As of 22 April 2009 this website is 'frozen' in time — see the current IFLA websites

This old website and all of its content will stay on as archive – http://archive.ifla.org

IFLANET home - International Federation of Library Associations and InstitutionsAnnual ConferenceSearchContacts

62nd IFLA General Conference - Conference Proceedings - August 25-31, 1996

A Comparative Study of National Public Library Statistical Systems and Management in The United States and People's Republic of China

Yan Quan Liu
1) Doctoral Student
University of Wisconsin Madison
2) Librarian Obj.
University of Wisconsin Milwaukee


Through an analytical and comparative study, this paper examines the development and current status of the United States' and China's national library statistical systems, their natural functions and key system characteristics, usability of performance measures, and each system's strengths and shortcomings. A distinction is made between government centered and profession centered systems. Four f actors are identified influencing the systems' characteristics. By finding commonalities and discrepancies between the two countries' statistical systems, the article provides library professionals, governmental administrators, and educators with the opportunity to learn from each other's practices and experiences, to conduct further investigations into the purpose and characteristics of the stat istical systems, and ultimately, to improve nationwide library planning and evaluation in their own country.



National statistical measures of library operations have been used in both the United States and People's Republic of China for over half a century. These national systems of statistical measurement provide a variety of information concerning service effectiveness, operational management, staff administration, national planning, and professional competitiveness. However, compared with the effort s expended in collection and dissemination of the numerical information, analyses of the differing character of national systems or of the contributions of these systems have rarely received attention. A study that compares characteristics of these systems can increase understanding of the function of each and can suggest approaches to improving their effectiveness.

Objectives of this study therefore are fourfold. First, the national statistical system in both the U.S. and P.R. China are described. Second, key characteristics of these system are examined. These include system function, data manipulation, and standardization and implementation. Third, the relation of the data utilization and library performance measures is discussed. Initial comparison betwe en two countries of the national systems is supplied, and is based on the latest information on each system that I can find. Finally, an attempt is made to provide some insight into possible future developments as well as a system assessment.

Description of the National Systems Government Centered versus Profession Centered Systems

With long library histories and large populations of public libraries, both countries have a natural need to trace meaningful numbers about libraries. In the U.S., plans to collect library statistics on a national basis began approximately one hundred years ago. The first milestone in public library statistics occurred in the end of 19th century. In a matter of nine years, after the establishmen t of the U.S. Office of Education in 1867, "a massive report, Public libraries in the United States, containing descriptive and statistical data was issued" (Adams, 1990). In P.R. China, public library statistics can be traced back to the beginning of this century. The first nationwide library survey authorized by Ministry of Culture of China was distributed in 1950 (Li, 1988). This was ju st one year earlier than the dispatch of a very simple questionnaire from the UNESCO Office of Statistics in 1951. The practice of collecting national public library statistics in both the U.S. and China preceded the UNESCO's request for international collaboration on library data collection (Thi, 1987).

After years of efforts to find a better way to report useful and up to date data on public libraries, librarians in both the countries have built up their own unique statistical systems on national public libraries. Such systems, this study's subjects, can be categorized as government centered systems or professional or non governmental systems. The distinction between the two kinds of systems i s based on the clients served by each system and their missions. The systems of the U.S. and China, respectively, are described below.

The Systems in the United States

  1. Formally started in 1988 (Lynch, 1989), the current U.S. government public library data system is funded by the National Center for Education Statistics (NCES), U.S. Department of Education Office of Educational Research and Improvement, and conducted through the Federal State Cooperative System for Public Library Data (FSCS). The support of the system includes the local public librarians wh o provide the data; the FSCS State Data Coordinators who collect and verify data at the state level; the Chief Officers of State Library Agencies; the American Library Association (ALA); and the National Commission on Libraries and Information Science (NCLIS) which collects data disks from Data Coordinators. To run this system, NCES is the executive agency and funds the program; FSCS is responsib le for questionnaire design and definitions of standardization. Over ninety five percent of public libraries anywhere in the country respond to the system yearly. Its clients include governmental officers, legislators, and professionals. As of 1995, the system issues the U.S. public library data on paper, diskette and the Internet each year.

