As of May 21, 2004, the National Library of Canada (NLC) and the National Archives of Canada (NAC) became Library and Archives Canada (LAC).

The Use of NLC MARC Records in Canadian Libraries

Phase I: University and Large Urban Public Libraries

Executive Summary

The Graduate School of Library and Information Studies, McGill University, undertook a study on behalf of the National Library of Canada (NLC) to determine the dollar value of savings incurred by Canadian university and large urban public libraries as a result of using Canadiana printed monograph cataloguing records (including federal government documents) generated by the NLC rather than cataloguing these items themselves, and to propose a methodology by which this Phase I study could be extended to other Canadian libraries.

This study was conducted between January and March 2002, using three methodological approaches: questionnaires were sent to 90 Canadian university and college libraries, and to 30 member libraries of the Council of Administrators of Large Urban Public Libraries (CALUPL); follow-up telephone interviews were held with 18 university and 12 public libraries; and a sample of 100 bibliographic records for Canadiana printed documents was selected by the NLC from its 1999 cataloguing production and then compared with records in a sample of 20 university and 10 public library OPACs to determine the extent to which NLC records form the basis for copy cataloguing by other libraries.

The three methodologies together, when implemented with care and following guidelines set out in the Report, worked well to collect the required information, but none is perfect alone. The quantitative data collected remain estimates rather than precise figures, and it is extremely difficult to collect absolute data because the libraries themselves do not record it.

The record matching suggests that only 4% of the Canadiana documents catalogued by the NLC overall find their way into university and large public library catalogues. University libraries on average hold 7% of monograph titles and 11% of government document titles catalogued by NLC; the comparable figures for large public libraries are 2% in each case.

Questionnaire analysis (58% response rate) revealed that the saving per library through using NLC records as the basis for copy cataloguing rather than originally cataloguing items was $16,400 per annum for university libraries and $7,800 for large urban public libraries. An extrapolation to all university and large public libraries suggests an annual saving of $1,476,000 for all Canadian university libraries, and $249,000 for all Canadian large urban public libraries, a total of $1,725,000 as a result of NLC catalogue record use for copy cataloguing. The Report explains in some detail the problems encountered in establishing these figures and the riders that must be attached to their interpretation.

Many libraries make use of NLC name or series authority data (35% frequently and 47% occasionally). Libraries using NLC copy in their acquisitions processes or for other bibliographic purposes reap additional savings. The monetary benefits accruing to the libraries from these services and activities have not been quantified in this Report.

The most frequently used single sources of NLC catalogue records cited by the libraries were Amicus Online (75%), AG Canada (40%) and OCLC Online (35%), although other libraries' OPACs via Z39.50 gateways were cited by for 76% of the libraries and CIP data by 56%.

Both the questionnaires and the telephone survey reveal a considerable level of satisfaction amongst respondents for the cataloguing services provided by NLC. The main suggestions for improvements focused upon providing more detail in NLC records (especially LC classification numbers and subject headings, Dewey Decimal numbers, and subject headings for Canadian government documents); greater timeliness in the production of records; better coverage of government documents, non-print materials of various kinds, serials, local materials and Internet resources; harmonization of headings with LC, and better access to Amicus via Z39.50.

It seems probable that the two categories of libraries examined in this Phase I project are the easiest for which to compute savings accruing from the availability NLC copy catalogue data. They are limited in number, normally maintain detailed cataloguing statistics, and provide externally searchable OPACs using MARC record format. Nevertheless, a variety of problems were encountered in undertaking this study. Its extension to other kinds of libraries in all likelihood would require modifications to the methodology.

For more information on this study, please contact: