Minutes of the June 6, 2008 COPAFS Meeting

Ed Spar. Executive Director’s Report.

Ed Spar started the meeting by noting that Census Bureau director Steve Murdock was in attendance, and asking if he would make a few remarks. Murdock thanked the COPAFS attendees for their support of census activities, and described the recent changes due to the problems with the handheld computers. Murdock stressed that contrary to reports that all blame rests with the Census Bureau, it is a combination of the Census Bureau’s managing of the contract, and the contractor not completing some of its duties. The director described the supplemental funding requested for the 2010 re-plan – funding needed to keep the contractor working on its activities in the revised plan.

Asked about plans for the use of replacement questionnaires in 2010, Murdock said it is still the plan, but that it could change depending on the budget. He also confirmed that the 2008 Dress Rehearsal is proceeding, but without follow up for non-response—because of the late change to paper-based methods.

Spar then started his Executive Director’s Report with a review of budgets – noting that BLS faces some major problems due to a sudden drop in its budget. If the cuts hold, BLS faces some difficult choices – such as whether to eliminate or cut back on the American Time Use Survey, or to preserve the ATUS and cut something else. Ed Sondik of NCHS described the budget challenges faced by his agency. In particular, he described efforts to keep the vital statistics system going despite severe budget cuts. Spar also pointed to significant cuts in the works for BEA, but described the BJS situation as “so far so good,” but also one of “be careful what you wish for.” A CNSTAT panel has recommended that BJS do more subnational work, but now they have to get funding for this work. Turning to transportation, Spar noted that the National Household Transportation Study will go forward. The survey had been in doubt, as the Bureau of Transportation Statistics lacked funds for it, but the Federal Highway Administration is able to fund it.

Spar gave us a heads up about a metropolitan areas conference planned for the first quarter of 2009. The conference coincides with OMB’s once per decade review of the standards for defining such areas. No major changes are anticipated, but there is a need to review the definition of combined areas, and to address issues posed by the fact that the ACS will provide annual data on commuting – and therefore a basis for re-defining metro areas every year. The potential complications of annual metro area updates are tricky enough to make this another “be careful what you wish for” situation.

This year’s remaining COPAFS meetings are September 12 and December 5.

The Interagency Forum on Aging Related Statistics – An Overview.

Edward Sondik. National Center for Health Statistics

Sondik described the mission of the Interagency Forum on Aging Related Statistics as encouraging cooperation and collaboration among federal agencies to improve the quality and utility of data on the aging population. The forum began in the mid 1980s with NCHS, NIA and the Census Burau, and expanded in the mid 1990s to include six additional agencies. As part of its mission, the forum holds workshops and meetings, and produces publications designed to identify needs and issues related to the aging population.

Sondick distributed copies of a publication Older Americans 2008: Key Indicators of Well-Being as an example of their work. The objective is to present indicators that are solid – based on reliable nationwide data, objectively based on research, and balanced among a variety of topics. The definition of “older” is often age 65+, but varies by measure and source of data.

Major topics covered by the report include the number of older Americans, race and ethnic composition, marital status, educational attainment, living arrangements, and older veterans. Reviewing some of the major tables, Sondik noted the projected rapid increase in the older population, and the increase in the percent that are living alone. Turning to health status, he noted that life expectancy has been increasing, although more for women than men. A key objective of the forum is to inform policymakers, and Sondik cited the relatively low rank of the U.S. on life expectancy as an indication to policymakers of the improvements to be made in this area.

Sondik then reviewed data from a number of other areas, such as chronic health conditions and functional limitations, where data show older people have fewer limitations in the ability to walk several blocks or lift 10 pounds. Somewhat surprisingly, the data on diet quality indicate that 90 percent of older Americans get the recommended amount of fruit in their diet, but the data on physical activity are not as positive. Time use data suggest that much of older Americans’ leisure time is spent watching TV. The report also presents data on nursing home utilization (the rate for population 85+ has dropped in recent years), as well as health literacy—where the data suggest that many older Americans living alone are not well equipped to make their own health care decisions.

Sondik finished up by describing areas where more data are needed – including caregiving, elder abuse, mental health, pension measures and residential care. And he referred us to the forum’s website at

The 2010 Census Partnership Program – An Overview.

Timothy Olson. U.S. Census Bureau.
Lisa Blumerman. U.S. Census Bureau.

As Olson described it, the objective of census partnerships is to reach into local communities – especially those with hard to count populations – and generate public awareness that leads to increased mail response. To underscore the importance of the 2010 partnerships, Olson described the challenging environment faced by the 2010 census – including the post September 11 psyche, the growing population diversity, privacy concerns, deepening distrust of government, and (perhaps most important) the charged debate over immigration. As Olson put it, 2010 is a different world than 2000.