  2. . The Public Library Data Service (PLDS) is the major public library professional system of the United States. The system, in which data were collected starting in 1987 (Johnson, 1993), is supported by Public Library Association (PLA), a division of American Library Association (ALA). Its executive agencies have included School of Library and Information Studies at University of Wisconsin Mad ison, and Library Research Center at the University of Illinois at Urbana Champaign. As a voluntary data collection project this system annually collects data from public libraries of all sizes, almost covering libraries serving more than 100,000 population (79.8% were included in 1995). Data collected in February/March are published in an annual report in June. Clients include public library dir ectors and library educators. The system is designed by the public library profession for its own use. Running for eight years, this system has become an invaluable reference for librarians, trustees, governmental officials and the media.

The Systems in People's Republic of China

  1. As a part of the statistical system on national social cultural affairs, statistics on public libraries in China are officially reported by the Ministry of Culture. The Ministry gathers library data through local government agencies and reports the data to the Bureau of National Statistics, the national statistical office. Data obtained at the end of each year are reported early the next yea r, and are published in The Annual Report of Public Library Statistical Records (PLSR). This government funded system was established in early 1950s, and has been continuing for four decades, although information during Cultural Revolution is fragmentary.

  2. Using the basic pattern of data design in PLSR, the National Library and Bureau of Library Administration (a unit of the Ministry of Culture) have conducted national public library survey and published Overall Conditions of China Public Libraries (OCCPL) irregularly every several years starting in 1959 (OCCPL, 1989). Statistical information published includes not only numerical data of the libraries but also narrative descriptions of the history and physical facilities. The summerized accounts were only for provincial or city public libraries. The purpose of this system is to report the current development of public librarianship and to serve the public library profession. Its clients include library administrators and researchers. Responses rates varied from year to year, a nd usually lower than PLSR.

Features of these four systems can be seen from this table:

Table not available, please contact Author.

System characteristics, Commonalities and Discrepancies

Built on different social cultural backgrounds, the statistical systems contain their own unique features. In some aspects, such as central organization and relationship between central organization and local participation, variation of national systems may make comparison difficult. However, on the other hand, some characteristics, such as system functions, data manipulation, standardization, a nd publications, have international concerns and commonalities that can be analyzed for the comparison.

System functions

Both of these systems include functions that address the following needs or issues:

  1. Accountability. By tracking movement over time and historical trends/accounts, the intention is that the development and evolution of libraries nationwide can be traced through a national system.

  2. Feasibility and measurability. Data elements can be established and measured to reflect library key conditions and services that were particularly considered when the US systems were founded (Zweizig, 1979). Data established should have generizability and also serve accountability.

  3. Comparability and uniformity. Data can be used within and between libraries for library quality control and service improvement, and commonly to exchange information. "As states achieve greater uniformity of data, there will be greater comparability of data" (Adams, 1990).

  4. Reliability and validity. Accuracy of data reported is emphasized.

  5. Accessibility. Data are open to public. NCES data were available through Web and Gopher starting last year.

Although the aims of all these systems are to support administration and decision making, their individual objectives are not really quite the same. In China the government system tends to use statistical data for making budget guidelines or for library information exchange. The U.S. professional system on other hand intends to use these data for library professionals to make observations on the effectiveness of library services, to demonstrate accountibility, to identify staff development needs, and to improve service design. NCES conducts its surveys in fulfillment of its legislative mission "to collect, and analyze, and disseminate statistics and other data related to education in the United States and in other nations" (USCA, 1988). It intends to primarily serve federal government e xecutive and legislative needs.

Data manipulation

The differences between the U.S. and Chinese systems on data manipulation illustrate the systems' intention and expertise, and are often seen within many aspects such as elements structure, survey instrument, data treatment and statistical techniques used. Basically, contrasted to Chinese systems, information selected in the U.S. systems includes not only a public library and its service, but a lso its communities served. Data in the U.S. systems are treated analytically, not only as input. Relations between library resources and service (what a library has and what is used) are reported by percentage distributions, and relations between library output and communities served are calculated as per capita figures. Statistical analysis is these systems' important feature.

