Joint Newsletter of the Government Statistics Section and the Social Statistics Section of the American Statistical Association

Summer 1996

Government Statistics Section Chair's Report

Gerald Gates, Bureau of the Census

It is with sadness that I prepare my first report as 1996 chair of GSS. The Section has lost a cherished colleague and friend, Maria Gonzalez, who passed away on February 19. Maria was a long-time member of ASA and GSS and was very active in this section's activities. As the 1994 GSS Chair, Maria provided leadership to the section and to me personally with her counsel and encouragement while serving as the program chair. I know I speak for many others in recognizing this wonderful person for all she has done for her agency, for Federal statistics, and for her community. The GSS is pleased to cosponsor a session in Maria's memory at the Annual Meetings in Chicago.

The GSS has two new faces this year on its executive committee. Linda Gage of the California State Finance Department is the 1997 Chair Elect, and Michael Cohen of the University of Maryland is the 1997 Program Chair Elect. Certainly not strangers to GSS, Linda and Michael have our support and encouragement as they continue to provide our members with a forum to address the needs of government statistics.

The GSS Executive Committee last met on March 7, 1996. The board has scheduled the 1996 GSS business meeting to take place in Chicago on Monday, August 5, beginning at 6:00 p.m. Check your program for the room location. The executive board will meet immediately before the business meeting. After the business meeting, you are encouraged to form dinner groups to continue discussions, form new acquaintances, or just have a good time. The business meeting will include a discussion of the GSS relationship to the Social Statistics Section. As evident by the title of this news-letter, GSS and SSS are already closely aligned sections and share many of the same objectives and concerns. We even have joint membership opportunities. Our discussion will focus on the desirability of further consolidation of these sections.

GSS is rapidly moving into the electronic age with the establishment of a Home Page on the World Wide Web. The ASA Council of Sections Governing Board has required that sections that have WEB pages provide, at a minimum, their mission statement, member services, and a list of officers. We have all these and more. Our WEB site is expanding rapidly and now includes links to many sites of interest to government statisticians, electronic versions of our newsletter, and the program for our annual meetings with links to papers we have received in electronic form. If you have suggestions for additional information you would like to see, please visit our site at and leave a message for Al Tupek.

In March, the Executive Committee agreed to produce both a paper and electronic version of our proceedings from the 1996 Joint Program. We will use the 1996 Proceedings to assess the feasibility of moving to an electronic-only version to reduce the publications costs to the section. That means you will be able to purchase a printed version as well as access electronic versions of papers through our WEB site. There are many issues yet to be resolved before we become fully electronic, including how to best receive the electronic versions from authors, preserving the quality of the statistical charts and figures, and how to provide printed versions upon request. By the time of our meeting in Chicago we should have printed versions of the section brochure developed by Mary Heim. The brochure is modeled after those used by other sections and will be available at the ASA registration tables for prospective members. The brochure also features the new GSS logo created by Joan Turek. Be sure to check it out.

In closing, let me say how much I appreciate the opportunity to serve the section as it faces the challenges and opportunities of better meeting the diverse needs of its Federal, state, and local government members, as well as non-government members, in the years ahead. If you have any suggestions, compliments, or complaints, please do not hesitate to e-mail me at

Social Statistics Section Chair's Report

Daniel Weinberg, Bureau of the Census

We have made progress on setting up an Internet presence for the Section. Miron Straf (Chair­Elect) will lead the effort; contact him with your ideas ( Margo Anderson (past Program Chair) is also leading an effort to develop continuing education and workshop ideas. She needs volunteers to help, so volunteer (, 414­229­3969)!

The Section is co­sponsoring a symposium May 17th at New York University, RECENT DEVELOPMENTS IN SAMPLING METHODS. For information, contact the Symposium Coordinators, Sean Chen (Assistant Professor of Statistics) or Mark S. Handcock (Associate Professor of Statistics) at New York University: phone: (212) 998­0475, Fax: (212) 995­4003, E­mail:

We have exciting plans for the Joint Statistical Meetings in August in Chicago which are described in this Newsletter by our program chair, Susan Miskura. There will be invited sessions, poster sessions and luncheon roundtables. Ed Goldfield has volunteered to keep an eye out for papers of publishable quality so the authors can be encouraged to submit their work to JASA.I will highlight for you some of the key activities and plans that we discussed. We are making excellent progress in putting together our program for Chicago. Carolyn Shettle and Michael Cohen have outlined the program elsewhere in this newsletter and I think you will see that we have many exciting sessions focused on topics of general interest to our members. Please try to support your section by attending sessions of interest to you. If you find the session topics not of interest to you, let Michael know what subjects you would like to see covered at next year's meetings.

A topic of discussion at the SSS Executive Committee meeting will be investigating opportunities for closer contact and cooperation with the Government Statistics Section (as illustrated by this joint newsletter). Please send your ideas and comments to any member of the committee or to me (, 301­763­8550) on this or any other concerns.

See you in Chicago!


CHICAGO, AUGUST 4 - 8, 1996

SSS Program Highlights in Chicago

Susan Miskura, 1996 Program Chair

With four invited paper sessions, five contributed sessions, nine poster sessions, and seven luncheon roundtables, the Social Statistics Section Program for Chicago will offer a variety of sessions designed to encompass our members' many interests.

Organized by Linda Gage, California State Department of Finance, our invited session, "Statesman, Social Scientist, Statistician: Memorial Session in Honor of Clifford C. Clogg," is co-sponsored by several sections. Kathy Wallman will chair the session, where Cliff will be remembered by Leo Goodman, University of California at Berkeley; Martha F. Riche, Census Bureau; Ken Wachter, University of California at Berkeley; and Don Rubin, Harvard University.

Widespread interest in the status of decennial census planning will be addressed in the invited session, "Census 2000: Re-engineering, Content, and Public Perception," organized by Patricia Becker, APB Associates. Ed Spar, COPAFS, will chair this session, with panelists Robert W. Marx, Census Bureau; Keith Rust, Westat, Inc.; TerriAnn Lowenthal, The Rothleder-Lowenthal Group; and Patricia Becker commenting on the status of Census planning and representing a range of views on design issues. The session will include a dedication by Katherine Wallman to the memory of Maria Elena Gonzalez, among whose many interests and contributions to the profession were several that influenced the decennial census.

Our two other invited sessions will cover several matters related to children and families. Edith McArthur, National Center for Education Statistics, has organized the session "Developments in the Coordination and Measurement of Data on Children and Families." In this session, Jerry West, NCES, will describe "Developments in the Measurement of Early Childhood Educational Experiences"; Mike Pergamit, BLS, will discuss data collection issues for the National Longitudinal Survey of Youth; Matt Stegner and Laura Lippman, NCES, have authored a presentation on government-wide issues of coordination; and Deborah Phillips, the National Research Council, will discuss data collection methodology. We are pleased to have Linda Waite, University of Chicago and PRC, and former president of the Population Association of America, as the discussant for this session.

"Evaluation of the Family and Medical Leave Act," organized by Mary K. Batcher, IRS, and chaired by Donna Lenhoff, Department of Labor, completes our roster of invited paper sessions. The panel for this session includes Kerry Levin, Westat, Inc.; Ellen Bravo, 9 to 5; Kate McGonagle, University of Michigan; and Sherry Dale, National Association of Women Business Owners.

The five contributed paper sessions sponsored by SSS encompass a wide range of technical methodologies, analytical findings, and social issues. It is always a challenge to the program chair to select the right combinations of papers for sessions by topic, methods, and results; so please be sure to check the Joint Meetings Programs to plan the sessions you want to attend and specific papers you want to hear. Our contributed paper session titles include "Survey Design and Measurement for Social Characteristics," "Statistical Techniques in Health and Mortality Studies," "Analytical Modeling for Social Science Applications," "Analytical Assessments of Social Conditions and Issues," and "Improving the Design, Collection, and Analysis of Social Data."

Ed Spar, incoming Social Statistics Section Program Chair, has arranged seven luncheon roundtable sessions that will be well worth attending. Topics include using administrative records as a future data source, led by Patty Becker; experimenting with measuring race and ethnicity, with N. Clyde Tucker, BLS; Privacy and the internet, discussed by Jerry Gates, Census Bureau; the new approach to measuring poverty, with Connie Citro, National Academy of Sciences; using the Survey of Income and Program Participation, with Meyer Zitter, National Academy of Sciences; the Continuous Measurement Program, featuring Jay Waite of the Census Bureau; and the role of the federal statistical system in the era of block grants, led by Ed Spar. We know this wide array of topics and knowledgeable discussion leaders will attract much interest.

Last but not least, don't forget the poster sessions, where applications of a number of modeling and analytical techniques will be available for your information and questions.

Thanks to all of the session organizers, presenters and panelists, chairpersons, section officers and others whose efforts have gone into building the program for Chicago. We hope to see you at the sessions and the section business meeting.

GSS Program Highlights in Chicago

Carolyn F. Shettle, National Science Foundation, 1996 Program Chair, and Michael L. Cohen, Committee on National Statistics, 1997 Program Chair

In spite of the furloughs and snowstorms that placed obstacles in the way of our plans for the 1996 Joint Statistical Meetings (JSM), GSS has a strong program scheduled for Chicago. Thanks to the many GSS members who have volunteered their time and talents to put this together!

