Home » Columns, Science Policy

Controlling Decennial Census Costs Major Theme of Senate Hearing

1 June 2011 1,436 views No Comment

Early planning, research, technology, and fixed term cited often

This column is written to inform ASA members about what the ASA is doing to promote the inclusion of statistics in policymaking and the funding of statistics research. To suggest science policy topics for the ASA to address, contact ASA Director of Science Policy Steve Pierson at pierson@amstat.org.

Contributing Editor
Pierson-color copySteve Pierson earned his PhD in physics from the University of Minnesota. He spent eight years in the Physics Department of Worcester Polytechnic Institute and later became head of government relations at the American Physical Society before joining the ASA as director of science policy.

How to control costs of the 2020 decennial census dominated testimony and questioning during an April 6 Senate hearing, titled “Census: Learning Lessons from 2010, Planning for 2020.” The many approaches the U.S. Census Bureau is taking to address the cost of the next decennial census—while maintaining data quality—and realizing that prudent funding for 2020 decennial census planning now will save money down the road will be part of the message when trying to convince Congress to support the Fiscal Year 2012 (FY12) Census Bureau budget request.

2010 Census Preliminary Evaluation
According to U.S. Census Bureau Director Bob Groves, early evaluation of empirical performance indicators suggests the decennial did well. The mail return rate in the 2010 Census matched the mail return rate for the short form of the 2000 Census. Since there was still a long form in 2000—the American Community Survey has since replaced it—the 2010 mail response rate “was actually better” than the combined short and long form return rate from 2000, Groves told the panel. This occurred, the director suggested, because of the mailing strategy, including a replacement form, the expanded partnership program, and more targeted advertising.

One sour note was that the number of households from which the bureau received information from “proxies”—neighbors, building managers, or others—rather than direct information, increased from 17% in 2000 to 22% in 2010. There was also a slight decrease in some item response rates compared to 2000, with the largest decline coming in the Hispanic Origin question. The difference was 94% in 2000 compared to 92.8% in 2010.

Another positive indicator providing confidence in the results, Groves contended, occurred because the census count matched the demographic analysis population estimates. The count was 308.7 million people. The mid-range estimate (out of five) was 308.5 million. The bureau is also in the midst of its post-enumeration survey that samples 170,000 housing units to help verify the results of the count. One preliminary result Groves reported was that address list matching significantly improved over 2000.

Excerpted from the “Census Director Outlines Lessons Learned From 2010 Count” section of Silver’s April 18, COSSA Washington Update.

Chaired by Sen. Thomas Carper (D-DE), chair of the Census Bureau authorizing subcommittee, and joined by subcommittee ranking member Sen. Scott Brown (R-MA) and Sen. Tom Coburn (R-OK), the first panel of the hearing featured Census Bureau Director Robert Groves, Commerce Department Inspector General Todd Zinser, and Government Accountability Office (GAO) Strategic Issues Director Robert Goldenkoff. Remarkably, the oversight witnesses—Zinser and Goldenkoff—had little criticism of Groves’s strategy for the 2020 decennial census planning.

With the cost of enumerating households rising from $16 in 1970 to $98 in 2010—both figures in 2010 dollars—GAO has estimated that the 2020 decennial could rise to as much as $30 billion if the current design model is used. The 2010 decennial census cost $13 billion. Groves reported the Census Bureau’s goal to “design a 2020 census that costs less per housing unit than the 2010 census, while maintaining the quality of the results.”

Among the three witnesses, the following common themes for how to reign in census costs emerged:

  • Collect census data through multiple modes of technology to make responding to the questionnaire as convenient as possible, whether the modes be the phone, handheld device, Internet, or face-to-face. Implicit to these capabilities is having information technology architecture that can support multiple modes of data collection, including modes yet to emerge.
  • Use administrative records (e.g., drivers licenses or records from the Veterans Administration or the Social Security Administration) to supplement the enumeration process
  • Continuously update address lists and maps, instead of intensely pushing toward decade’s end
  • Risk and operations management changes such as shorter and smaller tests to ensure subsequent tests can incorporate test results, or systems development that doesn’t require first-use perfection
  • Execute system and procedure innovations that benefit the entire institution
  • Better cost-estimating and budget processes
    • Carper, Zinser, and Goldenkoff also spoke about the importance of a fixed term for the census director, who currently serves at the pleasure of the president. The average tenure of a census director since 1969, according to Goldenkoff, is three years, which makes the long-term planning so essential to a successful and cost-effective census.

      Another theme discussed during the hearing was the importance of learning census lessons from countries such as Canada, Brazil, and the United Kingdom.

      With spending cuts dominating congressional budget discussions, it will be a challenge to convince lawmakers to fully fund the FY12 request for the Census Bureau. For FY11, the Census Bureau was funded for $93 million less than their request. The FY12 request—$242 million less than the FY11 request because of 2010 decennial census wrap-up—contains $67 million for 2020 decennial planning, $9 million to enhance use of the administrative records, and funds for exploring the sample size of the American Community Survey and executing the Economic Census and the Census of Governments.

      Praise for Groves
      Sen. Tom Coburn and Sen. Thomas Carper effusively expressed their support for U.S. Census Bureau Director Bob Groves. Adding to his past praise of Groves as being “just what the doctor ordered,” Coburn said, “I want to say publicly how enthused I am that we have very super competent leadership at the census. And I have great faith in Dr. Groves. I’ve seen what he took on, how he accomplished his mission, and his commitment to using science to make his organization more efficient. And so I’m one of your big backers.”

      Referring to a Senate legislative proposal to give the census director a fixed five-year term starting in 2012, Coburn asked, “Are we going to get to keep you?” Carper seconded the thought, saying he hoped Groves would be the first director to fill a fixed five-year term.

      Mindful of the current fiscal environment, the FY12 budget request contains $20 million in savings across the Census Bureau from administrative changes and $16 million in savings from eliminating nine programs, including the Statistical Abstract program.

      With the role Census Bureau data play in informing policy and guiding more than $500 billion in federal spending, as well as the potential to save money on the 2020 Census through prudent investments now, there is a strong case to be made for fully funding the FY12 Census Bureau requested budget. The challenge is making sure lawmakers hear the case repeatedly from a variety of valued sources. ASA members can and should play an important role in communicating this message to their U.S. senators and representatives.

      1 Star2 Stars3 Stars4 Stars5 Stars (No Ratings Yet)

      Comments are closed.