The data element as a crucial part of the statistical system embodies and condenses a system's platform and guidelines. NCES includes a total of 44 items collected from each library 33 basic data elements including population of legal services area, service outlets, staffing, collection, circulation, visits and facilities; and 11 identifying items including name, address, interlibrary relatio nship and administrative structure. There are also an additional 12 items collected on each public library outlet and state library outlet. PLDS has more than 50 data elements that are arranged into five categories: Library Identification, Financial Information, Library Resources and Community Measure, Annual Use Figures, and Output Measures and Role Selection. Additionally, a sixth category cha nges from year to year (e.g., 1995 was Technology in Public Libraries, and 1994 was Children's Services Survey). Each category is formed by three parts: instruction, data elements, and summary table and comparison charts. Formulas used to derive numbers, notations used in summary tables, and the questionnaire used to collect these data are listed as appendixes. Software used for data manipulation in both U.S. systems include DECPLUS (Data Entry, Conversion, Table, Output Program), LOTUS, dBASE, and Fourth Dimensions.

Data elements used in the Chinese systems, as in the U.S. systems, cover basic resources (funding, building, and holdings) and key services of each public library, but unlike the U.S. systems, do not cover population of the community served. The 48 data elements grouped within 18 categories in PLSR, include library name, number of staff, annual circulation, number of seats, and great detail on c ollections and expenditures. With almost the same categories of data as PLSR, OCCPL has only 11 data elements in its 1989 publication. Methods of statistical analysis and techniques for comparing detailed percentages have not been used as much as the U.S. systems. Some data elements such as size of library facilities are useful and valuable to show how resources are allocated. To communicate the nature of service provided in China, the number of seats and size of reading rooms can be major indicators for these closed stack libraries. Opposed to this, the U.S. systems don't report these items because their stacks are open and seats do not appropriately express the use of libraries.


A legitimate usability and comparison of data are based on quality of standards used for data collection. To keep data consistent and reduce error and confusion, U.S. systems use standardized definition to define each itemized element. During the preparation for developing a comprehensive national system, FSCS conducted a pilot project in 1985 to standardize the collection of public library data . The definitions used in this system are from the National Information Standards Organization (NISO) manual Z39.7 (NCES 1995 and 1991). PLDS was designed in conjunction with the Public Library Development Program by the New Standards Task Force of PLA (PLDS, 1988). Terms used in its questionnaire were carefully defined. As a check list of the standardized definition, Entity Data Element Definiti ons are published with data in both systems' publications. Code systems accompanying each data element ensure data quantification.

Use of Statistics for Measuring Performance

Library statistics generated are expected to describe and compare effectiveness and efficiency of libraries. Even though endeavors of library performance measures in both the U.S. and China can be traced back to 1960s (Zhu, 1994; Kaplan, 1964), the natural use of library statistics to support improvement and facilitation of management in public libraries was really recognized only in the past de cade.

In particular, PLDS was initiated to play a role of measuring library output nationwide with uniform measurements. Five major measures adopted in PLDS include circulation per capita, registration as a percent of the population, library holdings per capita, collection turnover, and reference transactions per capita (Johnson, 1992). These had been identified as useful in promoting efficiency, cost effectiveness and increased comparability in public libraries by a number of contributors, such as DeProspo et al. (1973), Zweizig and Rodger (1982) and Van House et al. (1987). Systems designers who worked for both NCES and PLDS developed the national systems with the uniform measurements in order to make them a powerful tool for performance measures in public libraries nationwide.

The measurements reported in NCES also concentrate on uses and the services a library provides for its community. Its focus is "what a public library delivers to the community it serves" (Lynch, 1990). Example measures are percentages of registration of population and per registered borrower, in library use, rates of reference completion, and title filled. Facilities or non service related infor mation such as library funding and staff salaries is measured, but the emphasis is on effectiveness and efficiency of uses and the services a library provides to its community. Instances of these measurements include per capita holdings, per capital operating expenditures, and per capital library operating income.