Meeting the budget challenges:

Government agencies at all levels are being challenged to improve customer services while cutting costs. Many of our sessions allow participants to share information on how they are meeting this challenge. We have an invited session, organized by Wendy Alvey, IRS, that will directly address "The Effect of Downsizing on Federal Statistics." The session will be chaired by Hermann Habermann, former Chief Statistician of the United States, who is currently at the United Nations. Panelists include Katherine K. Wallman, Chief Statistician of the United States; Ivan Fellegi, Chief Statistician of Canada; Edward Spar, Executive Director of the Council of Professional Associations on Federal Statistics; Jennifer Madans, Acting Deputy Director, National Center for Health Statistics; Fred Knickerbocker, Director's Office, U.S. Bureau of the Census; and Linda Gage, California State Finance Department. With such a diverse and distinguished set of panelists, discussion of this difficult issue will undoubtedly be both lively and informative.

Our invited session on downsizing will be complemented by several other sessions on related topics. Lynda T. Carlson, EIA, has organized and will chair a panel session, "The Pros and Cons of Privatizing Government Statistics," that will examine one of the frequently proposed solutions to "doing more with less." Her panelists include R. Paul Moore, RTI; Brad Edmondson, American Demographics; TerriAnn Lowenthal, a consultant formerly with the House Subcommittee on Census and Population; and David B. McMillen, Senate Subcommittee on Government Information and Regulation.

Denice M. Myers, NASS, has organized a session, "Re­engineering within Government Statistical Agencies," that will look at a second proposed solution to the challenge to improve services while reducing costs. This session includes papers by G. David Williamson and Donald R. Betts, CDC, "The Quantitative Methods Enhancement Program: CDC's Innovative Career Development Opportunity"; Jack G. Gambino, Statistics Canada, "The New Design of the Canadian Labour Force Survey"; Ralph Bradley, BLS, "The Use of Scanner Data to Reduce the Mean Squared Error of the CPI"; and Anne Steuer and Michelle Benson, IRS, "Exploring Current and Potential Markets for Electronic Filing (ELF) Using a Tree­Structured Analysis." Ron Bosecker, NASS, will chair the session and Cynthia Z. F. Clark, Census, will be the discussant.

Gerald W. Gates, Census, has organized a panel session entitled, "Administrative Records Can Play a Greater Role in National Statistics if....," to explore the feasibility of placing greater reliance on administrative records to reduce data collection costs. He will be joined in this discussion by Robert W. Marx, Census; Gordon Priest, Statistics Canada; John Czajka, Mathematica Policy Research; and John Long, Census. This session will be complemented by a paper session on "Using Administrative Records for Generating Government Statistics," organized by Kent Marquis, Census. Kent will present a paper entitled "Towards an Error Profile for Estimates from Administrative Records." The other papers in this session are: Paul M. Siegel and John Coder, Census, "Use of Administrative Records to Produce Postcensal (1993) Income and Poverty Estimates for All U.S. Counties"; Fritz Scheuren, GWU, and William E. Winkler, Census, "Recursive Merging and Analysis of Administrative Lists and Data"; and Peter Sailer and Michael Weber, IRS, "Creating Household Data from Individual Income Tax Returns." Daniel Melnick, NSF, will chair the session and Thomas B. Jabine, consultant, will be the discussant.

Two sessions will focus on how best to satisfy our customer needs. "Satisfying the Public's Needs for Statistics," organized by Denice M. Myers, NASS, and chaired by Gypsy Banks, CFSA, USDA, examines efforts at several statistical agencies to improve their customer services. The following papers will be given in this session: M.J.G. Luppes and R.V. Goedegebuure, Statistics Netherlands, "Private Needs and Public Use of Statistical Information"; Mary Ann Guadagno, Andrew A. White, Robert J. Weinzimer, and Elizabeth W. Jackson, NCHS, "Maximizing the Value of Discussion Groups in Customer Satisfaction Surveys"; Carolyn F. Shettle, Alan I. Rapoport, and Alan R. Tupek, NSF, "Performance Measurement and Benchmarking at the National Science Foundation"; and Robert H. Rutchik, EIA, "Maximizing Feedback to Increase Customer Use on the Energy Information Administration World Wide Web Site." Nancy Kirkendall, EIA, will be the discussant.

Mary Ann Guadagno, NCHS, has organized a related panel session, entitled "Customer Satisfaction Surveys: Looking Towards the Future." The session is chaired by Andrew A. White, NCHS. The panelists are Jerry L. Coffey, OMB; Lynda T. Carlson, EIA; Mary Batcher, IRS; Dick Eisinger, SSA; and Clyde Tucker, BLS.

Improving the quality of government statistics:

The pressures to improve government statistics do not abate during budget crises. Indeed, the need for good quantitative information for setting governmental priorities may well increase when funding is tight. It is, therefore, reassuring that several sessions discuss enhancements to existing government statistics.

Our second invited session will focus on the widely publicized research issues related to improvements in the Consumer Price Index (CPI) and is entitled "Statistics under the Spotlight: Improving the Consumer Price Index." It was organized by Brent R. Moulton, BLS, and will be chaired by Jack E. Triplett, Chief Economist, Bureau of Economic Analysis. Panelists include Katharine Abraham, Commissioner, BLS; William A. Barnett, Washington University; Robert J. Gordon, Northwestern University; Kirk M. Wolter, NORC; and Arnold Zellner, University of Chicago. This session will certainly be a must for all of us interested in understanding the relationship between politics and statistics as well as those interested in the technical aspects of the CPI research.

Olivia Carter­Pokras, HHS, has organized a follow­up session to last year's invited session on "The Categorization of Race and Ethnicity." Pat Golden, NCHS, will chair the panel session. The panelists will be Katherine K. Wallman, OMB; Clyde Tucker, BLS; David Cantor, Westat; and Nampeo McKenney, Census. They will report on their ongoing efforts to review OMB race and ethnic coding and will elicit opinions from JSM participants on what changes, if any, should be made to the current standard codes.

Carolyn F. Shettle, NSF, has organized and will chair a session on "The Effectiveness of Monetary Incentives in Government Surveys." Don A. Dillman, Washington State University, will talk about "Token Financial Incentives and the Reduction of Nonresponse Error in Mail Surveys." Robert Baumgartner, Hagler Bailly Consulting, Inc., will present a paper on "Prepaid Monetary Incentives and Mail Survey Response Rates." Geraldine M. Mooney, Mathematica Policy Research, and Carolyn F. Shettle will prepare a presentation on "The Impact of Incentives in a Government Survey." Robert Groves, Joint Program in Survey Methodology, and Jerry L. Coffey, OMB, will discuss the papers and the question of whether the Federal government should consider permitting token monetary incentives in government surveys.

Thesia Garner and Stephanie Shipp, BLS, have organized a session entitled "How Much Income Do You Need? How Much Do You Need to Spend?" This session will examine some of the recent research related to measuring subjective aspects of economic well­being. Lee Rainwater, Harvard and the Luxembourg Income Study, will chair the session. The papers to be presented are: Denton R. Vaughan, SSA, "Self­assessments of Income Needs and Financial Circumstances: Two Decades of Seeking a Place in Federal Household Surveys"; Robert Kominski and Kathleen Short, Census, "Developing Extended Measures of Well­Being: Minimum Income"; Thesia Garner, Linda Stinson, Brian Kojetin, David Johnson, and Stephanie Shipp, BLS, "Subjective Assessments of Economic Well­Being: Understanding the Concepts and the Language." Norman Bradburn, NORC, and Lee Rainwater will discuss the findings.

Improving confidentiality protection and dissemination:

One of the basic tenets of the customer service movement within the government statistics community is that producing high quality statistics is a necessary, but not sufficient, condition for serving our customers well. Another key ingredient is ensuring that data are easily accessible to policymakers and other customers. Two sessions focus on this aspect of our work.

Denice M. Myers, NASS, has organized a session looking at issues related to the "Electronic Dissemination of Statistical Information." The session will be chaired by Jimmie Givens, NCHS. Several agency representatives who participated in the recent Federal Committee on Statistical Methodology subcommittee on this topic will present information about electronic dissemination in their agencies. Robert J. Weinzimer, NCHS, will talk about "Electronic Dissemination of Statistical Information from the National Center for Health Statistics." Jewel T. Barr, NASS, will share information on the "Electronic Dissemination of Time­Sensitive Agricultural Data and Statistics at the National Agricultural Statistics Service." Oya Y. Rieger, Cornell, will relay information about the experiences of the "USDA Economics and Statistics System." In addition, Robert Hinton, Indiana University, will provide a librarian's perspective on this issue in a talk entitled "Electronic Dissemination of Government Information: GPO and the Federal Depository Library Program in Transition." Robert Zarnetske, BTS, will be the discussant.

Virginia de Wolf, BLS, has organized a session that will focus on overcoming one of the major impediments to making data more easily available to our users ­­ the need to protect the confidentiality of our respondents. She will chair this session entitled, "Avoiding Statistical Disclosure," that will examine new statistical approaches to increasing the amount of data available to respondents while ensuring confidentiality. Gordon Sande, Sande and Associates, will provide "Comments on Methods and Comparisons of Software for Cell Suppression." Michael J. Griffey and Jacob Bournazian, EIA, will co­author a paper for the session on "Disclosure Avoidance Techniques Used in Petroleum Marketing Data." Richard Moore, Timothy Evans, and Laura Zayatz, Census, will provide information on "New Directions in Disclosure Limitation at the U.S. Census Bureau." Lawrence H. Cox, EPA, will talk about "Using Mathematical Networks for Disclosure Limitation." George T. Duncan, Carnegie­Mellon, will serve as discussant.

