GSS Newsletter SSS
American Statistical Association
Sections Get Ready for Dallas..
GSS Ready for JSM!
The Government Statistics Section (GSS) has a full program for this year’s annual meetings -- featuring presentations, panel discussions, and luncheons on sessions dealing with a wide array of topics, from the aging workforce and changing classification codes to privacy, confidentiality, and data access issues. Sessions highlight the latest developments at the Federal, State, and local government levels. Special thanks go to Pat Doyle (1998 Program Chair) and Lynda Carlson (1999 Program Chair) for all of their efforts to line up an outstanding program!
GSS is especially excited to offer three invited sessions at this year’s annual meetings:
In addition, GSS has lined up six very timely Special Contributed sessions -- which have been organized around a common theme and provide a little more time for speakers to present their research in greater depth. Three of these sessions are in panel format, so they should provide for lively discussion -- on random-digit dialing samples for household surveys, measurement of survey quality, and the transition for the Standard Industrial Classification system to the new NAICS industry codes. The other three are "regular" paper sessions, focusing on the latest developments in statistical disclosure limitation, restricted access to Federal data, and statistical uses of administrative records. A Regular Contributed session on improving user access to data is also in the lineup.
For those of you who are looking for a lively luncheon discussion, GSS also has six roundtable groups set up for Tuesday, August 11 at 12:30PM. Roundtable sessions are designed to provide an overview of an issue and then to allow researchers and practitioners to share their experiences. This year’s sessions cover topics well suited to that format as they concentrate on new concerns and data systems --
These are really exciting sessions — please support them and attend! Just remember that the luncheon round-tables are fee events, so be sure to sign up early -- tickets for your choice table may go fast!
Remember to double check all dates and times of sessions you wish to attend, since last-minute changes can occur. Also, be sure to look for sessions other Sections have organized that are co-sponsored by GSS (like many of the Social Statistics Section programs) -- this means that the GSS Program Chair feels the sessions may be of particular interest to our members.
Finally, if you are in Dallas, definitely plan to attend the Government Statistics Section Members Meeting on Monday, August 10, at 5:00 PM in the Dardenelles Room of the Wyndham-Anatole Hotel. Come hear what is new, meet colleagues, make new friends, and become involved in GSS. This is an open meeting and you are welcome!
SSS Program All Set!
The Social Statistics Section (SSS) has lined up an exciting program! We have three invited paper sessions, two special contributed sessions, two regular contributed sessions, one poster session, and seven luncheon roundtables. Also in the plans is a new event -- an open reception for SSS members and friends. (Note that this is separate from our regularly scheduled business meeting and reception.) Plan to be there!
The invited sessions kick off the SSS program with a session on Public Policy Using Statistics, Monday, August 10, 8:30 to 10:20 AM (#45). This is the annual Roger Herriot Memorial session and will begin with the presentation of the Herriot Award to Roderick Harrison and Clyde Tucker, followed by presentations on estimation of global temperatures, by John Adams, James Hammitt, and James Hodges; surveying the disabled, by Allan Sampson; and designing and analyzing social experiments, by Robert Bell and Phyllis Ellickson. Fred Mosteller will be the discussant of this session organized by Sally Morton and John Rolph.
An invited panel on Welfare Reform: How Will We Know What Happened? was organized by Stephanie Shipp and will feature panelists William Brister, Daniel Weinberg, Barbara Gault, Stephen Bell, and Patricia Ruggles, presenting their different views on the potential impact of the 1996 Personal Responsibility and Work Opportunity Reconciliation Act. William Hobby will direct the discussion, adding his perspective, based on welfare reform efforts in Texas. It will be held Wednesday, August 12, at 2:00 PM (#232). A special Social Statistics Section reception will follow this session at 4:00 PM -- be sure to come and join us!
The third invited session is on Statistical Methodologies for Census 2000, organized and chaired by John Thompson, head of the Decennial Census efforts at the Bureau of the Census. Preston Jay Waite and Howard Hogan will describe the statistical methodologies planned for the upcoming census, followed by discussion comments from Mary Mulry, Keith Rust, and Thomas Hoefeller. The session, to be held on Thursday, August 13, at 10:30 AM (#276), promises to be well worth hanging around for!
