Summary of December 12, 2003 COPAFS Meeting

COPAFS chair Don Muff opened the meeting with the election of the 2004 COPAFS Board. The new board was voted in by acclamation. We then went to Ed Spar's executive director's report.

There are still no final 2004 budget numbers (and probably won't be until February), but Ed had us look at the latest budget sheet, which indicates where the conference committee numbers are headed. The good news for the Census Bureau is that they are getting most of what they requested, and the conference report has specific language confirming that the ACS will move ahead. The bad news is that, because Census will absorb some cuts, MAF/TIGER enhancements will not proceed as planned. BLS actually shows an amount greater than requested, reflecting the current popularity of programs such as the mass layoff statistics. NCHS is making out better than it had been, thanks in part (according to Ed) to the education efforts on its behalf by supporters including COPAFS. The situation at BJS is said to be uncertain, as many of its functions are being considered for outsourcing.

NCES faces similar uncertainty, as many of its functions are being transferred to the Institute of Educational Science. As Ed put it, we are seeing the degradation of a statistical agency. OMB has released a list of statistical agencies and functions determined to be inherently governmental (for Circular A-76 purposes), but Ed has not seen it yet. There was comment that the issue of outsourcing is a source of angst and bewilderment, with some wondering, for example, how BLS functions might be deemed inherently governmental, while those at BJS are considered commercial.

Next, Ed described a proposal to move the Bureau of Transportation Statistics to what he called an "odds and ends" division (Research and Special Programs Administration) of the Department of Transportation. Under this arrangement, the head of BTS would no longer be a Senate appointment and report to the Secretary of Transportation, and Ed described it as a potential disaster for BTS. COPAFS helped save BTS from an earlier proposal for its elimination, and Ed said COPAFS is talking with Secretary Mineta's office about the proposed transfer.

Ed wrapped up by describing the success of the November 4 conference on the new metropolitan areas, and plugging the conference on the reporting of economic statistics (December 18 at the National Press Club). Meeting dates for 2004 are March 12, June 11, September 17 and December 10.

The Economic Research Service: Programs and Products

Phil Fulton. Economic Research Service

Phil described ERS as the social science research agency at the Department of Agriculture. ERS addresses a broad range of questions for users including policy makers, federal agencies, researchers, trade associations, businesses, and the media. Outlets for ERS information include its website, research publications, Amber Waves magazine, congressionally mandated studies, journals and professional conferences.

Amber Waves, described as "a window into ERS research," publishes highlights of research findings five times a year, and Phil recommended that COPAFS members check it out. Phil also described the Agricultural Resources Management Survey (ARMS), calling it an Ag version of the CPS. The ARMS is an annual survey of about one percent of farms, and collects information on production practices, costs, and the financial condition of farms and farm households.

ERS research topics are varied, with many relating to "today's farms and Farmers" (for example, the structure of agricultural production, the diversity and economic well-being of farm households, and farm income). An interesting finding from this area is that farm families are actually better off than the average U.S. household, if one includes the value of land and off-farm wealth and income. Other topics include farm and commodity outlook, research and technology (such as the adoption of bio-engineered crops), global markets, policy, food markets, food safety, diet and health, food assistance, natural resources (such as manure management), and rural America.

The presentation concluded with a demo of the ERS website.

A Redesigned Questionnaire for the 2004 Survey of Income and Program Participation

Pat Doyle. U.S. Census Bureau

Started in 1984, SIPP is a multi-panel, longitudinal survey asking households about labor force status, income, assets, and (of course) program participation. Respondents are interviewed every four months for a period of two to four years. A paper instrument was used until the process was automated with CAPI in 1996. However, with automation, response rates dropped from about 95 percent to below 90 percent, and attrition increased. As Pat described it, people got nervous, and there was a decision to change. A Continuous Instrument Improvement Group (CIIG) was formed with the objective of reducing burden (to stem attrition), improve response and data quality, and improve field operations. Much of the effort focused on the development of a new instrument for the 2004 panel.

SIPP now has that new instrument, and Pat noted that the difference is not so much in content as in the way questions are asked. For example, not all respondents have to be asked all questions. All households are asked basic questions about recipiency, but now screeners based on income often permit the interviewer to forgo asking, for example, a mansion-dwelling respondent about food stamp use, or a financially struggling respondent about stock ownership. The new instrument also uses an expanded reference period to address "seam bias" relating to the timing of response relative to monthly transitions in program participation. Response rates are improved by giving respondents flexibility in reporting periodicity of earnings and asset income (hourly, bi-monthly, annual, etc.). And in Wave 2 interviews (if respondents give permission), respondents can be reminded of a previous response, and asked if it is still valid. Pat also described other improvements including cosmetic changes, improved field training, a 2000-based sample, and the use of incentives.

Summing up, Pat acknowledged that some at the Census Bureau are nervous about all the changes, but expressed high confidence in the SIPP staff, and in the new instrument's ability to achieve improved field operations, improved data quality, and better rapport between interviewers and respondents.

Puerto Rico: Building the Link to Meaningful Data

Lu Jeppesen and Tony Netto. SeekData, Inc.