For 2010, the partnership program is being integrated with the communications program, and has three major objectives. First is to increase mail response. For budget purposes the Census Bureau is projecting a 64 percent mail response, but they are hoping for and working to achieve something higher. The second goal is to improve accuracy and decrease the differential undercount. Overcoming the environmental challenges will be a key to achieving this goal. The third goal is to improve cooperation with enumerators. Speaking of enumerators, Olson described the recruiting challenge for 2010, with the Bureau having to recruit 3.8 million applicants to hire 1.3 million workers across all census operations.

Partnership Program tools include a 2010 census public website, partnership toolkits, and customizable materials – for example, permitting partners to translate promotional materials to languages other than the 18 for which standard support is provided. Olson also described the 2010 Census Planning Database, which identifies hard to count populations at the tract level. There was some question over exactly how much of this database is publically available now, so Olson and Lisa Blumerman promised to get back with a clarification.

Next, Olson described plans for partnership implementation, the Tribal Liaison Program, targeted outreach to hard-to-count populations, Complete Count Committees, and the Language Assistance Program, including the first time use of a bi-lingual census questionnaire in some areas. In response to a question, Olson said the current plan is still to provide replacement questionnaires in English only – even in areas where bi-lingual forms were mailed – but that this plan is being reconsidered.

Olson said the partnership program’s current staff is about 120, and will peak at about 680 from January 2009 to June 2010 – then ramp down to 48 in 2011. Partnership staff often are chosen for their multi-lingual capabilities, status as community leaders, and instincts as community organizers. Recruiting this staff is a critical challenge.

The broad timeline for the partnership program is as follows.

2008 Recruit partners
2009 Train partners
2010 Engage the population.

In the Q&A session Olson stressed that, contrary to the impression that some people have, the partnership program is not a grant program. The Census Bureau does not provide funding to partners.

Update from The Census Project.

Mary Jo Hoeksema. Population Association of America and The Census Project.

In a special lunchtime presentation, Hoeksema first discussed census funding for FY2008 – where just to get through the current fiscal year, the Census Bureau needs about $210 million in supplemental funding due to the problems with the handheld computers, and the late change to paper-based NRFU.

The additional census funding is part of an Iraq War spending bill that includes supplemental spending for a number of domestic programs. A major obstacle is that the House is determined to make cuts in the domestic portions of the bill, and it is not known whether the census spending will be reduced as part of those cuts.

Timing is tight as the Census Bureau needs to either pay Harris Corporation for its work on address canvassing, or provide 30 days notice if they cannot pay. Either way, the funding question needs to be resolved by mid August.

Asked if the Census Bureau could possibly get supplemental funding separate from the larger bill, Hoeksema said that while unusual, there has been talk of the possibility, as Congress does seem to recognize the imperative nature of the census count.

Looking to FY2009, Hoeksema noted that with the re-plan, 2009 costs will now be higher than the original 2009 budget request. The Census Bureau is due to provide details on how much additional funding is needed, and why, and the House is scheduled to mark up the funding bill Thursday June 12. So there is hope that we will soon have an early indication on 2009 funding. However, Hoeksema cautioned that final passage is another matter, and that this being a presidential election year, we are almost certainly be dealing with continuing resolutions. She summed up the situation with the observation that congressional appropriators are saying reassuring things, but you never know. v

Local Update of Census Addresses (LUCA) – New York City’s Experience.

Joseph Salvo. New York City Department of City Planning.

Salvo provided background on the LUCA program, in which local governments get Title 13 clearance so they can compare the Census Bureau’s address list with their own, and recommend additions and deletions to hopefully improve the census count. As Salvo described it, LUCA is critical because of how critical the address list is to the census.

In the New York City LUCA effort, the Census Bureau’s Master Address File (MAF) was compared to NYC administrative files on property assessments, certificates of occupancy of new construction and conversions, and electric utility accounts. They geocoded the Census Bureau’s address list using their own system, and then compared against the NYC lists. The lists were summarized at the building level, and patterns of discrepancies were examined. For example, they might find a group of buildings where NYC files show three units each, but the Census list indicates only two per building.

The NYC LUCA effort adds 196,973 units over the census list, with the largest number added to Manhattan, where Salvo indicated there has been considerable development not identified by the Census Bureau. Of the additional units, about 103,000 reflect new construction or conversions, reflecting the fact that there has been much more construction and conversion activity in the 2000s than in the 1990s.