SSS has also organized Special Contributed sessions on income inequality (Monday, 2:00 PM) and on the new Survey of Program Dynamics (Tuesday, 2:00 PM). Regular Contributed sessions are scheduled on social policy measurement issues (Tuesday, 8:30 AM) and on statistics for health, disability, and evidence in the courts (Wednesday, 8:30 AM). Also on Tuesday, August 11, from noon until 2:00 PM, six SSS poster presentations will be on display -- come on by for an opportunity to talk one-on-one with the author.
See the next article for a full description of the SSS luncheon roundtables and be sure to order your ticket ahead of time. The hot topics fill up fast!
Finally, SSS will hold its Social Statistics Members Meeting and Mixer on Monday, August 10, at 6:00 PM in the Lalique Room, Wyndham-Anatole Hotel. Come hear what is going on in the Section and provide some feedback of your own. Then join us for an informal get together and get to know some of your colleagues a little better.
SSS Luncheon Roundtables
The Social Statistics Section has planned seven roundtable lunches for the Joint Statistical Meetings in Dallas. Five of the topics involve measurement issues associated with social statistics. Two of the topics are based more on politics but involve the use of statistics. The measurement topics include poverty; intra-area cost of living differences; race and ethnicity reporting under the new Federal guidelines; time-use; and defining metropolitan and non-metropolitan areas. The ‘political topics’ include whether and how to use sampling as part of the decennial census and the impact of recent immigration policy changes. Come join us for one of these lively and current discussions. The roundtables are planned for Wednesday August 12 around noon.
So be sure to plan on attending one of these lively and current roundtable lunches. Look in your registration packet for information on how to register.
Hearty congratulations to four of our own Government and Social Statistics members who join the latest class of ASA Fellows! The kudos go to Margo Anderson, University of Wisconsin -- Milwaukee; Robert M. Bell -- The Rand Corp.; Ronald S. Fecso -- National Science Foundation; and Arthur B. Kennickell -- Federal Reserve Board, all of whom are being recognized for their exceptional contributions to statistics and the statistical community. The awards will be presented on Tuesday evening at the JSM in Dallas.
The Votes Are In....
Congratulations to newly-elected Section officers:
Chair-Elect -- Patricia J. Doyle, Bureau of the Census
Program Chair-Elect -- John L. Czajka, Mathematica Policy Research, Inc.
Secretary-Treasurer -- Dorothy Harshbarger, Alabama Center for Health Statistics
Publications Officer -- Signe Wetrogan, Bureau of the Census
Chair-Elect -- Judith Tanur, University of New York, Stony Brook
Program Chair-Elect -- Elizabeth Stasny, Ohio State University
Publications Officer -- Kathleen S. Short, Bureau of the Census
Best wishes, also, to our Section members who were elected to other ASA offices!
A Historical Perspective
As the summer looms and the Dallas meetings get closer, I have begun musing about how the world of statistics looks to the 'lay' public. Statistics sometimes get a bad rap. Mark Twain's quip about 'lies, damn lies, and statistics,' comes to mind, as do my students’
complaints about why they have to square all those deviations from the mean if they are just going to sum them and take the square root in the next step. Who thought that up? (Students who take their required courses in quantitative analysis in the history department are also often math-phobic liberal arts types who ask such 'irrelevant' questions.)
Or we can turn to a subject close to my own heart -- the decennial population census -- and watch Congress and the Clinton Administration spar over the 2000 Census Plan. While the President made a fund-raising and public relations trip to Houston recently to endorse the 2000 Census Plan, Rep. Dan Miller (R-FL), Chair of the House Subcommittee on the Census, accused the President of "peddling statistical snake-oil." And Rep. John Boehner (R-OH), head of the House Republican Conference, charged the President with politicizing the census and said that sampling "corrupts a basic sense of fairness by treating people as numbers that can be estimated, rather than individuals who have a right to be counted."