Research related to defense:

Thursday will feature two sessions related to research in the area of defense and national security. The first of these, organized by Donald P. Gaver, Naval Postgraduate School, grows out of work done by the Committee on National Statistics and is entitled, "Statistical Topics in National Security." The papers are: David M. Rocke, UC Davis, and George W. Downs, Princeton, "The Problem of Preference Change in International Security Agreements"; John Rolfe, UCLA, "Statistical Thinking to Improve DOD Testing and Evaluation"; Donald P. Gaver and Patricia A. Jacobs, Naval Postgraduate School, "Testing for and Removing Design Faults: A Missile Example"; Herman Chernoff, Harvard, "Experimental Design in Operational Testing under Limited Experimentation." Michael L. Cohen, Committee on National Statistics, will be the discussant.

The final GSS session of the JSM will be one organized by Wray Smith, Synectics, and is entitled "Policy Analysis with Education and Defense Manpower Survey Data." It will be chaired by John Czajka, Mathematica Policy Research. The papers in this session will be prepared by researcher teams associated with Synectics and the Defense Manpower Data Center: Gary Huang, Stanley Weng, and Fan Zhang, "Effect of High School Programs on Out­Migration of Rural Graduates"; Steven Fink, Sandeep Bhalla, Sameena Salvucci, and Mary Weltin, "Quality of Life among Military Recruiters: What Improvements May be Made"; Stanley Guterman, Rita Bureika, Sameena Salvucci, and Betty Maxfield, "Evaluation of Employment Assistance Programs for Spouses of Junior Enlisted Military Personnel: Focus Groups and Survey Design"; Lisa Hudson, Barbara J. George, John Helmick, and Timothy Elig, "Policy Options for Educating Military Children: Domestic Dependent Schools and Public Systems"; Timothy W. Elig and M. Sue Hay, "Estimating Sexual Harassment in the Military: Getting the Numbers Right."

Luncheon Roundtables:

The GSS has five luncheon roundtables scheduled for Tuesday, August 6. These roundtables offer an opportunity to discuss informally some of the key issues facing governmental statisticians. Two of these sessions deal with themes raised in our invited sessions. Hermann Habermann, United Nations, will lead a discussion entitled "Consolidation of U.S. Statistical Agencies: A Pipedream or An Unfortunate Eventuality." Robert Casady, Bureau of Labor Statistics, will lead a discussion of "Current Statistical Problems with Definitions, Sample Design, and Estimation for the Consumer Price Index."

The remaining three round tables will delve into quality improvement issues that are not treated in the more formal GSS sessions. John F. Long, Census, will lead a discussion of "Approaches to Metropolitan and Non­Metropolitan Areas for 2000." Jennifer Madans, National Center for Health Statistics, will lead a group examining the issue of "Striking a Delicate Balance: Content Issues in the Redesign of the National Health Interview Survey." Finally, Carolyn Hay, Census, will provide a forum for discussion entitled "Preliminary Indications from the 2000 Census Test of the Use of Innovative Short­Form Questionnaires and New Plans for Long­Form Questionnaires."

The GSS program committee is very excited about the roundtables it has organized and urges all members to make luncheon reservations early to ensure a seat.

Other Events:

GSS is also sponsoring three poster presentations. Sameena Salvucci, Tai A. Phan, and Michael Chang, Synectics, will have a poster on the "School District Data Book." Genevieve Strahan, NCHS, will have one on "Collecting Expense Data in the 1995 National Nursing Home Survey." John P. Timm, Census, will have a poster on "Automating the Private School Survey Universe Update Operation."

Also, please do not forget to review the May issue of AmStat News for the offerings of other sections that are co­sponsored by GSS. Finally, be sure to attend the GSS business meeting on Monday, August 5, at 6:00 p.m. After the meeting attendees will be provided with an opportunity to form informal groups to dine at nearby restaurants.

See you in Chicago!

Here We Go Again ­ It's Budget Season

Edward J. Spar, Executive Director, Council of Professional Associations on Federal Statistics

It's still too early to say how hard a year this will be for the federal statistical agencies. However, if last year is any indication, it will probably be another difficult year (for details of the funding history in fiscal years 1995 through 1997, including a comparison of requested vs. actual allocations for 1996, see the table at the end of this article.) Indeed, the Census Bureau, the Bureau's of Labor and Economic Statistics, the Energy Information Administration, the National Center for Education Statistics, and the Bureau of Justice Statistics are still operating under Continuing Resolutions. And given that this is an election year, it would not be unreasonable to see some of the proposed FY 1997 budgets passed before November. Every agency is fully aware of need on the part of Congress to make cuts across the board, and in general, the funding levels for FY 1997 that have been proposed for the agencies provide moderate increases over the resources appropriated in FY 1996. The largest increase comes for the Census Bureau, which is gearing up for the 2000 Census. Below are further details on FY 1997 programs for each of the 11 major agencies.

Bureau of the Census:

The FY 1997 budget request includes funds to expand and improve service statistics; to remedy shortcomings in principal economic indicators, such as for retail and wholesale trade; and to continue work on restructuring the existing Industrial Classification system and implement the new North American Industry Classification System. The Periodic Censuses and Programs request reflects the transition from planning for the Decennial Census to implementation, including conducting tests and planning for the data collection and processing facets of the Census; planning for the 1997 Economic Censuses; CASIC development; and continued work on developing the American Community Survey, a new approach to collecting decennial census data through continuous measurement.

Bureau of Labor Statistics (BLS):

The BLS request includes funds for the continuation of the six-year process to revise the Consumer Price Index (CPI). The BLS requests funds to conduct major revisions of the CPI approximately every 10 years. These revisions, scheduled after each decennial census, allow the BLS to incorporate changes in the distribution of population into the CPI samples of areas and housing units, in addition to updating the CPI "market basket" of goods and services to reflect changes in consumers' buying habits. The updated market basket would be introduced in January of 1998. The proposed revision also will enable BLS to incorporate improvements in technology and index methodology, and to redesign survey questionnaires and computer systems to make the index more accurate, reliable, and timely. BLS is also requesting funding to provide for the BLS component of the government-wide effort to revise the Standard Industrial Classification.

Bureau of Economic Analysis (BEA):

The funding request includes an increase of $7.2 million to support improvements in the economic accounts. BEA will focus on two initiatives: the Mid­Decade Strategic plan for maintaining and improving GDP and other economic accounts data, and the Information Technology Strategic Plan which brings together BEA's customer service, mid­decade review and computer re­engineering efforts. BEA has begun a critical program to move to an integrated micro­computer network environment in FY 1997 and eliminate its 1970's-vintage mainframe. This environment will increase the accuracy, reliability, and timeliness of BEA's data.

Statistics of Income, Internal Revenue Service:

To accommodate the lower level of funding compared to FY 1995, some of the program changes under consideration include reducing the size of the SOI corporation and individual income tax return samples; cancelling the SOI individual income tax return program; dropping certain tax return schedules data from the individual income tax return program; curtailing annual SOI high­income study analyses; and limiting the scope of the present annual SOI estate tax return studies. Although many in the statistical community have voiced their concern, the Statistics of Income operation has been merged with compliance research.

National Agricultural Statistics Service (NASS):

The FY 1997 budget includes $17.5 million for the Census of Agriculture, which is proposed to be transferred from the Bureau of the Census to NASS. This action will integrate the agricultural census activities with NASS's agricultural survey program. By merging the two programs, major efficiencies will be attained in building a more complete list of farm and ranch operators and reducing the reporting burden on agricultural producers.

The FY 1997 budget also includes funding for Integrated Pest Management (IPM) which will allow NASS to measure progress toward meeting the United States Department of Agriculture goal of achieving IPM adoption on 75 percent of the Nation's crop acreage by the year 2000. Funding is requested for a post­harvest pesticide survey which would measure pesticide and other chemical usage on commodities after they leave the farm. Increased funding is also requested to expand the Pesticide Data Program.

Economic Research Service:

An increase in funding of $1.1 million is being requested for data acquisition and analysis of farm practices linked to environmental conditions and economic performance of farms. Under the proposed program, ERS would collaborate with NASS to design and conduct a survey and carry out economic analysis that relates farm financial conditions with data on adoption of production practices, the cost of alternatives, and the environmental and resource conditions under which they are used. The resulting economic information will support the evaluation of key United States Department of Agriculture conservation and environmental initiatives.

Energy Information Administration:

EIA's proposed direct FY 1997 budget request is $66.1 million, a decrease of $6.2 million from FY 1996. EIA is also requesting $4.8 million on a reimbursable basis with the Department of Energy Office of Energy Efficiency and Renewable Energy. The reimbursement is required to support the core EIA data and modeling activities needed to set energy efficiency program needs and measure program results. This includes efficiency and renewable data collection and analyses, end­use energy consumption surveys, greenhouse gas data collection studies, mid­term energy demand modeling, and integrated end­use energy data compilation. With the reduction of the EIA budget over the past two years, EIA has eliminated or reduced the scope of ten publications; reduced mid­term forecasting and related analysis by 35 percent; eliminated or reduced the frequency of nine data collections; and eliminated computer disaster recovery capabilities.

National Center for Health Statistics:

The proposed increases provide for the implementation of the next National Health and Nutrition Examination Survey, a major effort to obtain objectively measured health data from direct physical examinations, in 1998. This funding level will also provide for continued operation of major surveys and data programs of NCHS, such as the National Vital Statistics System, and the National Health Care Survey. Major initiatives at NCHS include the redesign of the National Interview Survey, including a questionnaire redesign and a move to computerized data collection. NCHS is also taking an increasing leadership role in coordinating and streamlining Federal health statistics through the Health and Human Services Survey Integration Plan. This plan involves using the National Health Interview Survey as the sampling frame for the National Medical Expenditures Panel Survey, conducted by the Agency for Health Care Policy and Research.

