Courtesy of Flickr user Faramarz Hashemi, under a Creative Commons license.

I mentioned last week that car travel in America appears to have peaked backed in 2004. Since then, "vehicle miles traveled" per person in the U.S. have been falling or flat-lining, prompting a fascinating debate over whether we're witnessing some fundamental shift in the American relationship to the car, or some economic blip instead.

Timothy J. Garceau, a Ph.D. candidate in geography at the University of Connecticut, and professors Carol Atkinson-Palombo and Norman Garrick offer a different way to think about the answer. In research they presented this week at the annual meeting of the Transportation Research Board, they looked at travel data not at the national level, but by state instead.

Their results further challenge the argument that Americans have merely been driving less of late because of the bad economy: Washington state experienced "peak car travel" all the way back in 1992, and Nevada, Idaho, Kentucky, Oregon, Rhode Island and Virginia all did before the new millennium. By this measure, peak car happened in D.C. in 1996.

"The longevity of this phenomenon at the state level," the researchers write, "provides evidence that peak car travel in the U.S. is a more permanent phenomenon than previously thought."

By 2011, the last year for which they gathered data, 48 of 50 states had peaked in miles traveled per capita. The two outliers: Alabama and North Dakota, where an energy boom has made the state a national exception on many fronts.

The state data Garceau, Atkinson-Palombo and Garrick have gathered raise interesting questions about why driving patterns differ from one state to the next and how those patterns are connected to the economy. But first, a quick tour of the data.

Here are the earliest states to experience peak car, starting with Washington in 1992:

Rural states like Wyoming have seen some of the biggest swings, while change in California has been much more modest:

Even Florida peaked before the housing bust and the recession:

Mississippi has been on the decline since 2008 after a stunning growth in driving over the previous 25 years:

And our two outliers:

Over this same time, Garceau, Atkinson-Palombo and Garrick also note that car travel has decoupled from economic activity in many parts of the country. For most of the last century, nationwide, personal wealth (in GDP per capita) and car travel (in VMT) have risen in tandem. The relationship between the two has traveled in both directions: Our economy has in part been powered by transportation, by the movement of goods, by farmers hauling their produce to market, by logistics. Meanwhile, the more wealth we've had, the more we've driven, as families could afford multiples cars, new homes in the outer suburbs, and more family vacations.

Some of the earliest states above to experience peak car, though, illustrate that car travel can decline even as the economy grows. Washington, Oregon and Utah have experienced this. The below chart illustrates that they are now effectively getting more GDP per mile driven than 20 years ago. Alabama, meanwhile, has experienced the reverse:

Over the 2000s, 30 other states in this data no longer appeared to have any significant relationship between VMT and GDP. Why? One potential explanation is that our economy is increasingly powered by sectors, like tech, that rely less on transportation. Or maybe we're interested in spending our wealth in different ways today.