Technology Projects: Thin line between Success and Failure

This article from today's edition of the Wall Street Journal really leaves me scratching my head. Here are the facts:

- The Census Bureau is scrapping the use of new, hand-held devices to complete the 2010 census, for which they paid $600 million to a high-tech company for development
- The cost of NOT using them will add an additional $3 billion to complete the census
- The effort to fix the devices so they can be used in the future will double the value of the contract to this high-tech company, raising it to $1.3 billion

But, here is what really makes me crazy. According to the contracted company,
the devices reportedly operated with 99.5% accuracy. The reason that they won't be used is reported as the lack of "comfort level" of using them by the census bureau. According to the census bureau, the reason for the failure boiled down to scope creep.

So, the bottom line is that as a result of spending $600 million to successfully complete a project (99.5% accuracy has to be considered a success), the government will now spend an additional $3 billion to complete the census, and will then tack on an additional $700 million to "fix" the successful project - all so that the agency can be more "comfortable" using the devices. All of which could have been resolved with better requirements management throughout the project.

So, who ends up footing the bill for the $3.7 billion cost of poor requirements management? The Project Manager? Nope - the U.S. taxpayers.

Wow.