National Center for Education Statistics:

For FY 1997, a budget request of $82.8 million is proposed, an increase of $4.0 million over FY 1996. Of the total, $32.8 million is requested to support the National Assessment of Educational Progress (NAEP). The FY 1997 request includes funds for completing analysis of the 1996 assessments in mathematics and science; conducting an arts probe in grade 8; planning for proposed assessments of reading, writing, and civics in 1998; evaluation and validation studies; and the activiteis of the National Assessment Governing Board. The request also includes $50 million required to support the ongoing statistics program of NCES. The program collects and reports data on pre-kindergarten through post-secondary education, including; general statistics about the condition of education and providing data to monitor education reform and improvement.

Bureau of Justice Statistics:

For FY 1997, continuing programs will include annual data on criminal victimization, populations under correctional supervision, federal criminal offenders, federal case processing, and criminal justice expenditures and employment. Periodic data series are undertaken to provide statistical information on felony convictions, state court processing, the composition and characteristics of correctional populations, prosecutor practices and policies, and the administration of law enforcement agencies and correctional facilities.

BJS is requesting funds to assist states in identifying, classifying, collecting, and transmitting data on warrants, arrests, convictions, and protection orders relating to stalking and domestic violence, and making such information available through the national criminal history record network. Funding is also requested to conduct the National Study on Campus Sexual Assault. Data collections include the first­ever analysis of campus police agencies; a survey of college campus administrators; and data from victims on the response of the educational institution and law enforcement authorities.

Bureau of Transportation Statistics:

Key activities in FY 1997 for the Bureau include development of transportation system performance measures, development of tools to integrate Geographic Information Systems and transportation network models for national transportation analysis, and expansion of the Bureau's National Transportation Library accessible via the BTS Internet home page. Other initiatives include initiation of data collection for the 1997 Commodity Flow Survey, release of data products from the 1995 American Travel Survey, initiation of a State and Metropolitan Technical Assistance Program, and modernization of the OAI data processing system.

The Plan for Census 2000

Martha Farnsworth Riche, Director, and Robert W. Marx, Associate Director for Decennial Census, Bureau of the Census

From apportioning the Congress to understanding our Nation's communities, the decennial census is key to our knowledge of who we are. There was widespread dissatisfaction with the 1990 result - it cost too much and accomplished too little. To correct this, the overarching goal of Census 2000 is to be faster, less costly, and more accurate. The following objectives are fundamental to our efforts:

n To make every effort - from simpler, user-friendly forms design to the design of field operations - to count every household.

n To implement an open process that diverse groups and interests can understand and support.

n To eliminate the "differential undercount" of racial and ethnic groups.

n To produce a "one number census" that is right the first time and that unites us as Americans rather than dividing us as litigants.

Strategies: The plan for Census 2000 is built around four core strategies:

n Strategy One: Build Partnership at Every Stage of the Process. The Census Bureau can't do everything alone: We need to find "best in class" partners to help us accomplish our objectives. This means:

Partnership with state, local, and tribal governments and community groups. They know local conditions and circumstances better than the Census Bureau ever will. They can help us correct our maps and address lists, tell us where to put our forms so that people will find them, and alert us to other local problems.

Partnership with the U.S. Postal Service. The Census Bureau spent too much time and money developing address lists in the past. This time, we'll use the Postal Service list and avoid costly duplication of effort.

Partnership through privatization. The Census Bureau can't be "world class" in every stage of a decennial process that it manages only once every ten years. It needs to outsource some aspects of the process, for example, Facilities Management, Advertising and Promotion and Training.

n Strategy Two: Keep It Simple. The simpler and easier the decennial census is, the more accurate and less expensive it will be. For example:

User-friendly forms. More powerful computers allow us to use forms that are easier to read and fill out. Private design firms are working with us to implement designs that help people understand why we ask each question.

Multiple contacts. The direct marketing industry has learned that repeated contacts and reminders pay big dividends. We will use that technique.

More ways to respond. In 1990, you had to find your form in the mail. In Census 2000, the form will find you. We'll put forms at stores and malls, in civic or community centers, in schools, and other places where people go. We'll have them in many languages. And we'll have a well-publicized toll-free telephone number people can call if that's more convenient.

n Strategy Three: Use Technology Intelligently. In 2000, the decennial census will finally enter the Information Age.

Digital "capture" of forms. The 1990 census was microfilmed and keypunched. In Census 2000, we'll scan the forms directly into computers that use "optical character recognition" - software that reads handwriting.

"Matching" software. Sophisticated software will allow us to spot duplications - for example, a husband returns the family form in the mail while a wife fills one out over the phone at work.

"Point and click" data tabulation. Any data seeker will be able to get the information they want instantly from the Census 2000 data set, instead of buying thick books that may or may not have what they're looking for.

n Strategy Four: Use Statistical Methods. Sampling and statistical estimation have been part of decennial census processes for more than 55 years. Most statisticians agree that using widely-accepted statistical methods in Census 2000 will produce a better census at less cost. The new approaches include:

After reaching a target of 90 percent response in each county, we'll draw a sample of 1-in-10 nonrespondents and use their responses as a basis for completing the job. This use of sampling for the final increment of the population will ensure that Census 2000 is built on a solid core of responses and save enough money to allow us to pay for a quality check.

We'll take a second sample of the population - about one-half percent - and work them hard to get accurate results, then use their answers as a basis for checking all results. This quality check procedure will lead to a "one number census" with accuracy that is very high for all States, for all Congressional Districts, for the larger urban areas, and for other populous areas. For less populous geographic areas, there will be less certainty about the totals for any specific block, neighborhood, or village. But overall, Census 2000 will be much more accurate, particularly for traditionally undercounted population groups.

We are on solid Constitutional grounds. Our proposal will withstand all legal challenges. Some have suggested sampling sooner, but we are hesitant to do so because it changes the political and social nature of the census. Guaranteeing that we'll find at least 90 percent of the people in every county preserves the traditional character of the decennial census.

Relative or Absolute­­New Light on the Behavior of Poverty Lines Over Time

Gordon M. Fisher, Department of Health and Human Services

One of the central questions in discussions about poverty lines is whether they should be updated in "absolute" or "relative" fashion. (Once established, an absolute poverty line is updated for price changes only, while a relative poverty line is updated for changes in the median or mean income or consumption of the general population.)

In American debates on this aspect of poverty lines, the tendency has been to put forward theoretical arguments for using one or the other form of updating. However, it is not necessary to confine oneself to theoretical arguments on this issue. One can also consider the empirical historical behavior of poverty lines, particularly before the 1960's. (During that period, unofficial poverty lines were developed and then superseded after some period of time by new unofficial poverty lines.) The American evidence on the historical behavior of poverty lines (evidence that many people are not aware of) has been summarized in a longer paper by the author, together with similar evidence from Britain, Canada, and Australia.

This extensive body of American evidence demonstrates that successive poverty lines developed as absolute poverty lines show a pattern of getting higher in real terms as the real income of the general population rises. (This phenomenon has been termed "the income elasticity of the poverty line.") This evidence includes "expert"­devised minimum budgets prepared over six decades; "subjective" low­income figures in the form of national responses to a Gallup Poll question over four decades; and the recorded common knowledge of experts on poverty lines and family budgets from about 1900 to 1970. The recently released report of the National Research Council's Panel on Poverty and Family Assistance (C.F. Citro and R.T. Michael (editors), Measuring Poverty: A New Approach) referred to this historical evidence half a dozen times.

From roughly 1905 to 1960, American budget experts developed a number of "standard budgets" (item­by­item "market baskets") at different levels of living. In 1966, Oscar Ornati analyzed about 60 standard budgets prepared during the 1905­1960 period, and classified them as being at "minimum subsistence," "minimum adequacy," and "minimum comfort" levels. (His "minimum subsistence" category corresponds to our concept of poverty.) A 1973 study by Robert W. Kilpatrick showed that Ornati's minimum subsistence figures over this 55­year period rose 0.75 percent in real terms for each 1.0 percent increase in the real disposable income per capita of the general population.

An examination of early poverty lines and budgets not considered by Ornati confirms and extends this picture. The poverty/subsistence figures examined (like Ornati's budgets) were all derived as absolute poverty lines. Yet over time, these successive absolute poverty lines rose in real terms as the income of the general population rose. Poverty lines and minimum subsistence budgets before World War I were, in constant dollars, generally between 43 and 54 percent of Mollie Orshansky's poverty threshold for 1963. By 1923, Dorothy Douglas' "minimum of subsistence level" was equal to 53 percent to 68 percent of Orshansky's threshold. A U.S. Works Progress Administration "emergency" budget for the Depression year of 1935 was equal to 65 percent of Orshansky's poverty threshold. Robert Lampman's low­income line for 1957 was equal to 88 percent of Orshansky's poverty threshold.

Evidence from an overlapping but more recent period (extending up to the 1990's) comes from the Gallup Poll. Since 1946, the Gallup Poll has repeatedly asked the American public the following question: "What is the smallest amount of money a family of four (husband, wife, and two children) needs each week to get along in this community?" The average response to this "get­along" question has been higher than the Orshansky poverty line, but it seems reasonable to assume that the relationship between the "get­along" amount and family income is a good indicator of how the public's perception of the poverty line would vary over time in relation to family income (if a "poverty" poll question had been asked). Half a dozen analyses have found that the "get­along" amount rises by between 0.6 and 1.0 percent for every 1.0 percent increase in the income of the general population.

