Ken Fisher at Ars Technica points out what everybody with even a rudimentary understanding of mathematics or just simple logic*, knew when it was proposed, simply the shift in the DST enacted by Congress would not (and did not) accomplish what it was intended for, namely saving energy.
The US government's plan to boost energy savings by moving Daylight Saving Time forward by three weeks was apparently a waste of time and effort, as the technological foibles Americans experienced failed to give way to any measurable energy savings.
While the change caused no major infrastructure problems in the country, plenty of electronics and computer systems that were designed with the original DST switchover date (first Sunday in April) failed to update. The inconvenience was minor, and the potential savings were great. Or so we were told by the politicians behind the move.
As it turns out, the US Department of Energy (and almost everyone else except members of Congress) was correct when they predicted that there would be little energy savings. This echoed concerns voiced after a similar experiment was attempted in Australia. Critics pointed out a basic fact: the gains in the morning will be offset by the losses at night, and vice-versa, at both ends of the switch. That appears to be exactly what happened.
I would point that for American business the inconvenience was more than just minor, as every computer system had to be evaulated for a potential DST impact. As someone that works in the computer industry, it was shades of Y2K all over again.
It would have taken a lot of energy savings to cover the waste in human productivity associated with the DST switch. The fact that it generated no savings is just a travesty. But hey, at least they looked busy! Just wait until they get into the CO2-offset regulation business! They'll really look like they are doing something.