Last week Commerce Department Secretary Carlos Gutierrez told a House panel that the Census Bureau was dropping plans to use newly developed handheld computers to collect information from Americans who did not mail in census forms for the 2010 census. In his testimony, he said the handhelds were part of a larger plan to make the census "better, faster, and simpler."

The plan, Gutierrez said, was to address the increasing problems that the bureau is facing that threaten the accuracy of the census, including a larger population, the changing shape and diversity of American families, and a decreased response rate to the census because of a growing distrust of government and because of privacy concerns. These problems have led to lower productivity of the temporary workers the bureau hires to go door to door to count Americans, which requires hiring even more temporary workers to make up for the lost work. Gutierrez said the bureau developed the GPS enabled handhelds to collect more accurate address locations to make it easier for the workers to find residences.

While some of the steps in the revised census plan were offered up as solutions to these problems, the early arguments given to use the handhelds had nothing to do with a "better, faster and simpler" or, more important, accurate census, as Gutierrez claimed. Jay Waite, the No. 2 manager at the Census Bureau, told me during an interview on June 6, 2007, that the bureau decided to develop the handhelds based on the nearly single argument that they would save money. (Gutierrez never said anything about a "better, faster, simpler and cheaper" census.) The 2000 decennial census had cost $6.5 billion and estimates showed the next census would surpass $10 billion.

That's where risk management experts say the bureau made its first, and most critical, mistake. Risk experts say the first question to ask before launching such a program is: "What are you trying to do?" or "What problem are you trying to solve?" As Gutierrez testified, the answer to that question would have been a more accurate count. For Waite, it was to reduce cost. (Ironically, the $445 million in savings Waite's business plan promised was long ago eaten up in budget overruns and will now cost billions of dollars more than the original estimate of $11.3 billion.) Both answers may or may not have led to developing handhelds, and we won’t know, unless all options are considered and analyzed. But given that accuracy is the primary complaint heard around census time (because of the political and financial implications of the count), maybe the bureau should have pursued policies that would have led to a more accurate count, and handhelds, at least based on the arguments presented for them seven years ago, would have not been the answer.