Other additional units reflect non-matches with the census files – sometimes units in existing buildings that have been subdivided or “cut up” – some “informally” or without approval. Many of these are in areas where most of the population is foreign born, and may be fearful of responding to the census. Salvo said the city has some data on these violations, and just wants to help the Census Bureau include these dwellings in the 2010 census. Field counts conducted by the city as part of LUCA confirm that the city’s files also miss some of these additional units in existing buildings, and Salvo showed pictures of some interesting examples. Because some of these dwellings are so unconventional, Salvo suggests that the Census needs special operations in areas where they are concentrated.

Salvo also explained that the shortfall in the Census address list is not as simple as the addresses not being included in the U.S. Postal Service Delivery Sequence File (DSF). The addresses may be in the DSF, but if they are on new streets or have addresses that otherwise cannot be geocoded, they are held in a file of uncodable addresses. MAF/TIGER enhancement does not address this problem because it focuses on spatial precision rather than improving the geocoding scheme to recognize new addresses. Salvo also commented that it is not possible to canvass all addresses in New York City, but one wonders if the canvassing effort could be productively focused on the list of uncodable addresses. In the Q&A session, there was comment that the Census Bureau is supposed to follow up on such addresses throughout the decade, but that the regional offices are not sufficiently staffed to do this.

Looking at “what’s next,” Salvo said the Census Bureau will field check the city’s LUCA work in the spring of 2009, and report their findings in fall 2009. The city can then appeal the Census Bureau’s determinations from fall 2009 through early 2010, and Salvo reported that the city is preparing to appeal – defensively assuming that the Census Bureau will reject all of the new addresses.

In summing up, Salvo noted that the city and the Census Bureau agree on the number of housing units in NYC – they just don’t identify the same buildings. Asked if they have identified addresses that should be removed from the MAF, Salvo said they have not had time to focus on deletes, but he added that their work makes it apparent that there are many duplicate census addresses. Asked how much the city is spending on LUCA, Salvo reported that they have 24 people working full time on LUCA, and that the city has provided $200,000 in funding. Altogether, he said the total expenditure is easily about a half million dollars.

Development of the NSF/Census Business Research and Development Survey (BRDS).

Lynda Carlson. National Science Foundation.

Describing her presentation as a follow up to the presentation on innovation measures (at the March COPAFS meeting), Carlson said the “new” BRDS will be an important part of the federal statistical portfolio. It will be a comprehensive (and sometimes only) source of statistics on R&D and innovation. It is also a component of an NSF mandate to provide information on science and engineering enterprise, as well as NSF’s support of the American Competitiveness Initiative/America Competes Act of 2007.

The new BRDS is to be a redesign of an existing survey that started in 1953, but has not really been updated since. It is an annual survey at the firm (as opposed to establishment) level, with a sample of about 32,000 R&D performing companies. The survey collects data on five mandatory items including total and federal R&D, sales, employment, and R&D by state. The survey is being redesigned because, as Carlson put it, the business context of the 2000s is very different from that of the 1950s. For example, businesses (not governments) are now the largest source of R&D expenditures, academia (not business) is now the basic research performer, services have become more important than manufacturing, large companies are no longer as dominant, and businesses are much more global in focus.

The redesigned survey follows through on NSF recommendations related to the need for information on industrial R&D. Among the objectives are to determine the relevant data that companies have and how to get it – such as through industry expert panels, and company contacts. Among the lessons learned are that different types of data are stored in different parts of companies – no one person has access to all the data, and in some cases, the desired information is not knowable. A challenge will be getting the survey to the most knowledgeable respondents in a variety of companies – the idea being that the “right person” can provide the “right data.”

Survey content covers four major components including financial measures of R&D activity, R&D employment data, nature and technical aspects of R&D, and intellectual property. The goal is to make it easy for companies to report, so there is a choice of paper and web response options, and the a short form has been developed with guidance from Don Dillman and extensive cognitive and usability testing.

The BRDS timeline calls for redesign efforts through 2008, the launch of a pilot redesigned survey in January 2009, and evaluation throughout 2009. The redesigned BRDS is to launch in January 2010 (collecting data on 2009), with delivery of the 2009 data scheduled for December 14, 2010.

Future plans call for the addition of innovation and other questions to the survey, the consideration of industry specific modules, and a pilot survey of firms with 1-4 employees. The objective is to continuously review and update survey content, methodology and processing, and to maintain data quality while reducing respondent burden.

Concerns of COPAFS Constituents

No concerns were raised, and the meeting was adjourned.