SeekData is a private company playing a leading role in the development of data products for Puerto Rico-where the demand for data has exceeded availability. As Lu described it, addresses and address coding are important to "building the link to meaningful data" for Puerto Rico.

However, there are many challenges in building address-based systems and data products for Puerto Rico-where conventional addresses are much less common than in the U.S. Among the challenges are "anomalies" including Urbanizations, which Lu described as similar to gated communities, and which require a fourth address line. Further challenges are provided by the use of building names in addresses, alpha-numeric house numbering, and multiple same-named streets in close proximity.

As part of building the link, SeekData has built a lexicon to deal with common misspellings, and an "aka file" to identify where roads change name when passing through a town.

Next, Tony described the ambitious work they do in "maintaining the link." Using state of the art GPS equipment and sometimes satellite imagery, SeekData collects feature and attributre data for all new roads, and has been awarded a (Commerce/Census) contract to geocode Puertro Rican addresses and to collect information on housing units and roads that are new since April 2000. As part of this work, they have developed an inventory of new addresses and new construction. Tony also described work in which they have compared TIGER information with their GPS-based work, and taken the initiative to move TIGER roads to match their home-grown data. As Tony explained, they have a major interest in the Census Bureau's MAF/TIGER enhancements, but realize that corrections for Puerto Rico are likely to be a relatively low priority for the Census Bureau program.

Lu wrapped up with further descriptions of the challenges presented by Puerto Rican addresses, and an invitation to learn more about the company at its website

Insights Into the OMB Clearance Process for Federal Statistical Surveys

Brian Harris-Kojetin, Office of Statistical Policy, OMB
Karen Lee. Office of Information and Regulatory Affairs, OMB

Karen explained that OMB's Office of Information and Regulatory Affairs (OIRA) oversees and coordinates the information collected by the federal government, coordinates statistical policies, budgets, and standards, and oversees government information policy. As part of these responsibilities, OMB is required by the Paperwork Reduction Act to review federal data collection activities. The purpose is to improve the quality and utility of federal information, and to reduce the paperwork burden on the public.

The review process takes about 180 days, starting with a 60 day Federal Register notice period, followed by the addressing of public comments, and an OMB review period ranging from 30 to 60 days. The objective is to ensure that the information collected is not overly burdensome, not duplicative, and has practical utility. Karen then described the process, and the types of information that agencies must submit to OMB-including a description of the data to be collected, justification for the collection, and steps to ensure that burden is minimized. She noted that there are two types of reviews-regular and emergency-and that any information submitted becomes public.

Brian described the work of the Statistical Policy staff, and characterized it as an exercise in having a fresh set of eyes looking at the data collection process. Staff serve as desk officers for principle statistical agencies, and often consult with each other on statistical collections. When Brian noted that OMB staff "have teeth," mostly prior to data collection, he was asked if they have the authority to stop a data collection in the field. He said they do, but that he does not recall that happening. Asked how stretched their staff is, Brian noted that there were once 90 people working in Statistical Policy, but they now have a total of seven.

Brian noted that OMB is concerned with (and agencies need to submit information on) universe and respondent selection, procedures for collecting information, methods to maximize response, measures to address the potential for non-response bias, and justification for the use of incentives (this last item to avoid shifting burden from respondents to taxpayers).

Karen wrapped up the presentation by offering hints on how to make OMB review less stressful (allow enough time, and communicate with OMB). She also indicated that their current activities include the preparation of "clearance package," updating a Statistical Policy Directive on Standards for Statistical Surveys, and providing OMB training opportunities.

Concerns of COPAFS Constituents

No concerns were raised, and the meeting was adjourned.

  • Dick Kulka RTI International
  • Ralph Rector The Heritage Foundation
  • Jane Smith Fenestra Technologies
  • Rick Ayers ESRI
  • Colleen Flannery Census
  • Susan Lapham ESSI
  • Samantha Friedman American Sociological Assn
  • Lee Herring American Sociological Assn
  • Melissa Shea Sabre Systems
  • Fred Cananaugh Saber Systems
  • Seth Grimes Alta Plana Corp
  • Judie Mopsik Abt Associates
  • Joe Garrett Mathematica Policy Research
  • Kelvin Pollard Population Reference Bureau
  • Chuck Waite NABE
  • Jerry Fletcher AAEA
  • Tom Witt AUBER
  • Maurine Haver NABE
  • Dorothy Harshbarger NAPHSIS
  • Linda Jacobsen APDU
  • William Kandel RSS
  • Ken Hodges PAA, Claritas
  • Sarah Zapolsky AARP
  • Clyde Tucker ASA
  • Jonel Peckyno-Haley Temple University-ISR
  • John Munyon SeekData
  • Lu Jeppesen SeekData
  • Phil Fulton USDA/ERS
  • Mary Reardon USDA/ERS
  • Ed Goldfield Census, CNSTAT
  • Bob Parker GAO
  • Tim Tang NEA
  • Stephen Tordella Decision Demographics
  • Joanne Pascale Census
  • Anna Chen Census
  • Nancy Bates Census
  • Jeff Moore Census
  • Mary Moien NCHS
  • Brian Harris-Kojetin OMB
  • Tony Netto SeekData