My goodness, I wondered, what will my students say the next time I tell them to 'estimate' a coefficient? Will they tell me that those survey respondents deserve to be counted, not estimated?!
But then I returned to my 'historical' roots, and consoled myself with a bit of past census 'history' -- grimly remembering Santayana's old saw that "those who cannot remember the past are condemned to fulfill it." All this discussion about sampling -- for nonresponse follow up or for adjusting for the undercount -- brings to mind the laments of Francis Amasa Walker, a remarkably candid nineteenth century Census Director, who fretted over census accuracy and did not have the tool of sampling to help him. (It hadn't been invented yet.) Walker oversaw the 1870 census and came back to Washington in the late 1870s to ask Congress for reform of the census law. He knew he had problems with census accuracy and described his difficulties in vivid detail. "Many persons," he told Congress, "have a 'usual place of abode' only in a qualified sense. They are here or there as business or pleasure or necessity required. This class does not comprise the vicious or the poor alone. It embraces large numbers of persons of ample means, often of wealth, whose local ties are very slight, who drift about from place to place, because they are found in hotels and fashionable boarding houses."
Walker asked Congress to provide him with sufficient staff to shorten the enumeration period from the 100 days then allowed to 2 weeks in populous areas. "A protracted enumeration is essentially vicious," he concluded. "All that can be done by administration, under the best provisions of law, is to reduce the error within moderate limits." In 1878 Congress listened, reappointed Walker Superintendent, and gave him a number of the legislative changes he requested for the 1880 census. They included the creation of the Geography Division to map the country and create enumeration districts, the authority to appoint field staff to take the census, and major expansion of the then temporary census office in Washington, DC. They did not insulate Walker from political fallout after the census, despite the consensus of his contemporaries and later observers that he was a fine administrator.
Years later Walker reflected on his experience, fretted about the "inevitable errors in the traditional census" and proposed something he called "an instantaneous photograph of the people as they are at a given moment." He was not sanguine that he would live to see such improvements in census accuracy. "It may come about in time that the people," he mused, "out of patience with the inevitable errors of the traditional census, and weary with the quarrels and recriminations between States and cities necessary attendant upon it, will unanimously agree to waive the theoretical objections to the photo-graphic method, as possibly, probably, and in some degree certainly, affecting unequally the basis of representation, and will accept the latter system as good enough for political purposes, and as vastly more satisfactory from all other points of view."
The issues raised by Walker's 'photographic method' have strong similarities to the current controversies about sampling. We'll see how it goes...
NAICS: The Manual is Here
Paul T. Zeisset
You may have noticed a number of presentations on the JSM schedule in Dallas which deal with "NAICS." Well, America has a new system for classifying industries, and it will affect data you use. The U.S. Government is retiring the familiar Standard Industrial Classification (SIC) system in favor of the new North American Industry Classification System (NAICS), a scheme developed jointly by the United States, Canada, and Mexico. Now, with the publication of the official NAICS Manual, you can begin to do something about it.
NAICS (pronounced "nakes") identifies over 350 new industries recognized for the first time. Some new industries reflect "high tech" developments, such as fiber optic cable manufacturing and satellite telecommunications. Other new industries simply recognize changes in the way business is done -- e.g., bed and breakfast inns, environmental consulting, warehouse clubs, HMO medical centers, and diet and weight reduction centers.
Using 20 broad sectors (as opposed to 10 under the SIC) and 6-digit codes (instead of the 4 used by SICs), NAICS has codes for 1,170 industries in the U.S., as opposed to 1,004 under the old SIC system. Because the focus of the codes is on grouping industries with similar production methods together, instead of defining them by product, the 4-digit SICs cannot be simply expanded to obtain the new NAICS codes. A substantial number of 4-digit SICs can be reconstructed from NAICS categories, either because the industry was not changed (other than in code) or because new industries were defined as subdivisions of old ones. On the other hand, many other industries have been changed more profoundly, leading to breaks in the availability of time series data. For example, some office supply stores previously classified as wholesalers are being reclassified into corresponding retail industries if they sell primarily through storefront locations similar to other retail establishments. This change affects not only the compar-ability of individual industries, but whole sectors, as well.