Another significant (although neglected) source of evidence about the income elasticity of the poverty line is the common knowledge of experts on poverty lines and family budgets before 1970, as documented in quotations from their writings. There is one such quotation from 1841, over a dozen from the 1900­1959 period, and over a dozen from the 1960's. It is clear that the income elasticity of the poverty line was well known to these experts, and that they were quite familiar with the underlying social processes involved. One quotation which illustrates these social processes with particular clarity was written in 1938 by Carroll R. Daugherty: "A standard budget worked out in the [1890's], for example, would have no place for electric appliances, automobiles, spinach, radios, and many other things which found a place on the 1938 comfort model. The budget of 1950 will undoubtedly make the present one look as antiquated as the hobble skirt." Some of the quotations make ironic reading in the light of subsequent history, as when the Social Security Administration's Ida Merriam (Mollie Orshansky's boss) wrote in 1967 that "It is easy to observe that poverty in the U.S. today cannot meaningfully be defined in the same way as in the U.S. of 1900....obviously today's [poverty] measure, even if corrected year by year for changes in the price level...should not be acceptable twenty, ten or perhaps even five years hence." Others publicly recognizing the income elasticity of the poverty line during the 1960's included Rose Friedman and the Republican minority of the Joint Economic Committee of Congress.

In the light of this extensive American evidence from standard budgets, the Gallup Poll, and the common knowledge of experts, it becomes clear that the 1968­1969 decision to adjust the official poverty thresholds only for price changes (and not for changes in the general standard of living) represents a single major historical anomaly. The anomaly of the 1968­1969 decision is highlighted by the fact that when the Council of Economic Advisers' $3,000 poverty line and then the Orshansky poverty thresholds had been adopted only five years earlier, they had been known to be 14 to 19 percent higher in real terms than unofficial poverty lines in use only six to fifteen years earlier.

The force of this American historical evidence is strengthened when one realizes that the income elasticity of the poverty line results from social processes that have continued­­indeed, have perhaps even intensified­­since the 1960's. These social processes can be summarized as follows: As technology progresses and the general standard of living rises, new consumption items are introduced. They may at first be purchased and used only by upper­income families; however, they gradually diffuse to middle­ and lower­income levels. Things originally viewed as luxuries­­for instance, indoor plumbing, telephones, and automobiles­­come to be seen as conveniences and then as necessities. In addition, changes in the ways in which society is organized (sometimes in response to new "necessities") may make it more expensive for the poor to accomplish a given goal­­as when widespread car ownership and increasing suburbanization lead to a deterioration in public transportation, and the poor are forced to buy cars or hire taxis in order to get to places where public transit used to take them. Finally, the general upgrading of social standards can make things more expensive for the poor­­as when housing code requirements that all houses have indoor plumbing added to the cost of housing.

In the light of these social processes, the only kind of American society in which it would be sociologically justified to have had the same fixed­constant­dollar poverty line since the mid­1960's would be a society in which there had been essentially no technological change or innovation since 1960. In a society in which there have been extensive technological changes and innovations since 1960, the appropriate course would be to have a poverty line which is higher than the poverty line developed in the 1960's and which does rise in real terms over time in response to increases in the general standard of living­­as recommended by the report of the NRC's Panel on Poverty and Family Assistance.

For additional information on this subject, you may contact Gordon Fisher by telephone (202-690-6143) or e-mail ( The views expressed here are those of the author, and do not represent the position of the Department of Health and Human Services.

The 1996 Panel Survey of Income and Program Participation Redesign

William Lloyd, Bureau of the Census

In April 1996, the Survey of Income and Program Participation (SIPP) completed a transition from a paper survey to a computer-assisted personal interview (CAPI) survey. In addition to this, a number of changes were implemented that make the SIPP/CAPI different than the traditional paper-and-pencil SIPP survey. This paper describes the important aspects of the 1996 Panel implementation of the redesigned SIPP/CAPI survey.

Survey Design:

In the paper SIPP, a new survey panel of 20,000 respondents was introduced in February of each year, and respondents remained in sample for approximately 3 years for 9 interviews. Interviews took place every four months, and persons who moved were followed and interviewed at their new address. However, our data users identified several problematic areas. The sample sizes were too small to analyze certain population subgroups. Further, the panel lengths were too short to observe a sufficient number of completed program spells.

The redesigned SIPP/CAPI introduced a new panel of 36,700 households in April 1996, and will interview them every four months (13 times) for a four-year period. A new SIPP Panel will be introduced in February 2000, and every 4 years thereafter. The combined effect of the larger panel size and the longer panels will help us to observe a much greater number of complete spells of program participation and will allow improved analysis of population subgroups. This shift in survey design primarily reflects a change in priority for the survey. The paper SIPP survey design, with its annual panel introduction, reflected the importance of cross-sectional estimates. The new SIPP/CAPI design takes the same number of annual interviews and puts them all in one larger panel, whose main emphasis is longitudinal information.

There will be no overlapping panels with the 1996 SIPP/CAPI Panel. In the paper SIPP, overlapping panels allowed for linkage of multiple SIPP data sets, and for continuity in the time-series data. It also allowed analysts to examine larger sets of cross-sectional data collected from similar, overlapping panels. We examined several statistical issues that arise from the decision not to use overlapping panels: Will there be an increase in the time-in-sample bias to the data?; Should there be a significant increase in item or person non-response?; and, Will the lack of overlapping panels adversely affect cross-sectional data estimates?

The research we have done indicates that the Census Bureau should be able to compensate for any of these effects, and thus these issues should not adversely affect SIPP data quality. The Census Bureau will closely examine the collected data to see what effects, if any, the change in sample design has on SIPP data quality.

Questionnaire Design:

The SIPP/CAPI questionnaire is administered on a laptop computer, which affords the following benefits:

n Integrated data editing as it is being collected, including range and consistency checks;

n Dependent interviewing in subsequent waves, using information reported previously by the respondent;

n The elimination of several pre-processing clerical and keying operations, as well as the initial pre-edit of the keyed data;

n Greatly reduced time and expense in transmitting completed survey data from the field representatives to the regional offices and headquarters, since the data will be transmitted via high-speed modems over existing telephone lines; and

n Reduced time span from data collection to production of the final data file.

Although the questionnaire content remains essentially the same, the wording and arrangement of questions have changed slightly because of dependent interviewing techniques. Beginning with the Wave 2 interview, interviewers can feed back and verify information reported from the Wave 1 interview, reducing the amount of time needed to collect new information in subsequent waves. As a part of the total redesign process, we performed cognitive interviews with test subjects to review the wording of questions with high non-response rates and questions that were found to be awkward. We tested the new wording for specific questions through several field tests that allowed us to examine how respondents answered the newly-revised questions. Listed below are some of the significant features found in the SIPP/CAPI questionnaire:

Special Features of the CAPI Questionnaire:

In the CAPI questionnaire, a number of features have been built into the instrument to make data collection easier. Field Representatives (FRs) can back up 1 screen at a time, or jump back to any previous point in the interview to make corrections in the collected data. Then, once the correction has been made, the FR can press a key to jump forward to the next unanswered question. The FR can enter explanatory notes on any question screen in the instrument. The SIPP/CAPI instrument allows the FR to skip any eligible person's interview, and then return to that person's interview later in the session. A Navigation Grid is available to show the FR how much of the interview has been completed, and what remains to be finished. Individual question screens can be designed to display a roster of specific sub-groups of persons in the household, showing who in the household a question may pertain to. Help screens, containing explanatory information and definitions, are available in virtually every question screen. The SIPP/CAPI instrument can "spawn" new cases if extra units are found at the sample household in Wave 1, or if persons or households move from the original address to a new location in later waves. In the SIPP, movers are followed to ensure that we collect a longitudinal set of household data for the entire survey period. Finally, information reported in previous waves can be fed back to the instrument in subsequent waves, thus reducing the amount of time needed to collect new or revised information.

Topical Modules:

In addition to the basic set of "core" items which are asked in each interview, we also ask additional questions on various topics which relate to a person's income-earning potential. These "topical modules" feature questions on issues such as: migration history, employment history, education and work training history, marital history, and fertility history. We will ask an "assets and liabilities" topical module that gathers information on personal and household net worth. Another module will gather annual tax data, and still another will look at extended measures of a person's well-being. These topical modules allow us to examine the conditions under which a person's economic situation has been built, and how it is likely to change over time. The SIPP/CAPI instrument will use information reported earlier in the interview to determine what questions, if any, should be asked of each household member.

Data Products and Reports:

The Census Bureau will release a series of wave-based cross­sectional, topical module, and multiple-wave longitudinal reports from the 1996 Panel. We will also regularly release public use files containing the core data on income recipiency and program participation and topical module data, where applicable. Wave files will be released that contain both core and topical module data. Once all wave files are released, we prepare a longitudinal data file for each panel. We will also release longitudinal files for calendar years as well.

Wave-based cross-sectional data would be used to create a basic set of statistics from the SIPP core. These statistics would be used to evaluate the survey, to profile the participants in various programs, to examine the characteristics of the population in need, and to examine how the economy is changing. Such statistics would include average monthly estimates of: Median household income, Number of workers and their median earnings, Number of persons in poverty, Persons with labor force activities, Persons who spent time looking for work or on layoff, Participants in government programs such as AFDC and other public assistance, Social Security and SSI, school lunches, food stamps, Medicare and Medicaid, public or subsidized housing, and unemployment and workers' compensation. These statistics (crossed by race, age, and other characteristics) from each SIPP wave would be the basis for reports (or fact sheets) on the economic situation of Americans and their families.

