Summary of the December 10, 2004 COPAFS Meeting

COPAFS Chair Dun Muff started the meeting by introducing past Chair Maurine Haver, who presented the slate of nominees for the 2005 COPAFS board. The nominees were:

Chair: Sarah Zapolsky
Vice Chair: Dick Kulka
Secretary: Ken Hodges
Treasurer: Judie Mopsik
At large: Maurine Haver
Ralph Rector
Felice Levine
Bob McGuckin.

The nominees were approved by voice vote, and we went to Ed Spar for his Executive Director’s report.

Ed welcomed the Conference Board as a new COPAFS affiliate, and thanked Rick Ayers of ESRI for the Unlocking the Census with GIS publication, which they provided for complimentary distribution to COPAFS reps.

Ed then reviewed the latest budget numbers, commenting that they reflect the recently passed omnibus funding legislation, but not an across-the-board 0.8 percent downward adjustment that will be made to reconcile the House and Senate grand totals. Most notable is that the Census Bureau got the funding needed for full ACS implementation – a major recovery from the Senate mark. The census funding bill contained language requiring that the 2010 census include a “Some other race” category in the race question. Several stakeholder groups have pressed for retaining this category, but Ed commented that it is bothersome for something like this to be legislated, rather than left to the experts. Ed made special note of the full funding for NCHS – achieved through a lot of work on the part of NCHS proponents, including COPAFS.

With sadness, Ed reported the demise of COPAFS affiliate American Demographics magazine. After over 25 years of publication, the magazine was sold, and will be maintained only as a column in Advertising Age magazine

Dates for the 2005 COPAFS quarterly meetings are March 11, June 10, September 16, and December 9.

Current Activities at the National Center for Education Statistics.

Robert Lerner. National Center for Education Statistics.

Robert started with a clarification of Ed’s introduction—explaining that he is now the former commissioner of NCES because his present term has expired, and his reappointment is pending. Outside of the meeting, Ed explained that his reappointment has been blocked as part of some political maneuvering, and it remains to be seen how things will work out.

Robert described NCES as having two basic points of reference – 1) the federal statistical community and 2) the Institute of Educational Sciences. Much NCES research work involves statistical description, which he defined as the systematic recording of observations of what is out there, with respect to teachers, students, schools, and other aspects of education. These descriptive data serve as building blocks for explanatory and policy analyses, and are used to improve accountability systems.

The subject of drop outs is one of current interest, and one that poses some real measurement problems. As Robert described it, drop out (or conversely, “completion”) measures can be viewed as status measures (for students) or accountability measures (for schools). For example, graduation rates are typically defined as the percent of students graduating with a regular diploma in the “standard” number of years. This measure ignores those who return to school or later get a GED. Work is ongoing at NCES to determine how best to measure drop out/completion rates, and as Robert described it, some states are better than others at using these measures. The measurement issue is further complicated by private schools and the practice of home schooling. NCES is similarly involved in developing measures of completion for post-secondary education—where there are policy implications for financial aid programs.

NCES also has measures related to charter schools. Defined as a “public school of choice,” charter schools are chartered by local authorities with a mission to serve special populations or objectives, and may be waived from some of the requirements imposed on other public schools. There are about 1,500 charter schools nationwide, and NCES is working to determine how best to study them. Robert also described an early childhood longitudinal survey that NCES is conducting in response to policy interest in early childhood. As he noted, there has never been a survey following students from early childhood through high school—an approach that removes the cohort selection issues that plague other surveys.

NCES also is looking at teacher attributes, and their impact on student achievement. The objective is to identify what teacher attributes best predict student achievement. Results so far suggest that teachers’ verbal abilities are more important than advanced degrees, credentials, and certification, to student achievement.

A Review of the New Service Sector Economic Indicators

Frederick Knickerbocker. U.S. Census Bureau

Nick described that on September 13, the Census Bureau released the first data from the Quarterly Services Survey (QSS) – a new survey focused on the service sector. The release was greeted with enthusiasm and a question – What took you so long? The release also met with allegations that its timing was influenced by the politics of the presidential election, but Nick asserted that nothing could be further from the truth.

In fact the Census Bureau had been working toward better services data through the 1980s and 1990s – collecting more service sector data in the Economic Census, and in the Services Annual Survey (SAS). And the funding and interagency planning process for the QSS dates back to 2001.

