Census Plan Doesn't Add Up

A failed handheld initiative points to bigger tech problems in government.   

With much fanfare, the U.S. Census Bureau announced about two years ago that it was modernizing.

The bureau said it would issue a state-of-the-art handheld computer to each of the 2010 census takers to record the count at each household and its specific location (via GPS) and then upload all the data for processing. The prime contract went to Harris, which would provide Windows Mobile-powered devices using Sybase mobile middleware for deployment.

The 500,000 units would provide the government with the most effective method for retrieving accurate and complete data, at a price of $600 million.

Fast-forward to March. With little fanfare, Commerce Secretary Carlos Gutierrez, whose department oversees the bureau, announced that he was curtailing this project due to cost overruns and systems that didn't work effectively.

The department would still buy 151,000 devices for the now-inflated cost of $1.3 billion, but they would be used primarily as GPSes, and the actual data would be collected the old-fashioned way-with paper and pencil.

It's obvious that the government didn't create a viable strategic plan up front. Further, it did not get its users involved in the design of a system that was user-friendly.

There were also too many chefs in the kitchen, as specs changed continuously, and Harris had to make changes on the fly at the same time it was trying to create the core system functions. Finally, the complexity involved caused the help desk costs to skyrocket-a major piece of the price tag increase.

Clearly, good project planning and practice is not a forte of the U.S. government, but Harris, which has done many government projects in the past, should have known better. How many enterprise projects have gone the same way?

So who are the winners and losers?

Microsoft lost out on its claim to power the largest handheld deployment ever, and Sybase will sell fewer licenses. Harris will lose its bragging rights to future projects, although it did get more revenue than originally planned even without completing the full scope of the project.

But the biggest losers are the American people. Any public corporation would have its stock price trashed and management replaced with a project fiasco such as this. But who in the government will pay the price for such a dismal failure? No one. And it will leave us with one more example of how our current public systems cannot cope with the transition necessary to bring them into the 21st century.

Jack Gold is the founder and principal analyst at J. Gold Associates, of Northboro, Mass.