In terms of creating multiple-wave longitudinal data products, the redesigned SIPP is quite different from the original SIPP in that, as the panel grows in length, our longitudinal reports (and data files) will be more interesting and useful. As the length of our observations grows, we will be better able to examine issues such as:

n Annual income and poverty estimates based on different definitions of income

n Persons experiencing unemployment and median unemployment spells

n The characteristics of persons ever participating in government programs, and persons who participated in each month

n Median program participation spells

n The characteristics of persons with lapses in health insurance coverage

n Median spells without insurance

n Family and household transitions

In addition, as our observation length grows, we will be able to better examine issues

Bureau of Transportation Statistics

Carolee Bush, Manager, Department of Transportation

Established by the Intermodal Surface Transportation Effiency Act (ISTEA) of 1991, the mission of the Bureau of Transportation Statistics (BTS) in the U.S. Department of Transportation is to improve the transportation knowledge base for public decision making, and to improve public awareness of the nation's transportation system and its consequences. BTS compiles, analyzes, and makes accessible information on the Nation's transportation system; collects information on intermodal transportation and other areas as needed; and strives to enhance the quality and effectiveness of government statistics. BTS is a statistical agency, a mapping agency, and an organization for transportation analysis, all rolled into one. It assembles and interprets information on the extent, use, condition, performance and consequences to the transportation system. The Bureau's analyses are summarized each year in the Transportation Statistics Annual Report, which also includes an assessment of data needs to improve our understanding of transportation.

BTS also is mandated to initiate data collection programs when existing information is inadequate for the Bureau's analyses or for the intermodal transportation data base required by Title V of ISTEA. The Bureau supplements data collection programs of other agencies whenever possible, and provides the lead only when the required data transcend or fall between existing programs. The Bureau's largest data collection programs are the Commodity Flow Survey (CFS) and the American Travel Survey (ATS). Both of these surveys are conducted by the Bureau of the Census to assist BTS in identifying where freight and people go by all modes of transportation. Data by state from the 1993 CFS are just beginning to be released; preliminary data from the ATS are expected in late fall.

BTS also has a Geographic Information System (GIS) Center and an Information Technology Center (ITC). The GIS Center is beginning to provide data integration and display services as well as analysis for the Transportation Statistics Annual Report and the National Transportation System. The ITC was started to support physical communications between DOT and its customers, to develop new technologies for collection of data, and to provide electronic library and archival services. The ITC gives the public access to BTS reports and data through a variety of channels, including the Internet and its home page (, its Electronic Bulletin Board System (1­800­363­4BTS), and Email (

Statistical information can also be obtained the old-fashioned way by calling 1­800­853­1351. To order BTS products or reach BTS staff, call 202­366­DATA or fax at 202­366­3640. Printed reports and CD­ROM's are free as long as quantities last. To obtain a product catalogue or general information, fax, email or send written requests to: Bureau of Transportation Statistics, 400 Seventh Street, S.W., Room 3430, Washington, DC 20590.


Committee on National Statistics Staff

The Committee on National Statistics (CNSTAT) at the National Academy of Sciences-National Research Council serves to contribute to a better understanding of important national issues by working to improve the statistical methods and information on which public policy decisions are based. The Committee recently released two reports: (1) the report of its workshop on racial and ethnic classification, Spotlight on Heterogeneity: The Federal Standards for Racial and Ethnic Classification, and (2) an interim report of the panel study for the Department of Defense, Statistical Methods for Testing and Evaluating Defense Systems.

New studies:

The Committee recently convened two panel studies in the areas of transportation and public health and conducted a workshop on aging-related statistics. The Panel on Statistical Programs and Practices of the Bureau of Transportation Statistics, a joint study with the Transportation Research Board, is reviewing the data collection procedures and capabilities of the newly established Bureau of Transportation Statistics. The panel is chaired by CNSTAT member Janet L. Norwood of the Urban Institute; the study director is Constance F. Citro.

The CNSTAT Panel on Performance Measures and Data for Public Health Performance Partnership Grants is charged with assessing and recommending performance measures for local, state and federal officials to use in appraising progress in meeting objectives of public health Performance Partnership Grants (PPGs) in a number of areas, including mental health and substance abuse. An alternative to traditional categorical grant programs, PPGs offer states greater flexibility on funding priorities and program management, in return for their accepting greater accountability for results. The panel, sponsored by the Department of Health and Human Services, is chaired by CNSTAT member Edward B. Perrin, Department of Health Services, University of Washington; Jeffrey J. Koshel is the study director.

At the request of the National Institute on Aging, CNSTAT and the Committee on Population convened a workshop to discuss and establish priorities for data on the aging population. Participants reviewed major trends in the population at older ages, identified policy issues for which data may be needed, assessed the adequacy of current data collection to address these issues, and suggested changes to more effectively meet future data needs. The workshop was chaired by Dorothy Rice, Department of Social and Behavioral Sciences, School of Nursing, University of California, San Francisco.

Forthcoming studies:

CNSTAT is also developing two new panel studies. One, requested by the Congress, will review and assess the methods used by the Census Bureau to produce estimates of poverty for small geographic areas. The estimates are used in the allocation of many federal funds, including over $6 billion annually to school districts. This study is sponsored by the Department of Education.

The other new study, sponsored by the Bureau of Economic Analysis, will review the approaches to the valuation of natural and environmental resources and suggest improvements and further research. The study will contribute to BEA's plans for a satellite account to describe such resources in the context of the National Income and Product Accounts.

CNSTAT, at its February meeting, conducted a seminar and discussion on measuring performance in the federal government. The Committee plans for a seminar on longitudinal surveys at its forthcoming meeting in June.

The Committee on National Statistics is chaired by Norman M. Bradburn, National Opinion Research Center, University of Chicago; Miron L. Straf is Director.

For information about the Committee and its reports, see the CNSTAT home page, which is available in May, at or send e-mail to

Where to Obtain Demographic Data on Aging

National Institute on Aging Staff

Federal Web Sites of Interest:

The Census Bureau's home page at offers access to a variety of data and information on aging and the older population, such as statistics on population and housing, the economy, and the experimental TIGER digital mapping system. The International Programs Center (IPC) at the Bureau of the Census has created a computerized data base which provides detailed demographic and socioeconomic information about the aged in the United States and other countries, both industrialized and developing. The International Data Base on Aging (IDBA) contains the IPC's population projections for all countries of the world, as well as information for as many as 51 demographic and socioeconomic variables for each country. Link to

Recent reports and a list of publications from the National Center for Health Statistics (NCHS), including survey descriptions and analyses of both the Supplement on Aging (SOA) and the Longitudinal Study of Aging (LSOA), are available on the Internet through the CDC and NCHS home pages, at

A home page is under construction for the Interagency Forum on Aging­Related Statistics. In the interim, information can be obtained through the Census Bureau home page, subdirectories "population and housing," "social and demographic characteristics," and "age."

Web Sites of Selected Publicly Archived Data Sets:

Asset and Health Dynamics Among the Oldest Old (AHEAD): AHEAD is a longitudinal survey of a representative sample of U.S. individuals aged 70 and older and the families in which they reside. The main focus of AHEAD is to study the impacts and interrelationships of changes and transitions for older Americans in three areas: health, finances, and family. The AHEAD home page at provides: a user's guide with information on sample design, content, and field notes; information on accessing the data; and working papers, journal articles, and monographs.

German Socio­Economic Panel (GSOEP): The GSOEP is a longitudinal micro­data base containing socioeconomic information on private households in the newly united Federal Republic of Germany. It contains objective measures of time use, wages, benefits, and income, and subjective measures of life satisfaction, hopes, fears, and political involvement. The sample allows analysis of the German and foreign­born populations. Begun in 1984, there are 11 waves of data from West Germany. Since 1990, the former East Germany was added to the sample. There are now 5 waves of data that include eastern Germany. The GSOEP is linked with the Panel Study of Income Dynamics (PSID) which is described below. The GSOEP can be accessed at

Health and Retirement Study (HRS): The HRS is a nationally representative longitudinal survey of roughly 7,600 US households headed by an individual aged 51 to 61. The primary purpose of HRS is to study the transition between work and retirement with particular emphasis on sources of retirement income and health care needs. The HRS home page at provides: public release data sets, documentation, and results; abstracts and working papers; and links to other sites and files of interest.

Longitudinal Study of Aging (LSOA) and Supplements on Aging (SOA I and SOA II): The LSOA is a family of surveys beginning with the 1984 Supplement on Aging (SOA I) and follow­up interviews in 1986, 1988, and 1990. The sample consisted of about 7,500 persons aged 70 and older. It looks at functional capacity and transitions from independent to dependent states of health. In 1994­1995, the SOA II drew a new sample of 10,000 older persons and collected data on functional status, health care needs, living arrangements, and social supports. These measures can be compared with similar SOA I data. Survey descriptions and analyses of the data sets are available on the home pages of the Centers for Disease Control and of the National Center for Health Statistics (NCHS)

Luxembourg Income Study (LIS): Begun in 1983, the LIS provides comparable socioeconomic data collected in household surveys for more than 25 countries. These data enable researchers to perform cross­national analysis of social policy issues related to income distribution. The LIS data bank currently contains more than 60 data sets covering the period 1968 to 1992. Additional data for the mid­1990s will be added in 1996. Abstracts of working papers are available, and annual summer workshops introduce researchers to the data base. Link to

National Long Term Care Survey (NLTCS): The NLTCS is a nationally representative, 12­year longitudinal survey designed to study changes in the health and functional status of older Americans (aged 65+). It also tracks health expenditures, service use, and the availability of personal, family, and community resources. Link to

Panel Study of Income Dynamics (PSID): The PSID is a nationally representative, longitudinal survey that looks at the attitudes and behaviors of families and individuals within an economic context. Begun in 1968, the most recent versions of PSID data and supplements can be downloaded from this site at Documentation, errata, and a newsletter are also available.