Learn How to Use NAICS
A great deal of information is now available about NAICS on the World Wide Web. The Census Bureau maintains a major site --http://www.census.gov/naics -- featuring:
The correspondence tables now on the Web reflect the latest updates -- hundreds of wording changes and a few changes to industry codes made since the adoption of NAICS was formally announced in April 1997. The tables also define which industries are fully comparable among the U.S., Canada, and Mexico, which are common only to the U.S. and Canada, and which are unique to the U.S.
North American Industry Classification System -- United States, 1997, otherwise known as the NAICS Manual, is available ($32.50 hardcover, $28.50 soft-cover, $45 on CD-ROM) from the National Technical Information Service (1-800-553-NTIS) and Bernan Press (1-800-233-0504).
The printed NAICS Manual includes:
The index to NAICS codes on the CD-ROM is somewhat more detailed than the index in print. Each entry includes both a NAICS code and an SIC code, so it is possible to pinpoint the shifts between SIC and NAICS at the finest level.
With the publication of the NAICS Manual, users finally have a tool with which to put NAICS to work. Businesses can determine their new NAICS codes and vendors can start reclassifying their business lists according to the new scheme.
A new promotional brochure is also available. NAICS -- Calibrating a New Economy is available on request from Census Customer Services, (301) 457-4100. Single copies are free.
Still have questions? To ease the transition to NAICS, we have given our resident expert on industry classification, "Dr. NAICS," an e-mail address (firstname.lastname@example.org) and a toll-free telephone line (1-888-75NAICS).
Census 2000 Update
The political aspects of the 2000 Census fight become more evident every day, the
issue garnering increasing public attention as Census Day draws closer.
Meanwhile, Bureau staff are working hard to conduct the Dress Rehearsals, do their studies, prepare in a myriad of other ways for the world’s biggest field operation, respond to an unending stream of questions from Congress, and provide support to the newly established Census Monitoring Board.
Court Cases.--A three-judge U.S. District Court panel heard oral arguments on June 11 in the case filed by House Speaker Newt Gingrich challenging the constitutionality and legality of sampling in the census. The courtroom was packed.
Although the government argued the technical point that the lawsuit is premature, the judges appeared to understand that the decision must be made soon. On the fundamental points of legality and constitutionality, the two sides were pointing to different provisions of the Census Act (Title 13 of the U.S. Code). One provision states that, except for purposes of apportionment, the Census Bureau "shall" use sampling methods wherever possible. Another section authorizes the Secretary of Commerce to determine how the census will be taken, including the use of sampling.
An audio press conference the same day featured constitutional scholar and Harvard law professor Laurence Tribe. He called the lawsuit one of "the Emperor has no clothes," saying that it would not make sense for the framers of the Constitution to say that Congress should direct how the census should be taken and then to limit those methods.
Oral arguments in the other lawsuit, Gavin v. Clinton, will be heard by a three-judge panel in Roanoke, VA, on August 7. This year’s census appropriations bill directs that the court expedite consideration of the cases, with any appeal going directly to the Supreme Court. Even so, it is considered unlikely that a court ruling will come before next February, identified as the final "drop dead" date for deciding how the census will be done.
Dress Rehearsals and Census Planning.--John Thompson, Associate Director for the Decennial Census, reported on current activities and concerns at the June 11-12 meeting of the 2000 Census Advisory Committee to the Secretary of Commerce.
Generally, the Dress Rehearsals are going well. Census officials are elated with the capability and staying-power of local staff; 90 percent of applicants have been hired and the turnover rate is only 20 percent. This is attributed, in part, to the fact that Bureau wages are several dollars per hour greater than generally available for temporary workers in these communities.
Concern is being raised about the problems involved in providing questionnaires in various languages other than English (and Spanish). They have not found an automated solution which would permit putting two questionnaires (of different languages) in the same envelope for mailing. It’s also difficult to target linguistically isolated households.
One new idea which was received with considerable enthusiasm by the Advisory Committee is the prospect of expanding funding for walk-in Questionnaire Assistance Centers. The Bureau would hire local people to staff these centers, but space would be provided by local governments and private organizations.
