Committee on Privacy and Confidentiality
 Information about Statistical Disclosure Limitation Methods

What is meant by "statistical disclosure limitation methods"?

 A brief description of the issue.

 Federal agencies are often required by law or policy to protect the confidentiality of the information that they collect from persons, businesses, or other units. Before releasing statistical tables or microdata files, these agencies use a variety of statistical methods to protect their data and to ensure that the risk of disclosure is very small. In the past, such methods have been called disclosure-avoidance techniques; currently they are called statistical disclosure limitation methods or statistical disclosure control methods. The name change reflects a realization that a zero-risk of disclosure is an impossibly high standard and that the collection of ANY data entails some sort of risk, no matter how small.

 In essence, statistical agencies protect the confidentiality of the data that they collect by restricting the amount of information in tabular data products or microdata that they release (these are called "restricted data products"). Once released to the public, the agencies do not impose conditions on what can be done with the data. Descriptions of statistical methods to limit disclosure can be found in the references cited in the following section entitled How-to Manuals . Bibliographies and other useful references are also listed below. Additional information can be obtained from Statistics Netherlands' Statistical Disclosure Control Project .

 The other main option that statistical agencies have to protect confidentiality is to impose conditions on who can access data, the purposes for which the data can be used, where the data can be used, etc. Such conditions are called "restricted access procedures" Procedures for restricted data access can be viewed as complimentary to statistical methods to limit disclosure

How-to Manuals:

 To learn how to apply these statistical methods, the following sources are recommended:

NTIS Document Sales
 5285 Port Royal Road
 Springfield, VA 22616
 USA
 TEL: 703-486-4650

Note: Chapter 2, entitled "A Primer," is a useful place to start for those who need an introduction to the basics.

Willenborg, L. & de Waal, T. Statistical Disclosure Control in Practice . (Lecture Notes in Statistics # 111). NY: Springer, 1996.

Note: A brief abstract is available from the publisher.

Selected References

BIBLIOGRAPHIES:

To assist researchers and others who want to learn more about this area, the Committee scanned four existing bibliographies, two from the U.S. and two from the Netherlands. Eventually, on-line searches of these bibliographies will be possible.

Federal Committee on Statistical Methodology : The Federal Committee on Statistical Methodology (FCSM) has issued two report on this topic, each of which have excellent bibliographies:

  • Statistical Policy Working Paper # 2, Report on Statistical Disclosure and Disclosure-Avoidance Techniques , contains a bibliography of publications written in 1978 and earlier.
  • Statistical Policy Working Paper # 22, Report on Statistical Disclosure Limitation Methodology , has an annotated bibliography that updates the references in the earlier FCSM report.
  • Statistics Netherlands, Department of Statistical Methods : The Department has an active research program on methods to limit disclosure of statistical data (called "statistical disclosure control"). Relevant references can be found in these bibliographies:
  • Departmental papers and publications from 1986 to December 1994 are listed in a bibliography .
  • Papers and publications written between January to December 1995 are contained in an annotated bibliography .

Other Useful Publications:

  • Duncan, G.T., Jabine, T.B., & de Wolf, V.A. Private lives and public policies: Confidentiality and accessibility of government statistics . Washington, DC: National Academy Press, 1993.

Abstract: This book is the result of the deliberations of the Panel on Confidentiality and Data Access which was convened by the National Academy of Sciences-National Research Council's Committee on National Statistics and the Social Science Research Council. The Panel was charged with developing recommendations that could aid federal statistical agencies in their stewardship of data for policy decisions and research. Three areas were of paramount concern: protecting the interests of data subjects through procedures that ensure privacy and confidentiality, enhancing public confidence in the integrity of statistical and research data, and facilitating the responsible dissemination of data to users. The Table of Contents of the book follows:

    • Principles and problems
    • The framework of the study
    • Data subjects
    • Data users
    • Legislation
    • Technical and administrative procedures
    • Statistical data on organizations
    • Managing confidentiality and data access functions Chapter 6 contains the Panel's findings and recommendations on statistical techniques and on administrative procedures to protect confidentiality. These administrative procedures, called "restricted access" , impose conditions on who may have access, for what purposes, for what length of time, etc.

To order this publication visit the National Academy of Sciences website or contact:

National Academy Press
 BOX 285
 2101 Constitution Avenue, N.W.
 Washington, D.C. 20555
 TEL: 800-624-6242 or 202-334-3313
 

  •  EUROSTAT : Proceedings from its biennial seminar series on confidentiality:
  • 1992 seminar held in Dublin : International Seminar on Statistical Confidentiality: Proceedings of 8-10 September 1992. Bruxelles, Luxembourg: EUROSTAT, Statistical Office of the European Communities, 1993

Abstract: This statistical document comprises 39 papers presented at the international seminar on statistical confidentiality from 8 to 10 September 1992 in Dublin, Ireland. The seminar was jointly organized by Eurostat, the Statistical Office of the European Communities, and the International Statistical Institute, in collaboration with EU and EFTA member countries, Statistics Canada and the US Bureau of the Census. This seminar was organized in order to share information amongst producers, providers and users of data on the methods and systems used to ensure statistical confidentiality. (Author's abstract)

To purchase this publication contact Eurostat or in the US call this toll-free number:

Unipub
 4611-F Assembly Drive
 Lanham, MD 20706
 TEL: 800-274-4888
 FAX: 800-865-3450

1994 Seminar held in Luxembourg

  • 1996 Seminar held in Bled
  • Netherlands Society for Statistics and Operations Research , Statistica Neerlandica , 1992, vol. 46, no. 1

Abstract: This special volume contains papers given at a symposium in honour of Professor Robert J. Mokken on the occasion of his retirement as Director of Methods and Development of the Netherlands Central Bureau of Statistics (CBS). This conference, "International Symposium on Statistical Disclosure Avoidance," was held at CBS in Voorburg on December 13, 1990.

