(AP Photo/Michel Euler)

Back in 1784, hanging out in Paris and heady with Enlightenment, Benjamin Franklin had an idea. Struck by the fact that Parisians were sleeping during sunlight hours and then staying up late at night by candlelight, he calculated the number of candles that were being wasted -- and came up with a very impressive number, 64 million pounds worth of them. Franklin therefore jokingly proposed a massive schedule change, noting that a fortune could be saved through "the economy of using sunshine instead of candles," and even suggested at one point that perhaps cannons be fired at sunrise to get everybody out of bed.

Such was the germ of the idea that would eventually lead to daylight saving time -- that if we patterned our lives to rise and set with the sun itself, we'd save energy and money. Flash forward 230 years later, and this remains the basic reason why many of us will wake up this coming Sunday and realize that it's brighter outside than we're used to. That’s because "fall backward" (the return to standard time that takes effect early Sunday) in effect restores an hour of sunlight, which daylight saving time had placed in the evening, back to the morning.

But there's a problem with this (well-lit) picture. While we'll all surely appreciate the extra hour of sleep this Sunday morning, it is increasingly looking like Franklin's idea about saving energy was wrong. Genius though he was, he seems to have forgotten one thing: Moving around daylight hours doesn't only change how much people need to rely on artificial lighting, whether in the form of candles or modern halogens. It also changes the overall complexion of temperatures that we experience while we're awake. And in an age of heating and air conditioning, that may cause us to reach for the thermostat -- with big, probably negative nationwide impacts on energy use, greenhouse gas emissions, and your wallet.

It should be noted, incidentally, that this view of daylight saving time is not fully accepted. Indeed, some authorities still suggest that the massive scale shift in human behavior known as daylight saving time may save energy, rather than waste it. But the contention has come under attack by researchers, and indeed, one of the most impressive studies out there calls it into serious question.

The paper, published in 2011 in The Review of Economics and Statistics by Matthew Kotchen of Yale and Laura Grant of the University of California-Santa Barbara, took advantage of what the researchers dubbed a "natural experiment" with the effect of daylight saving time in the state of Indiana -- which has some serious time issues. Much of the state is on Eastern Time, but some western parts use Central Time. And until 2005 -- when the state passed a law to make daylight saving time uniform -- some Indiana counties also practiced saving time while others didn't. In effect there were three groups of counties as far as time went: Eastern Standard plus daylight saving, Central Standard plus daylight saving, and Eastern Standard without daylight saving.

This isn't just temporal chaos -- it's a scientific dream come true. Knowing the coming law (effective 2006) would switch everyone in Indiana to daylight saving time, the researchers realized they could jump in and examine before-and-after energy use in Indiana counties that were already using daylight saving, vs. those that had just switched to it. Not unlike a drug trial, there was a "control" group and a "treatment" group.

The researchers also had the cooperation of Duke Energy, which provided a massive data set of monthly utility bills for nearly 230,000 Indiana residents, organized by their locations. And they had weather data, meaning that they could chart energy use against temperature fluctuations (which are obviously a very central factor in heating and cooling). And the results, at least for followers of Franklin, were shocking: Daylight saving time increased energy use in the counties that had just switched to it, by about 1 percent during the period when it was in effect. The overall cost translated into $ 3.29 per person per year -- nearly $ 9 million overall across Indiana. And on top of that, the added pollution resulted in an additional $ 1.7 to $ 5.5 million per year in “social” costs. Ouch.

So what caused daylight saving to apparently backfire (rather than backlight)? "In the spring, you are basically making people wake up in the early morning, the coldest time of day, when they might turn up their heat," explains Yale's Kotchen. "And in the summer, if you take an hour of sunlight and you move it from the morning and put it in the evening, people are more likely to be running their air conditioner harder in the evening." That's because heat from the sun builds up over the course of the day, making that particular hour of sunlight hotter than it would have been if it occurred in the morning rather than the evening. So the "Benjamin Franklin effect" -- people using less artificial lighting -- turns out to be overwhelmed by heating and cooling choices.

Kotchen  warns that you can't robotically extrapolate these findings to other states, but he does think that Indiana is likely to be a fairly good test case. He also says that "I think 1 percent would be an underestimate for what you’d get for the whole country" in terms of added energy use. That's mainly because a lot of parts of the US use way more air conditioning than Indiana does. And needless to say, 1 percent or more increased energy use across America is a seriously big deal.

So on Sunday, when you wake up and it's sunnier than you expect, you can thank Ben Franklin -- but maybe also remember that he could have used a better study design.