Discussing the LUCA mailings which went to 17,000 communities with city-style addressing in February, about two-thirds have actually been contacted and over half have agreed to participate in the program. Thompson acknowledged that 1997's Master Address File re-engineering process has worked out well.
Sampling and Estimation.-- Work continues as planned under the direction of Howard Hogan, now Acting Chief of the Decennial Statistical Studies Division. Hogan has prepared written documents which explain the sampling and estimation process in significant detail, for public review and under-standing. In addition, Tommy Wright, of the Census Bureau, has a paper explaining the process in the June issue of the American Scientist. We recommend it as a way of understanding the Bureau’s methods and why they work.
Monitoring Board.-- The Census Monitoring Board held its first meeting on June 3. Most of their time was spent on administrative details, deciding how to divide up the annual $4 million budget, hire staff, and keep track of spending. The next meeting is scheduled for July 8.
Fiscal Year 1999 Budget.-- The House and Senate Appropriations Committees are moving forward on the FY1999 funding process. The Senate has allocated $848 million for 2000 census preparations, the amount requested by the President.
The House Appropriations Subcommittee, on the other hand, allocated $956 million but wants to allow only half the funds to be spent for the period through March 31, 1999. The remaining funding would be held up pending a report from the President on the full estimated cost of completing the census and new legislation permitting the funds to be spent. Rep. Alan Monahan (D-WV), the Subcommittee’s senior Democrat, expressed "dismay" that this plan violated last year’s agreement
regarding resolution of the sampling issue. Consideration by the full House Appropriations committee will come after the July 4 recess.
The 1998 Roger Herriot Award for Innovation in Federal Statistics goes to Roderick Harrison, Bureau of the Census, and Clyde Tucker, Bureau of Labor Statistics, for their outstanding contributions to research on race and ethnicity classifications. The work which
Harrison and Tucker led formed the basis for substantive revisions to OMB Circular No. 15, which establishes new racial and ethnic categories for Federal government reporting. The new classifications were devised to recognize more race and ethnic categories and to allow for persons who fall into multiple racial and ethnic groups. The award will kick off the session on Public Policy Using Statistics, on Monday, August 10, at 8:30 AM.
Shiskin Awards for Jacobs & Gastwirth
The Washington Statistical Society (WSS) and the National Association of Business Economists recently selected Joseph L. Gastwirth and Eva Jacobs to receive the Julius Shiskin Award for Economic Statistics.
Gastwirth, of George Washington University, was cited for his seminal work that clarified statistical properties of the Lorenz curve and related measures of income inequality and for his continuing contributions to statistical methodology needed for economic and labor analysis.
Jacobs, who recently retired from the Bureau of Labor Statistics, was recognized for her management of the Consumer Expenditure Survey (CES) Program, for her use of the CES data to analyze and interpret the economy, and for her responsiveness to customer needs.
The awards and accompanying honorariums were presented June 17, 1998, at the WSS Annual Dinner.
Some Upcoming Events
Below are some upcoming events that are of particular interest to GSS and SSS members:
Record Linkage Techniques -- 1997
This Proceedings is now available from the Record Linkage Workshop and Exposition, held in Arlington, VA, March 20-21, 1997. The volume presents recent papers by two of the pioneers in record linkage research -- Ivan Fellegi and Howard Newcombe -- as well as other work rising from increased privacy concerns due to record linkage; growing interest in exact matching as a means for more efficient use of scarce statistical resources; heightened importance of linkage technology for policy research, including for health care reform; issues of physical security of data; and the measurement of nondisclosure and reidentification risks in public-use microdata files. In addition to U.S., Canadian, and European papers presented at the conference, the Proceedings includes overheads from a tutorial on record linkage, information on some record linkage software, and a selection of papers which highlight developments in record linkage since the 1985 Workshop on Record Linkage Techniques.
For information on obtaining copies of Record Linkage Techniques -- 1997, contact email@example.com or fax Bill Winkler at (301) 457-2299. For an electronic copy, see http://www.nas.edu .
