Power splurge: the daylight-saving time myth.

LAST YEAR'S energy bill extended daylight-saving time (DST) by a month, on the theory that it would encourage Americans to save energy.A recent working paper from the Center for the Study of Energy Markets suggests otherwise.

An ideal study of daylight-saving time would randomly allocate time schemes across regions, a difficult experiment to execute. So Ryan Kellogg and Hendrik Wolff, Berkeley graduate students in economics, looked at the next best thing. When Australia hosted the Olympics in 2000, two states, Victoria and New South Wales, started DST two months early. Looking at Victoria, the state that didn't host the Olympics, and excluding the time span of the games (when usage patterns were aberrant), Kellogg and Wolff compared Victoria residents on DST to those on a normal schedule one state over.

Australians, they found, were apt to use more electricity under the DST regime. Waking up in darker, colder conditions during September and October prompted a dramatic spike in usage each morning. Victoria residents did ease up on electricity at night, but the decrease was not enough to make up for the morning surge. And while U.S. legislators are hoping shifting an hour of sunlight from the morning to the evening will help balance out energy patterns, the Australian experiment showed more-dramatic swings with DST than without.

Will the findings hold in the U.S. ? At the very least, Kellogg and Wolff predict, the DST extension "will fail to yield the anticipated energy savings." September in the Southern Hemisphere is equivalent to the American March, the very month Congress just tacked onto DST.