Implementation issues include the timing requirements of federal agencies and what industries to cover as part of “services.” The QSS currently covers NAICS sectors 51 (Information), 54 (Professional, scientific, and technical services) and 56 (Administrative and support and waste management and remediation services). Data are collected from just under 5,000 firms (not establishments) with paid employees. The sample is drawn from the SAS, and does not include the self-employed. By law, quarterly surveys are voluntary, and response so far has been good—with weighted rates of slightly over 80 percent, based on a mix of mail, Internet, telephone and fax response modes.

To illustrate the importance of the service sector, Nick noted that the sectors in the QSS have annual revenues of $7.8 trillion, compared with $3.8 trillion for manufacturing. He explained that a Census Bureau definition of services (consisting of 12 private services producing sectors in the Economic Census) accounts for about 55 percent of GDP. Adding wholesale and retail trade increases this to 70 percent, and adding government services would get one to 80 percent of GDP. The QSS reports sector totals, but not a grand total because other sectors likely will be added, and the Bureau does not want the press to mistakenly report “total services” numbers based on incomplete measures.

Looking ahead, Nick explained that the QSS is not yet seasonally adjusted, but that such adjustments will start after 16 quarters of data have been collected. They also plan to benchmark the QSS to the SAS, and will expand the QSS in 2005 to collect data on hospitals, nursing, and residential care facilities – increasing the sample size to about 6,000.

Nick described as “unfinished business,” the fact that data on finance, insurance, real estate, transportation, and utilities are collected only once every five years. These industries need to be moved first into the annual survey (SAS), and then to the quarterly QSS. The long term objective is to collect more data on products and activities across all services sectors.

During the follow up discussion, there was comment that the controversy over the QSS release was fueled by the lack of advance work with non-federal data users. There was also comment that this was not all the Census Bureau’s fault, but agreement that economists needed to be better prepared to respond to the inevitable press inquiries related to the initial data release.

An Update from Capitol Hill.

Jim Moore. Majority – House Committee on Government Reform.
David McMillen. Minority – House Committee on Government Reform.

Ursula Wojcieschowski, with the House Subcommittee on Technology, Information Policy, Intergovernmental Relations and the Census, was scheduled for this panel, but was unable to attend.

Jim Moore noted that some congressional reorganization is going on, but that he would continue to work on the census with Representative Tom Davis and the full committee. He noted that they were able to get the “bare minimum” required for ACS full implementation, and credited data user letter writing with making a difference. Jim described the situation as “touch and go” for a while, and commented that the recovery from such a cut was remarkable.

When asked if congress recognizes the need for long term commitment to the ACS, Jim said that there is implicitly such recognition, but noted that in recent weeks, everyone was focused on this year’s budget. Asked which points were most persuasive with the appropriators, Jim said it was those that identified local users, thus translating to votes. The disincentive of a 2010 long form also was credited with playing a part. David McMillen observed that the letters of support gave the House something tangible to support their higher mark. And timing was important. There was agreement that hitting the issue late (in the process), but at the critical time, worked very well.

Jim closed with an observation that the Department of Commerce and Secretary Evans have been very supportive of the ACS, but we will have to see what kind of support we get with the new Secretary.

David McMillen commented that legislators are funding Jay Waite’s three-legged stool (Census, ACS, MAF/TIGER), but asked that we be aware of the programs that are not being adequately funded – such as demographic analysis and the intercensal estimates program, on which the ACS depends. These programs involve relatively small amounts of money, but the Census Bureau is challenged to communicate their importance to appropriators.

For those who had not finished Christmas shopping, David recommended two forthcoming GAO reports—both looking back at the 2000 census. One looks at the 3 million duplications in Census 2000, and the other looks at data quality measures.

David noted the language in the omnibus funding bill calling for chief privacy officers at federal agencies—and commented that while it might not be a good idea for all agencies, it is a good idea for the Census Bureau. Asked if there was fallout on the Hill from the Census/Homeland Security controversy, Jim said there is no real concern from the majority side, and said they don’t want to get into the game of picking data users. David said concern on the minority side is with the threat to census response rates, and the need to recognize the reality that some small area data can be damaging to population groups – even when confidentiality is preserved. David expressed concern that we are not handling that discussion very well right now.

Estimates of the Legal and Unauthorized Foreign-Born Population

Jeffrey Passel. Urban Institute

Jeff described that he and others have been working for some time on residual measures of unauthorized immigration, with the objective of improving the Census Bureau’s population estimates. The residual approach involves subtracting data on legal immigration from data on total foreign born population to estimate unauthorized immigrants.