U.S. Mortality Estimates and Projections: Four data files using data from the Office of the Actuary, U.S. Social Security Administration, provide the latest mortality estimates and projections for the United States. Separate cohort life tables for males and females are presented, as are period life tables by gender. The data files contain seven standard life table functions, as well as five actuarial functions which involve financial calculations. Link to

U.S. Census Files on Elderly: The Population Studies Center at the University of Michigan has census data on­line for public use. These files include two related files pertaining to the older population from the 1990 Public Use Microdata Sample (PUMs). The first is the 3 percent sample of households that had at least one person aged 60 and older in the household. The second file was created by combining households with a respondent aged 60+ from the 5 percent PUMS with the 3 percent PUMs. This yields an 8 percent sample of elderly households in 1990. Both files are also available in state­specific files. Link to

Wisconsin Longitudinal Study (WLS) of Social and Psychological Factors in Aspiration and Attainment: The WLS is a long­term study of a random sample of 10,317 men and women who graduated from Wisconsin high schools in 1957. Survey data were collected from the original respondents or their parents in 1957, 1964, 1975, and 1992. Data and Program Library Services at the University of Wisconsin­Madison provides data and documentation on the WLS. Link to

Web Sites of Policy­Relevant Research About Health, Long­Term Care, and the Economic Aspects of Aging: The NIA has established nine Centers on the Demography of Aging (listed below) to provide innovative and policy­relevant research focused on two key areas: health and long­term care, and the economic aspects of aging.

n Duke University, Duke Center for Demographic Studies

n Johns Hopkins University, Population Center

n National Bureau of Economic Research (NBER)

n RAND Corporation, Labor and Population Program

n Syracuse University, Center for Demography and Economics of Aging

n University of California, Berkeley, Center on Economics and Demography of Aging

n University of Chicago, NORC Population Research Center

n University of Michigan Exploratory Center on Demography of Aging

n University of Pennsylvania, Population Aging Research Center

Livestock Surveys Improved

Sandra Smith, Department of Agriculture

The National Agricultural Statistics Service (NASS) has completed improvements to its Cattle on Feed estimation program and will make several improvements to Hogs and Pigs this year.

For Cattle on Feed, the Agency is now able to provide increased industry coverage while reducing sample sizes by more than 50 percent. Previously, monthly estimates were made of all cattle on feed in seven States (which accounted for 70 percent of the U.S. total) and six additional States quarterly (which brought coverage to over 80 percent). In the new program, cattle on feed is estimated monthly only for lots of 1,000 or more head capacity, but in all States. Monthly coverage is over 80 percent of the U.S. total. NASS worked with industry and data user representatives for the past 4 years to develop the changes. Comparable data for 1992­1995 are being provided each month for the 1,000-plus lots in the original seven monthly States.

The hog industry structure has changed dramatically in the past few years­­from 300,000 operations in 1989 to only 182,000 in 1995. Most of the Nation's pork production is now produced by large organizations under contracts.

Presently, 5,000 operations control about 50 percent of U.S. hogs and pigs. NASS is shifting its hog surveys from a traditional farm operation base to a business enterprise basis in order to properly account for the contractor/contractee arrangements. Hogs will no longer be included in the Agency's Quarterly Agricultural Survey program which collected crops and grain stocks information along with hogs.

NCHS Releases Data from Disability Survey

This summer will mark the first release of data from the 1994­1995 National Health Interview Survey on Disability (NHIS­D), the largest, most comprehensive survey on disability ever conducted in the United States. Designed to meet the data needs of a wide range of users, the NHIS­D represents the efforts of over a dozen federal agencies as well as private organizations, academic researchers and disability advocates. In addition to the broad range of health topics covered on the National Health Interview Survey (an annual survey based on household interviews with a sample of the nation's civilian, noninstitutionalized population), the disability component included questions on communication, mobility and sensory impairments, limitations in role functioning, personal care, and social functioning, and mental illness. Special batteries were developed for children on developmental milestones, school attendance, and special health care needs. Persons found to have moderate to severe disabilities in the initial interview received a follow-up interview which delves more deeply into the social, policy, and medical aspects of disability. Data from these follow-up interviews will be released later this year. Data from the initial interviews will be available in July on public use data tapes and on CD-ROM. For information about the survey and the data files, contact Gerry Hendershot, by e-mail at GEH2@NCH08A.EM.CDC.GOV or by fax at (301) 436-3484.

Statistics about Statisticians from the National Science Foundation

Carolyn F. Shettle

The National Science Foundation (NSF) is responsible for producing and disseminating quantitative information about scientists and engineers, including statisticians. Staff from NSF's Division of Science Resources Studies in cooperation with ASA's Committee on Career Development have been assembling information on statisticians that may be of interest to GSS and SSS members.

Examples of some of the information available are:

The number of doctoral degrees awarded annually in statistical fields (Math Probability and Statistics, Biometrics and Biostatistics, Econometrics, Psychometrics, and Social Statistics) by U.S. institutions increased from 287 in 1985 to 352 in 1994. This 23 percent increase is somewhat below the 38 percent increase for doctorates granted in all science and engineering fields during this period. (Selected Data on S&E Doctorate Awards: 1994 ­­ NSF 95­337)

In 1993, most employed individuals with doctorates in statistical fields reported that they had one of the following three occupations: post­secondary teacher (44 percent), statistician (23 percent), or manager (15 percent). The representation of statistically trained individuals among managers was approximately the same as that for all scientific fields; however, their representation among post­secondary teachers was relatively high, since only 31 percent of individuals with doctorates in all scientific fields reported employment as postsecondary teachers. (Special tabulations from the 1993 Survey of Doctorate Recipients)

Between 1980 and 1990, there was a 22 percent increase in the number of non­academically employed mathematical statisticians at all degree levels, from 27,200 to 33,200. (Nonacademic Scientists and Engineers: Trends from the 1980 and 1990 Censuses ­­ NSF 95­306)

From 1989 to 1993, the number of statisticians employed by the Federal government increased 5 percent from 3,352 to 3,504, compared to a 6 percent increase in total employment for science and engineering occupations. (Federal Scientists and Engineers: 1989­93 ­­ NSF 95­336)

If you are interested in more information about statisticians, look for our upcoming articles in AmStat News or attend the "Statistics on Statisticians" session at the Joint Statistical Meetings (Monday, August 5, at 10:30 a.m.). If you want more information about science and engineering (e.g., information about biologists, mathematicians, sociologists, or economists), you are encouraged to browse through SRS's newly redesigned home page ( Alternately, feel free to request the publications cited by calling (703) 306­1130, by sending an e­mail message to, or by sending a fax to (703) 306­0510.

Bureau of Justice Statistics

The Bureau of Justice Statistics has recently gone online via the internet though its own world wide web page. The site, whose address is, offers the latest criminal justice reports, spreadsheets and new releases in a variety of formats. Over 400 users a day have been accessing the site since it went online February 16. The BJS web page is also linked electronically to the University of Michigan's national Archive of Criminal Justice Data internet site. Users accessing this web site have the capability of downloading the dataset and codebooks for a particular BJS reports and performing their own data analysis on their computer. With over 43 gigabytes of information available through the Criminal Justice Archive, users can explore a wide range of subjects, such as crimes reported to the police, prison populations, federal case proceedings and criminal victimization.

BJS, which is the Department of Justice's statistical arm, collects information that profiles the nation's federal, state and local criminal justice systems from crime through sentence and punishment. It publishes the National Crime Victimization Survey, the National Prisoner Statistics and various other criminal justice data.

Major Data Development Projects-SAMHSA

Ronald Manderscheid, Department of Health and Human Services

The Substance Abuse and Mental Health Services Administration (SAMHSA) has recently awarded two contracts which will result in the assembly of data sets which should be of interest to health statisticians both inside and outside government.

A contract on Analyses for Transition to Health Care Reform was awarded in September 1995. This two-year contract will collect information on public-sector and private-sector developments in the area of managed behavioral health care (MBHC) through an array of mechanisms, including a comprehensive search of available published and unpublished literature, analysis of quantitative data, and a series of focus group and expert panel meetings. The project will identify current information and data gaps, will define additional specific information and data that will be needed for MBHC, and will conduct essential policy analyses to inform the field regarding key issues in MBHC. Federal agencies need more information in this area because states and private organizations are now proceeding with health care initiatives on their own after the non-passage of health care reform legislation at the national level. Among the issues raised by these developments are how to assure that appropriate care will be available for vulnerable populations, many with chronic disorders requiring continuing care; the impact of managed care on public health functions, such as disease surveillance, policy development, and quality assurance; and an assessment of the continuing necessity to preserve the safety net of public services.

A two-year contract has been awarded on Medicaid, Medicare and Managed Care Analyses. This project will acquire, analyze, and present state Medicaid data (for four to six states), national Medicare data, and private-sector managed care data. The focus will be on annual utilization and cost data. Analyses of these data will facilitate the development of comprehensive planning, and will help state mental health and substance abuse agencies adapt to changes in the Medicaid program; the project will also provide federal offices with needed information from these major funding streams, including baseline information for managed care. As a result of significant changes in recent years in the financing of mental health and substance abuse care, a need exists to fill gaps in the knowledge of the demographic characteristics of persons served with these funds, the intensity of services that they receive, how these services are provided, and the associated costs of these services. Although analyses under the contract are to be done for a single year only, a major motivation for undertaking the project is to determine the feasibility of adding subsequent annual data for each of the funding streams considered.

For further information on these contracts, contact Dr. Ronald Manderscheid on (301)443-3343.

Comings and Goings

Edward J. Sondik, Ph.D., is the newly­appointed Director of the National Center for Health Statistics, Centers for Disease Control and Prevention. At NCHS Dr. Sondik will direct a program of statistical and information systems which guide actions and policies to improve the health of the American people. NCHS is the Federal government's principal vital and health statistics agency, established almost forty years ago to monitor the nation's health.

