EDITOR'S DESK: When to pull the plug
People hate to write off major expenditures that bought little in return. Maybe that's why the government too often clings to projects it should have canceled tens or hundreds of millions of dollars earlier. Like gamblers who keep feeding a cold slot machine, they hope the next appropriation or task order will move the project back on track.
Occasionally, agencies break out of this behavior. The Interior Department, for example, fired BearingPoint Inc. from the long-overdue Financial and Business Management System project [GCN, Oct. 10, Page 13]. Interior had spent $63.1 million over two years, and was to receive another $43.6 million in 2006 for the SAP software implementation.
More often, agencies aren't sure what to do. The General Services Administration has hired KPMG LLP to look at its Preferred IT tracking system, another SAP implementation that Unisys Corp. has labored on since 2002. GSA received $28.4 million for it last year, but requested no dollars for 2006. The system can't seem to connect with other agencies' finance systems.
Then there are the agencies that keep spending no matter what. The Defense Department is being pressed by Congress to cancel the Defense Travel System. DTS, to be generous, sort of works. But critics say the project, led by BDM International Inc., is $200 million over budget and four years behind schedule.
One reason agencies respond so haphazardly to poorly performing contracts is that no metric exists for when to cancel. In fact, the Federal Acquisition Regulation allows for a lot of discretion in canceling a contract.
Savvy investors often predetermine when they will sell off stock according to a selected metric. Government agencies should build in similar stop-loss metrics before awarding a contract. Make them part of the deal, as both fair warning and incentive to the contractor, as well as to avoid protests upon termination.
The specifics of such metrics, though, should not be mandated in the FAR or by the Office of Management and Budget. The people who understand a project and its mission and technology goals are in the best situation to figure out when to pull the plug.