The distinction between legal and unauthorized immigration is not always clear. Among the “legals” are refugees, asylum approvals, Cuban-Hatian entrants, other entrants, IRCA legalizations, and INS new arrivals. Also, all pre-1980 entrants are assumed to be legal. Unauthorized immigrants are those not in the “legal” categories, and include overstayers, and entrants without inspection (EWIs). Jeff also described “quasi legals” such as asylum applicants, and others with temporary protective status.

The methods involve the ongoing estimation of the legal immigrant population, and assumptions about the undercount of this population (based on A.C.E.). Jeff cited numbers indicating total post-1980 immigration of about 20.5 million, of which 13.1 million are estimated to be legals. After adjustments for estimated undercount, an estimate of about 8.3 million unauthorized immigrants is derived. Jeff contrasted this estimate with pre-Census 2000 estimates putting the unauthorized population at only 5 to 6 million, based on CPS data. He noted that by late 2000, CPS data were suggesting more like 6 to 7 million unauthorized, and that when later re-weighted to Census 2000, the CPS suggests as high as 9.8 million. As Jeff described it, there appears to have been a rapid increase in unauthorized immigration starting in about 1998.

The recent estimates suggest that the foreign born population consists of about 30 percent legal permanent residents, 31 percent naturalized citizens, 28 percent unauthorized, 7 percent refugees and 3 percent temporary legal; and it appears that most of the unauthorized arrived in the last 10 years. The unauthorized are largely from Latin America, with 56 percent from Mexico, and 26 percent from “other Latin America.” About one third of the foreign born population is now from Mexico. Jeff observed that it is not uncommon for such a large proportion of the foreign born to be from one origin, but the fact that Mexicans in the U.S. are now about 10 percent of the size of the population of Mexico is unprecedented. Jeff reported further that the new flows from Mexico are dominated by the unauthorized, and that this population is spreading from states where they have been (and still are) concentrated, to new areas. Jeff pointed out that Mexico is changing a lot, and these changes have implications for the future U.S. population. For example, Mexico’s total fertility rate of 7.3 in 1960 has dropped to 2.4 now – although Mexican immigrants to the U.S. have a rate of 3.3.

Jeff concluded by reporting that very recent data (from multiple sources) suggest that total immigration to the U.S. may have peaked during the 1999-2001 period. The rates are still high, but may have dropped due to conditions following the September 11 terrorist attacks.

Concerns of COPAFS Constituencies

No concerns were raised, and the meeting was adjourned.

  • Maurine Haver, NABE
  • Tom Witt, AUBER
  • Jerry Fletcher, AAEA
  • Richard Forstall, AAG
  • Patricia Becker, APDU/SEMCC
  • Ken Hodges, PAA/Claritas
  • Michael P. Cohen, BTS
  • Linda Jacobsen, APDU
  • Sarah Zapolsky, AARP
  • Stephen Tordella, Decision Demographics
  • Dorothy Harshbarger, NAPHSIS
  • Ralph Rector, The Heritage Foundation
  • Thomas E. Brown, IASSIST
  • Nitin Kapoor, IBM
  • Felice Levine, AERA
  • Nicholas Zill, Westat
  • Dick Kulka, RTI
  • John Czajka, ASA/Mathematica
  • Fred Cavanaugh, Sabre Systems, Inc.
  • Seth Grimes, Alta Plana Corp.
  • Miron Straf, National Academies
  • James Wing, Beyond 20/20
  • Howard Leathers, UMD/AAEA
  • Nancy Bates, AAPOR
  • Carolee Bush, AAPOR
  • Lindsay Clark, Brookings
  • Margaret Martin, COPAFS
  • Susan Lapham, ESSI
  • Marilyn Seastrom, NCES
  • Robert Lerner, IES/NCES
  • Don Muff, Muff Consulting Services
  • Dan Levine, Westat
  • Gooloo Wunderlich, CNSTAT
  • Sameena Salvucci, Mathematica
  • Michael Lee Cohen, CNSTAT
  • Mesfin Mulatin, MayaTech Corp.
  • Kelvin Pollard, Population Reference Bureau
  • Charles Hulten, AEA
  • Dan Estersohn, Arbitron, Inc.
  • Jane Smith, Fenestra Tech.
  • John Munyon, SeeData, Inc.
  • Rick Ayers, ESRI
  • Doug Skuta, ESRI
  • Katherine Wallman, OMB
  • Susan Schechter, OMB

The Council of Professional Associations on Federal Statistics

20 F Street, NW, Suite 700, Washington, DC 20001 |

©2010 The Council of Professional Associations on Federal Statistics. All rights reserved.