorial libraries
• 208 academic libraries
• 430 special libraries
• 1 national library (the National Library of Canada).

The combined 1994-1995-1996 database reports information for 1,273 libraries, as follows:

• 906 public libraries and 5 provincial and territorial libraries
• 157 academic libraries
• 204 special libraries
• 1 national library (the National Library of Canada).

Library Patterns, 1996

The 1996 survey of the National Core Library Statistics Program shows that Canadian librarians and their staff in the public, academic, and special library sectors provided impressive cultural and economic benefits to the people of Canada. While many of these benefits are intangible and unmeasurable, the National Core Library Statistics Program was designed to capture several key indicators of the broad impact on Canadian society of library services provided by these three sectors.

One key indicator is question answering services. While many library services in public, academic, and special libraries are organized for self-help, Canadians also seek library staff assistance on a regular basis. In 1996, Canadian library users asked librarians and staff in these types of libraries more than 30 million questions, on average 2,000 inquiries per day every day of the year. This works out to at least one question asked of a librarian somewhere in Canada in 1996 by every woman, man, and school-aged child in the country. By far the greatest proportion of inquiries was addressed to public libraries, some 77%.

Another key indicator is library collection use. In general, Canadians show a high interest in the intellectual capital held by these libraries. In 1996, they borrowed well over 300 million publications (print and other materials) to satisfy their reading, viewing, listening, and other consultative needs. Approximately one million of these items were obtained through interlibrary loan arrangements. When on-site use of library holdings is also factored in (on the order of 50 to 150 million items), it is evident that Canadians are voracious and eager consumers of the cultural materials collected by librarians. (The Program does not collect statistics for on-site use of collections - the preceding range is an estimate based on unrelated research.)

On average, Canadians consulted well over one million library items per day in 1996. This works out to 12 or 13 publications annually or one item per month for every woman, man, and school-aged child in the nation. The annualized "turnover rate" in 1996 (the theoretical ratio of library items in use to library items in collections) was greater than one, when both off-site and in-library consultations are taken into account. The greatest proportion of borrowings by Canadian library users was from public libraries, 80%.

These question answering and collection services, along with other important library programs not captured in the National Core Library Statistics Program, came from an infrastructure investment in 1996 involving more than 25,000 library staff, 300 million publications, and 4,000 library service points operated by 1,600 institutions across the country.

A key indicator of the economic impact of libraries on Canadian society is library expenditures. Policy makers should take note of the economic purchasing power of Canadian librarians. In 1996, they spent more than $1.4 billion to provide library services in the three sectors comprising the National Core Library Statistics Program. Of this total, staffing accounted for almost $900 million and new collections for more than $300 million. Another $16 million was spent on capital projects and equipment to maintain and upgrade library service facilities across the country. Public libraries accounted for half of all expenditures.

Library sector analysis is one important approach to understanding Canadian library service activities, benefits, and impacts. This analysis shows that, of the three library sectors represented in the National Core Library Statistics Program in 1996, Canadians looked to public libraries more often than to any other type of library for their question answering and collection borrowing needs, 77% of all inquiries and 80% of all circulations. At the same time, public libraries accounted for only 54% of all library staff, 31% of all collection holdings, and 63% of all service points. In contrast, academic libraries had by far the largest collection infrastructure, 56% of all collection holdings, suggesting perhaps that they are organized for more self-assistance than are other sectors.

Jurisdictional analysis is another approach to service analysis and measurement. The most dramatic comparison here is between Ontario and Quebec libraries in the three sectors represented in the National Core Library Statistics Program in 1996. While Ontario accounted for 37% of the Canadian population represented in the Program in that year, the cultural and economic impact of its libraries on provincial residents was considerably higher: 52% of all inquiries; 41% of all circulation; 44% of all collection holdings; 45% of all library staff; 43% of all service points; 44% of all expenditures.

In contrast, Quebec libraries accounted for 25% of the Canadian population but their impact on provincial residents was considerably less:

7% of all inquiries; only 17% of all circulation and collection holdings; only 19% of all library staff; less than 8% of all service points; and only 21% of all expenditures. No other province or territory showed this much variance between population and library impact in either cultural or economic spheres.

