In 2007, a team of federal energy researchers faced this question: Could their work be done on the cheap? Their agency, the Energy Information Administration, was facing budget pressure. For 25 years, it had collected comprehensive data on commercial buildings, using an army of surveyors to walk America's streets and update its building rosters. But putting all those boots on the ground was expensive. So for the 2007 survey, to save money, the researchers hired a well-established contractor at the University of Chicago to use an experimental method. The plan failed. Two weeks ago, EIA announced it won't release results from its 2007 survey because the contractor's data were shoddy.