Focuses of measures in Chinese systems are based on their management intention, which includes not only how well a library is used, but also how well a library is developed. PLSR was designed to determine investment of government resources in library development. Measurements are basically functioning the same as data elements. The main measures, such as number of seats and volumes, size of libr ary stacks and reading rooms, and annual book purchasing budget are examined and used to compare quantity and quality of library development. OCCPL was originated to play the role of measuring effective use of a library. But measuring each library's resource in order to study the use of library funds also plays an important role. The different focuses between the two countries' systems reflect so cial cultural backgrounds and level of library developments.

During 1992 to 1994, a national assessment of public library work was made in China. To give a series of national criteria to the appraisal, the Library Bureau of Ministry of Culture issued The Appraisal Indicator Series for Provincial Libraries (AISPL), and Appraisal Standards for City/County Libraries (ASCCL). The purpose of the national evaluation is to promote level of professional work and quality of library service, and to accelerate development of public librarianship nationwide. AISPL comprises six categories and seventy two indicators 55 quantitative indicators and 17 qualitative indicators. These indicators comprise aspects from library service, to professional work, to library facilities, and are intended to be employed as standardized performance criteria used for public libraries with different roles, such as provincial, city, county and children's libraries. A main feature of the assessment is the use of quantified measurements to promote library development by allowing comparability cross libraries on a number of dimensions. Over 2000 public libraries nationwide participated this activity. AS a result, 1144 libraries received awards and were designated as adv anced units at one of three levels. AISPL and ASCCL together with OCCPL comprise a complete professional system of performance measures for public libraries of all sizes and functions.

Because of the great differences between the data elements collected in the U.S. and in China, and between the services designs of each country, possibilities for comparing statistics indicating library operations between the countries' public libraries were extremely limited. Comparative totals or averages by per head between the two countries would be easy to see if the key source, community served, was originally included in China’s public library statistics. Interesting results that could indicate some features of Chinese public libraries compared to those in the U.S.A. at this writing may be seen from the following table:

Pubic libraries in the U.S. and China by population.

Table not available, please contact Author.

* Chinese data are from UNESCO Statistical yearbook, 1993 reported by China PLSR, and U.S. data are from NCES Public Libraries in the United States: 1993.

* In 1993, the population in China was 1,196,360,000; and in U.S. 261,052,000.

A very short and visible comparison came out from the above data on the two country library operations is that U.S. public libraries had more circulation with less staff number and smaller collection size per library compared to Chinese libraries. This reflects, in part, the emphasis on in-library use in Chinese libraries.

Table not available, please contact Author.

  1. 1. Averages of library staff in each library are 17.2 in China and 12.54 in the U.S.
  2. 2. Averages of collection size in each library are 121,790 in China and 76,870 in the U.S.
  3. 3. Averages of loans to users (circulation) in each library are 45,310 in China and 177,620 in the U.S. annually.

Conclusion: Issues and Trends

Four factors appear to influence the system characteristics in these two countries:

  1. System's background. Genesis of a system affects its later development. The social and cultural environment influence the system's evolution;

  2. Funding sources. The organizing and executive agencies determine the system's properties;

  3. Program goals. Government statistics serve government needs. Professional systems center on the use of measures for library professionals;

  4. Data manipulation. Government systems have a higher response rate therefore a greater authority of statistics. Data are collected systematically moving from local to central. Professional systems have to conduct their national surveys for data collection directly from each outlet. Great focus on questions of professional interest and emphasis on measures of performance are beneficial to p ublic library colleagues.

    As a trend, the parallel systems in both countries, reflecting modern needs for statistics support of library performance measures, will continue to serve the management needs.

Efforts made by both governmental officers and library professionals in library statistics have been rewarded with the development of comprehensive statistical systems in both countries, in which statistics can be readily used for management to examine and appraise library characteristics. The system does not have the power to promote effectiveness and efficiency of library service on its own, b ut it does have the power to illuminate the policy of the service and to support the development of librarianship by means of statistics if they are appropriately used. The statistical systems used to examine library progress have evolved over time as a result of a variety of impulses and not as a result of a coherent strategy or plan.