In his position as NCHS Director, Dr. Sondik will also serve as senior advisor to the Secretary of Health and Human Services on health statistics. He will provide the Secretary with technical and policy advice on statistical and health information issues and will serve on the HHS Data Council. This new council will oversee the development of efficient data collection strategies that enhance the analytical capabilities of HHS. In announcing Dr. Sondik's appointment, HHS Secretary Donna E. Shalala said, "to get where we're going, we have to know where we've been. Dr. Sondik, NCHS and the new Data Council will help us develop the blueprints for a healthier America." Secretary Shalala has charged the Council with developing a department­wide data collection strategy, including coordination of surveys and oversight of survey and general statistical analysis.

Dr. Sondik joins NCHS after a 20­year career at the National Institutes of Health. He began his NIH career as Chief of the Program Analysis and Evaluation Branch at the National Heart, Lung and Blood Institute, before moving to the National Cancer Institute in 1982. At NCI, he served as Associate Director of the Surveillance Program and was then appointed Deputy Director, Division of Cancer Prevention and Control, in 1989. Most recently he served as Acting Director of NCI. He is a native of Connecticut and received his B.S. and M.S. degrees in Electrical Engineering from the University of Connecticut and his Ph.D. in Electrical Engineering from Stanford University.

Dr. Susan Offutt is now the Administrator of the Economic Research Service (ERS). From January 1992 until her recent selection, Dr. Offutt worked at the National Research Council (NRC) of the National Academy of Sciences where she was the executive director of the Board on Agriculture and assistant executive officer of NRC. From 1988 to 1992 she was chief of the Agriculture Branch at the Office of Management and Budget.

Dr. Offutt served as a section leader in the Resource and Technology Division of ERS from 1987 to 1988. She first joined ERS as a cooperative researcher from 1985 to 1987.


Daniel Weinberg, Chair, 1995­1996
Bureau of the Census
Housing and Household Economic Statistics Division
Iverson Mall, Room 307
Washington D.C. 20233­8500
(301) 763­8550 (voice); (301) 763­8412 (fax)

Miron Straff, Chair-elect, 1997
National Academy of Sciences
2101 Constitution Avenue N.W
Washington D.C. 20418-0006.
(202) 334-3097 (voice); (202) 334-3751 (fax)

Susan Miskura, 1996 Program Chair
(703) 360-8774 (voice)

Edward Spar, 1997 Program Chair
Executive Director
732 N. Washington St.
Alexandria, VA 22314-1943
(703) 836-0404 (voice); (703) 684-2037 (fax)
e-mail: spare@guvax.georgetown,edu

Virginia de Wolf, Secretary-Teasurer 1996-1996
Bureau of Labor Statistics,
Office of Research and Evaluation
PSB 4915
2 Massachusetts Avenue, N.E.
Washington, D.C. 20212­0001
(202) 606­7374 (voice); (202) 606­7426 (fax)

David Cantor, Publications Chair, 1995-1996


Gerald W. Gates, Chair 1996
Program and Policy Development Office
Bureau of the Census, Rm. 2430­3 (DIR)
Washington D.C. 20233­1000
(301) 457­2515 (voice); (301) 457­2654 (fax)

Linda Gage, Chair-elect 1997
California State Finance Department
915 L Street
Sccramento, CA 95814-3700
(916) 322-4084 (voice)

Carolyn F. Shettle, 1996 GSS Program Chair
Division of Science Resources Studies, Suite 965
National Science Foundation
4201 Wilson Boulevard
Arlington, VA 22230
(703) 306­1780 (voice); (703) 306­0510 (fax)

Michael L Cohen 1997 GSS Program Chair-elect
University of Maryland
9005 Walden Road
Silver Spring MD 20902-3826
(301) 495-0729; (202) 586-3045 (fax)

Rich Allen 1995-1996 Secretary-Treasurer
NASS/OA/Room 4117 So. Bldg.
12th & Independence Avenue, S.W.
Washington, D.C. 20250­2000
(202) 720­4333 (voice); (202) 720­9013 (fax)

Joan Lee Turek, Publications Chair, 1995-1996
Division of Modeling, Computer, and Technical Systems,
Officernment Statistics Session

Sunday, August 4, 1996

4:00 p.m.

Avoiding Statistical Disclosure

Type: Special Contributed - Paper

Chair/Organizer: Virginia de Wolf

Discussant: George T. Duncan

Monday, August 5, 1996

8:30 a.m.

How Much Income Do You Need? How Much do You Need to Spend?

Type: Special Contributed - Paper

Chair/Discussant: Lee Rainwater

Organizers: Thesia Garner and Stephanie Shipp

10:30 a.m.

The Pros and Cons of Privatizing Government Statistics

Type: Special Contributed - Panel

Chair/ Organizer: Lynda T. Carlson

4:00 p.m.

President s Invited Address

6:00 p.m.

GSS Business Meeting (followed by informal dinner groups)

Tuesday, August 6, 1996

8:30 a.m.

Using Administrative Records for Generating Government Statistics

Type: Special Contributed - Paper

Chair: Daniel Melnick

Discussant: Thomas B. Jabine

Organizer: Kent Marquis

Customer Satisfaction Surveys: Looking Towards the Future

Type: Special Contributed - Panel

Chair: Andrew A. White

Organizer: Mary Ann Guadagno

10:30 a.m.

Administrative Records Can Play a Greater Role in National Statistics if....

Type: Special Contributed - Panel

Chair/ Organizer: Gerald W. Gates

Satisfying the Publics Needs for Statistics

Type: Special Contributed - Paper

Chair: Gypsy Banks

Discussant: Nancy J. Kirkendall

Organizer: Denice M. Myers

12:00 noon

GSS Luncheon Roundtables

Discussion Hermann Habermann,
Leaders: Jennifer H. Madans, Robert J. Casady,
John F. Long, & Carolyn Hay

2:00 p.m.

Statistics under the Spotlight: Improving the Consumer Price Index

Type: Invited - Panel

Chair: Jack E. Triplett

Organizer: Brent R. Moulton

4:00 p.m.

Deming Lecture

Wednesday, August 7, 1996

8:30 a.m.

The Effectiveness of Monetary Incentives in Government Surveys

Type: Special Contributed - Paper

Chair/Organizer: Carolyn F. Shettle

Discussants: Robert Groves, Jerry Coffey

Re-engineering within Government Statistical Agencies

Type: Special Contributed - Paper

Chair: R. Ronald Bosecker

Discussant: Cynthia F. Clark

Organizer: Denice M. Myers

10:30 a.m.

The Effect of Downsizing on Federal Statistics

Type: Invited - panel

Chair: Hermann Habermann

Organizer: Wendy Alvey

12:00 noon

2:00 p.m.

Electronic Dissemination of Statistical Information

Type: Special Contributed - Paper

Chair: Jimmie D. Givens

Discussant: Robert Zarnetsky

Organizer: Denice M. Myers

4:00 p.m.

Fisher Lecture

Thursday, August 8, 1996

8:30 a.m.

Statistical Topics in National Security

Type: Special Contributed - Paper

Chair/Organizer: Donald P. Gaver

Discussant: Michael L. Cohen

The Categorization of Race and Ethnicity

Type: Special Contributed - Panel

Chair: Pat Golden

Organizer: Olivia Carter-Pokras

10:30 a.m.

Policy Analysis with Education and Defense Manpower Survey Data

Type: Special Contributed - Paper

Chair: John Czajka

Organizer: Wray Smith

Social Statistics Sessions

Sunday, August 4

4:00 p.m.

Improving the Design, Collection, and Analysis of Social Data

Type: Contributed

Chair: David P. Nichols

Monday, August 5

8:30 a.m.

Census 2000: Re­engineering, Content, and Public Perception

Type: Invited

Organizer: Patricia Becker

Chair: Edward J. Spar

Presenter: Robert W. Marx

Panelists: Keith F. Rust, TerriAnn Lowenthal,
Patricia Becker

A dedication to Maria Elena Gonzales will be made by Katherine Wallman.

10:30 a.m.

Analytical Assessments of Social Conditions and Issues

Type: Contributed

Chair: Margo Anderson

Analytical Modeling for Social Science Applications

Type: Contributed

Chair: Deborah Rumsey

12:00 noon

Social Statistics Poster Sessions

2:00 p.m.

Statesman, Social Scientist, Statistician: Memorial Session in Honor of Clifford C. Clogg

Type: Invited

Organizer: Linda Gage

Panelists: Leo Goodman, Martha F. Riche,
Katherine K. Wallman, Kenneth Wachter,
Donald B. Rubin

4:00 p.m.

President's Invited Address

Deming Lecture

6:00 p.m.

Social Statistics Section Business Meeting

Tuesday, August 6

10:30 a.m.

Statistical Techniques in Health and Mortality Studies

Type: Contributed

Chair: Enrique J. Lamas

Wednesday, August 7

8:30 a.m.

Developments in the Coordination and Measurement of Data on Children and Families

Type: Invited

Organizer: Edith K. McArthur

Presenters: Jerry West, matt Stegner, Mike Pergamit,
Deborah Phillips

Discussant: Linda Waite

10:30 a.m.

12:00 noon

Social Statistics Section Roundtable Luncheons

2:00 p.m.

Survey Design and Measurement for Social Characteristics

Type: Contributed

Chair: Robert A. ,Kominski

4:00 p.m.

6:00 p.m.

Fisher Lecture

Thursday, August 8

8:30 a.m.

Evaluation of the Family and Medical Leave Act

Type: Invited

Organizer: Mary K. Batcher

Panelists: Kerry Levin, Ellen Brave, Donna Lenhoff,
Kate McGonagle, Sherry Dale