Census Bureau facing huge cost increase, possible delays in 2010 effort
The 2010 Census is at serious risk of massive cost increases and delays caused by problems with a contract to develop handheld computers that the Census Bureau plans to use to collect data.
Because of the problems, costs for the decennial census could increase by between $600 million to $2 billion, according to testimony given by the Government Accountability Office Wednesday at a hearing of the Senate Homeland Security and Governmental Affairs Committee. The increase would be on top of the $11.8 billion the bureau estimates the census will cost.
The problems with the contract are so great that the GAO took the unusual step of adding the project to its list of high-risk programs during mid-year. Typically, GAO adds programs to its risk list every two years at the beginning of new Congresses. GAO has added programs during mid-sessions only four times.
"So serious is the crisis that if it is not dealt with immediately, it could threaten the integrity of the 2010 count," said Sen. Tom Coburn, R-Okla., in his opening statement at the hearing. "We now find ourselves facing a situation in which we might have to revert to conducting the census entirely by paper -- no differently than it was done in the 19th century."
The source of the trouble is the Field Data Collection Automation Contract, which the bureau awarded in 2006 to Harris Corp. for $595 million. Under the contract, Harris is developing handheld computers for 525,000 enumerators who will use them to collect data from Americans who do not send in their census forms. Since awarding the contract, the project has experienced problems, including changing system requirements, which have caused costs to increase and the bureau to miss deadlines.
At the hearing, Commerce Secretary Carlos Gutierrez acknowledged in blunt terms that "significant miscommunication concerning technical requirements between the Census Bureau and Harris" were a main reason for the failings. David Powner, director of information technology management issues at GAO, said final requirements for the FDCA program were not delivered to Harris until Jan. 16.
For years, the bureau's top executives have downplayed any perceptions that the contract was in trouble. Most recently, at a Dec. 11 hearing of the House Subcommittee on Information Policy, Census and National Archives, former Census Director Charles Louis Kincannon blamed the delays on Congress' inability to pass the bureau's budget and downplayed the nature of the program's shortcomings.
However, at Wednesday's hearing, Gutierrez testified that "the problem was more significant than had been conveyed in the Dec. 11 hearing" and admitted that the Census Bureau and Harris "could have done things differently and better over the past couple years."
"This is a case of classic risk mismanagement," said Bob Charette, a risk management expert who consults with federal agencies. "Very early in the contract, [the Census Bureau] should have realized they did not do some of things they should have for a large-scale automation like this. Doing it now is like locking the barn door after the horse is gone."
Gutierrez and Steven Murdock, who recently replaced Kincannon as head of the bureau, offered four options to fix the contract's problems. Under one, called the baseline option, Harris and the Census Bureau would complete the FDCA project as planned, while each of the other options progressively remove a number of functionalities of the handhelds, with those functions being completed by paper. The $600 million to $2 billion estimate of the cost increase depends largely on which option is chosen. But sources expressed doubt about that estimate because the bureau cannot accurately predict the cost and timeframe of the 2010 census at this time.
COMMENTS
- I worked for the 2000 Census and we did a fine job with paper and pencil. Erasers are a lot cheaper than tech support and if I am not mistaken anyone who was hired knew how to read and write so no training was needed for the enumerators. Hand helds might be a good idea ON paper but it's a little too close to zero hour for that now, wouldn't you say? Mary alford Posted August 11, 2008 11:30 AM
- Sounds to me as though once again we have someone in the Bush administration who is doing a heck of a job. Dr. Sally Caldwell Posted May 16, 2008 12:51 AM
- Seems to me that its a little late to turn back now. The bureau has already poured millions of dollars and countless hours of work on this program. With the scheduled test of the HHCs coming up in June it seems crazy to pull the plug now. Having worked in the 2000 Census I can tell you that paper is not easy. Every sheet of paper must be reviewed by hand to ensure accuracy. If the handhelds do work, then its a no-brainer. The question is - do they really work? The only way to find out is to go forward with the scheduled test in June. Of the four options noted only one has the Bureau staying with the handhelds for the biggest part of Census - Non-Response Follow-up. Going away from the handhelds for NRFU seems like an expensive - and risky - option. Count Me in 2010 Posted March 23, 2008 12:06 PM
RELATED STORIES
- Census program to use handheld computers said to be in 'serious trouble' 01/02/08
- GAO: Delays in systems testing put 2010 census at risk 12/12/07
- Commerce seeks emergency aid for Census Bureau 10/16/07
- GAO cites cost, schedule risks in IT projects supporting census 10/10/07
- Census Bureau exposes personal data on public Web site 03/09/07