Strengths of the US systems include sophisticated data manipulation, extensive standardization, and strong statistical analysis. Important library characteristics are represented well through statistical elements. Data reported are made easy for public to access. A weakness in NCES is the delay of at least two years in reporting data after collection. On the other hand, higher response rate in P LDS is only from libraries that serve large communities. Both Chinese systems have a lack of statistical analysis and information about the community served. The information shown in these systems in both countries has not been effectively used in research on library performance.

Further investigation. Social economic background is different for the two countries, but the needs for statistical information appears to be the same. Most functions of these statistical systems and their problems and issues faced are also not different. Questions that a system administrator may ask, including what data should be included and relation between cost and effectiveness of data coll ection, still remain. Understanding these systems' development and function is hampered by their being considered "inevitable" and "natural" when in fact they have developed to serve unstated but definable purposes. Study of two comparable systems from strongly contrasting cultures can reveal obscure characteristics and functions. Similarities and differences regarding the above categories can il luminate the inherent characteristics of the national systems, and will be worthwhile for understanding of the role and impact of nationwide library planning, evaluating, and service performance measures.

Appendix: Publications of the National Public Library Statistics in USA and China


Public Libraries in Fifty States and the District of Columbia: 1989. (1991). Washington, DC.: U.S. Dept. of Education, Office of Educational Research and Improvement.

Public Libraries in the United States: 1990 1993. National Center for Education Statistics (NCES). Washington, DC: U.S. Dept. of Education, Office of Educational Research and Improvement.

Report on Evaluation of Definitions Used in the Public Library Statistics Program. (1995). Prepared for the National Center for Educational Statistics by the Government Division, Bureau of the Census.


Public library data service statistical report: 1988 1991. Prepared by the Public Library Association, a division of the American Library Association. Chicago: The Association.

Statistical report ... Public Library Data Service: 1992 1995. Prepared by the Public Library Association, A Division of the American Library Association. Chicago: The Association.


The Annual Report of Public Library Statistical Records: 1950 1995. Ministry of Culture of China. Beijing: The Ministry Publishing House.


Overall Conditions of China Public Libraries: 1959, 1965, 1982, 1989. The Library Bureau and The National Library of China. Beijing: Academic and Periodical Publisher.


The author is grateful to Professor Douglas Zweizig for his invaluable assistance in writing this article.


Adams, S.E. (1990, January). Public library statistics: past, present, and future. Illinois Libraries, 74 76.

DeProspo, E., Altman E. & Beasley K. (1973). measures for public libraries. Chicago: American Library Association.

Johnson, D.W. (1993, September/October). Reflecting on the public library data services project: public libraries over years, 1987 91. Public Libraries, 259 269.

Li, Z. (1988). Librarianship in China: 1949 1986. Beijing: Bibliographic Literature Publisher. [in Chinese].

Liu, D. (1995). Take advance of the assessment and make a new step. Journal of the Library Science in China 2, 3 9. [in Chinese].10(10)

Lynch, J.M. (1989, Summer). Public library statistics: things are definitely looking up. The Bookmark (Albany, N.Y.) 47, 234 7.

Lynch, J.M. (1990). Measurement of library output: how is it related to research. In C.C. Curran & F.W. Summers (Eds), Library performance, accountability, and responsiveness: essays in honor of Ernest R. DeProspo (pp.1 8). Norwood: Ablex.

Thi, K.W. (1988). Four decades of international library statistics. IFLA Journal 14(2).

USCA (1988) United States Code Annotated: Section 406(B) of the General Education Provision Act, as amended 20 U.S.C. 1221e l. St. Paul: Minn. West Publishing Co.

Van House, A., Lynch, M., Zweizig, D. & Rodger, E. (1987). Output measures for public libraries, a manual of standardized procedures (2nd ed.). Chicago: American Library Association.

Zhan, C. (1995). A line of thought regarding the appraisal work for promoting the librarianship. Journal of the Library Science in China /5, 50 53. [in Chinese].

Zweizig, D. (1979). Approaches to the study of public library services and users: the recommendations of a study panel to the National Center for Education Statistics. Public Library Quarterly 1(3), 258 266.