E-Pubs for Canada
Statistics Canada has announced plans to discontinue standard print format of approximately 100 titles of its regular publications and replace them with electronic publications, which will be available over the Internet. The E-Pubs are electronic images of the print publications, available in Adobe Acrobat (PDF) format. To find out what is now available electronically, visit the Web site at http://www.statcan.ca; select Products and Services; click on Down-loadable Publications (for Free) or Downloadable Publications (For Fee/$).
First BTS Journal
The Bureau of Transportation Statistics (BTS) just released the first issue of its new journal. The Journal will present information on the nation’s transportation system and the latest developments in transportation data, theory, methods of analysis, and concepts relevant to transportation systems.
The premiere volume contains articles on measuring transportation in the U.S. economy; data analysis of truck accidents on freeway ramps; estimation of State-level truck activities in America; fare elasticity for urban passengers; a review of literature on the social cost of motor vehicle use; and an analysis of public transport demand in Europe.
For information on the BTS Journal, call (202) 366-DATA or e-mail firstname.lastname@example.org .
Also, FYI --
@amstat.org or fax (703) 684-2037. Please include your mailing address.
GSS Section Contacts
Cynthia Z.F. Clark, Chair (1998)
6928 Butternut Court
McLean, VA 22101-1506
Michael L. Cohen, Chair (1999)
9005 Walden Road
Silver Spring, MD 20901-3826
Linda H. Gage, Chair (1997)
Department of Finance
915 L Street
Sacramento, CA 95814
Pat J. Doyle, Program Chair (1998)
Demographic Surveys Division
Bureau of the Census
Washington, DC 20233
(301) 457-3795 email@example.com
Lynda Carlson, Program Chair (1999)
Energy Information Administration
1000 Independence Ave., SW EI-70
Washington, DC 20585
Yahia Ahmed, Secretary/Treasurer
Statistics of Income Division OP:RS:S
PO Box 2608
Washington, DC 20013-2608
Wendy L. Alvey, Publications Officer
Room 2430 FOB 3 Policy Office
Bureau of the Census
Washington, DC 20233-3700
Carolyn Shettle, Council of Sections Rep.
7475 Wisconsin Ave.
Bethesda, MD. 20814
Robert G. Lehnen, Rep. to COPAFS
342 North Senate Ave., 3rd Floor
Indianapolis, IN 46204-1708
(317) 261-3041 firstname.lastname@example.org
Gerald W. Gates, Rep. to COPAFS
Room 2430 FOB 3 Policy Office
Bureau of the Census
Washington, DC 20233-3700
(301) 457-2515 email@example.com
SSS Section Contacts
Margo Anderson, Chair (1998)
University of Wisconsin -- Milwaukee
Department of History
Milwaukee, MI 53201
(414) 229-4316 firstname.lastname@example.org
Edith K. McArthur, Chair (1999)
National Center for Education Statistics
555 New Jersey Avenue, NW
Washington, DC 20208
Miron L. Straf, Chair (1997)
Committee on National Statistics
National Academy of Sciences
2101 Constitution Avenue, NW
Washington, DC 20418
Carolee Bush, Program Chair (1998)
Bureau of Transportation Statistics
400 7th Street, SW K-20 Room 3430
Washington, DC 20590
Stephanie Shipp, Program Chair (1999)
HHES Division , Room 1071 FOB 3
Bureau of the Census
Washington, DC 20233
Martha Hill, Secretary/Treasurer
Institute for Social Research
University of Michigan
PO Box 1248
Ann Arbor, MI 48106-1248
Patricia C. Becker, Publications Officer
APB Associates/Southeast Michigan Census Council
17321 Telegraph Road Suite 204
Detroit, MI 48219-3143
Constance Citro, Council of Sections Rep
Committee on National Statistics
National Academy of Sciences
2101 Constitution Ave., NW
Washington, DC 20428-0006
ASA Staff Liaisons
Marie Argana (GSS)
Director, Member Services
American Statistical Association
732 N. Washington St.
Alexandria, VA 22314-1943
Barbara Milton (SSS)
732 N. Washington St.
Alexandria, VA 22314-1943