Library Trends, 1994-1996

A third way of understanding the cultural and economic impact of Canadian libraries in the three sectors participating in the National Core Library Statistics Program is by means of trend analysis. A multi-year comparative approach cuts across both sector analysis and jurisdictional analysis to provide deeper descriptions of library activity and impact.

Over the three years of data collected for the Program, 1994 through 1996, patterns of library resources and use have not changed appreciably. Median inquiries increased 18% over the three years, from 2,200 per library in 1994 to 2,600 in 1996. Median circulation did not change over the three years, however, hovering around 21,000 items per library. Median library staff remained also relatively constant around 3.6 full-time equivalent employees per library. Median collection size increased from 23,000 items in 1994 to 25,000 in 1996, a 9% increase. Median service points remained constant at one per library.

Median library expenditures in 1994 and 1996 were virtually unchanged at $127,000 per library, but 1995 figures showed a median of $141,000.

There are, however, several caveats to these findings. The first is that only a subset of all libraries participating in the 1996 National Core Library Statistics Program also reported data in 1994 and 1995, some 77% or a total of 1,273 participants in all three years compared to 1,643 in 1996. Moreover, not all of these 1,273 libraries reported data for every survey variable, with omissions in key variables ranging from 7% failing to report total expenditures to 32% not reporting user inquiries.

Finally, a third caveat is that the statistical pattern represented in every one of the variables is extremely skewed, not only among the three library sectors but also within distinct types of libraries. So, for example, the overall mean number of library user inquiries in 1996 was 26,000, but the median was only 2,000 and the range was more than 2 million. Library circulation was similarly skewed, with a mean of 190,000 items, a median of 20,000 items, and a range of almost 11 million items. Skewed patterns mean that data averaging both across and within sectors can mislead: variances and subgroup patterns are hidden.

In recognition of these limitations and of others relating to methodology, median values were deemed to be more stable than means for comparison across the three years of the Program.

Conclusions

At the present time, the National Core Library Statistics Program does not include school libraries and as a result the picture of library benefits and impacts presented here is necessarily incomplete. In addition, it is incomplete because, even for the three sectors currently reported in the National Core Library Statistics Program, not all institutions participated.

It should also be noted that a number of library service dimensions of considerable magnitude are not captured in the measurements of benefits and impacts reported by the National Core Library Statistics Program. Notable among these are overall user satisfaction, on-site use of collections, library program attendance, library membership, personal visits and "virtual" visits, and self-help electronic information access including the use of networked and multimedia CD-ROMs, Internet web pages, and library catalogues both on-site and remote access.

While many of the electronic services were in their infancy in 1996, standardized and universal measurement data to describe their impact on Canadian culture and the economy will increase in importance as information technology becomes more and more pervasive in the delivery of library and information services and as the shift from ownership to access intensifies.

The ultimate challenges for both library policy makers and Canadian library users are, first, to articulate how they value libraries and, second, to develop persuasive methods of articulating that value. Of particular concern is the avoidance of reductionist impulses, such as the conflation of library inquiries and circulation data that produces an aggregate but oversimplified measure of library "activity." Given the multidimensional nature of library services, benefits, and impacts, the articulation of library value must be correspondingly multidimensional, standardized, and universal.

Bibliography

1. Kinnucan, Mark T. "National Core Library Statistics Program." Feliciter 43 (September 1997): 44-51.

2. Manning, Ralph W. "Counting Canadian Libraries." Feliciter 43 (September 1997): 40-43.

3. National Core Library Statistics Program: Statistical Report 1994. Prepared by Mark T. Kinnucan for the National Library of Canada. URL: http://www.nlc-bnc.ca/coopprog/enclsp.htm.

4. National Core Library Statistics Program: Statistical Report 1995. Prepared by Mark T. Kinnucan for the National Library of Canada. URL: http://www.nlc-bnc.ca/coopprog/enclsp.htm. National Core Library Statistics Program: Statistical Report 1995

5. National Core Library Statistics Program: Statistical Report 1996. Prepared by Alvin M. Schrader and Michael R. Brundin for the National Library of Canada (in press).