 The symposium was chaired by Tore Dalenius. In addition to Dr. Dalenius's opening remarks, the volume contains five papers:

    • Keller, W.J. & Bethelehem, J.A. "Disclosure protection of microdata: Problems and solutions," pp. 5-19.
    • Skinner, C.J. "On identification disclosure and prediction disclosure for microdata," pp. 21-32.
    • Greenberg, B.V. & Zayatz, L.V. "Strategies for measuring risk in public use microdata files," pp. 33-48.
    • Mokken, R.J., Kooiman, R., Pannekoek, J., & Willenborg, L.C.R.J. "Disclosure risk for microdata," pp. 49-67.
    • Blien, U., Wirth, H., & Muler, M. "Disclosure risk for microdata stemming from official statistics," pp.69-82.

To order this publication contact the Netherlands Society for Statistics and Operations Research (VVS) at the following address:

Secretariaat VVS
 P.O. Box 282
 1850 AG Heiloo
 THE NETHERLANDS

Statistics Canada : Symposium 95: From Data to Information Methods and Systems . Ottawa, Canada: Statistics Canada, October 1996. (Catalogue No. 11-522-XPE)

Abstract: Symposium 95 was the twelfth in the series of international symposia on methodological issues sponsored by Statistics Canada. Each year the symposium focuses on a particular theme and in 1995 the emphasis was on the analysis and dissemination.

 This document contains over 30 papers. Of the nine sessions, two are of particular import to those interested in the privacy and confidentiality of statistical data: Session 3, "Access and Control of Data," and Session 5, "Technical Aspects of Confidentiality." The papers in these sessions are:

    • Brown, D.C.G. "Informational privacy and data protection," pp. 73-75.
    • Bouchard, G. "Population registers and protection of privacy: The experience of the BALSAC register since 1972," pp. 77-88.
    • Desramaux, L. "Legal/policy aspects of confidentiality and privacy," pp. 89-93.
    • de Waal, A.G. & Willenbourg, L.C.R.J. "Global recoding and local suppression in microdata," pp. 121-132.
    • Evans, B.T. & Zayatz, L. "Using noise for disclosure limitation of tabular establishment datam" pp. 133-141.
    • Boudreau, J.-R. "Assessment and reduction of disclosure risk in microdata files containing discrete data," pp. 143-153.

To order this publication visit Statistics Canada's website or send correspondence to:

Statistics Canada
 Business Survey Methods Division
 R.H. Coats Building, 11th Floor
 Tunney's Pasture
 Ottawa, Ontario
 KIA 0T6
 CANADA

For more information about this volume or others in the series, please contact:

John Kovar
 TEL: 613-951-8615
 FAX: 613-951-1462
 EMAIL: kovar@statcan.ca

Statistics Sweden . Journal of Official Statistics , 1993, vol. 9, no. 2.

Abstract: This is a special issue on confidentiality and data access. It is the result of a joint effort among three groups: the Panel on Confidentiality and Data Access of the National Academy of Sciences-National Research Council's Committee on National Statistics, the Social Science Research Council, and the Journal of Official Statistics .

 The Panel on Confidentiality and Data Access sponsored a "Conference on Disclosure Limitation Approaches and Data Access" in March 1991. The purpose of this conference was to examine the dual trends of increasing nonresponse to governmental surveys and the increasing concerns about confidentiality expressed by U.S. citizens. The papers presented at this conference provided the basis for the majority of articles published in this special volume of the Journal of Official Statistics . This issue contains three papers on statistical disclosure methods as well as insightful comments by three discussants:

    • Fuller, W.A. "Masking procedures for microdata disclosure limitation," pp. 383-406.
    • Little, R.J.A. "Statistical analysis of masked data," pp. 407-426.
    • Jabine, T.B. "Statistical disclosure limitation practices of United States statistical agencies," pp. 427-454.
    • Discussion: Greenberg, B.V., Rubin, D.B., & Willenborg, L. pp. 455-474. In addition, it includes a paper by S. Keller-McNulty and R. Unger, "Database systems: Inferential security," (pp. 475-499) which addresses the problems of data security and confidentiality from both the statistical and computer science points-of-view. This article synthesizes the research done in both disciplines and provides an extensive review of the literature

For information about the Journal of Official Statistics , please contact:

Journal of Official Statistics
 Statistics Sweden
 Box 24300 S - 104 51 Stockholm
 SWEDEN
 TEL (Chief Editor): + 46 8 783 4300
 FAX: + 46 8 783 4288

Statistics Sweden . Statistical Review , 1983, Third series, vol. 31.

bstract: This is a special issue in honour of Profession Tore E. Dalenius. It contains a list of Dr. Dalenius's publications up to that date, many of which concerned privacy and confidentiality of surveys. It should be noted that Dr. Dalenius made an important contribution to the 1978 FCSM pubication, Report on Statistical Disclosure and Disclosure-Avoidance Techniques , Statistical Policy Working Paper # 2, which is available on-line .

 Of  particular note are the following papers:

    • Frank, O. "Statistical disclosure control," pp. 173-179.
    • Cox, L.H. "Some mathematical problems arising from confidentiality concerns," pp. 179-191.
    • Lichtig, L.K. "Confidentiality issues in health statistics: An American perspective," pp. 191-202.
    • Duncan, J.W. "Confidentiality in a decentralized statistical system," pp. 203-206.