To some, it's the tax that could rescue the country's transportation budget. To others, it smacks of Big Brother.
Whatever its fate, the tax on vehicle miles traveled, or VMT tax, will first face the fears of a public that's increasingly suspicious of government power.
Meanwhile, those familiar with the technology will try to convince the public that if properly designed, it can assess the tax correctly even as it protects drivers' privacy.
Earlier this week, two former secretaries of Transportation joined a group of experts to propose the VMT tax as a long-term solution for transportation funding. Currently, the federal money for roads, bridges and transit comes from a tax on gasoline. But cars have gotten more fuel-efficient, and the tax hasn't been adjusted for inflation, so the system is going bankrupt.
If that doesn't change, greenhouse gas emissions could rise because the funding goes to new roads, transit and other programs that help traffic flow and move traffic to modes that cause fewer emissions.
The two secretaries, Norman Mineta and Samuel Skinner, urged Congress to phase in the VMT tax over a decade.
In a report from the Miller Center of Public Affairs at the University of Virginia, they acknowledged that the public will have privacy concerns about the tax, but "in reality, the infringement on personal privacy need not exceed that already associated with other technological conveniences such as cell phones and credit cards."
Technology and VMT experts agree, but they say the real problem is that the devices can be configured to be far more invasive -- or at least more invasive than people are ready to accept.
Small devices with big potential
In pilot studies, researchers have used small devices to measure how far a car has driven, when and where. Usually, GPS or cellular technology is used to transmit the data to a central computer, which figures out how much tax the driver owes.
That's a perfect system, from the transportation planner's point of view. The tax could be a flat rate -- every American would pay the same fee on every mile driven.
Or the tax could vary, for example to discourage driving on high-traffic highways or encourage driving outside rush hour. And because the tax reflects road use, the more work the road needs, the more money there is to do it.
But just because GPS and cellular technology can do this, privacy experts say, doesn't mean they should.
"If you think about it, you'll realize that your location history indicates where you sleep, where you work, who you sleep with, who you go to business meetings with, where you go to church, what political meetings you attend, what nightclubs you go to," said Peter Eckersley, senior staff technologist with the Electronic Frontier Foundation.
"These facts about people are astonishingly sensitive. And we don't want to build a permanent tracking system for those by accident," he said.
Tracking has been costly for some. Divorce lawyers, for example, have dug into databases of automatic bridge tolls -- such as the E-ZPass in greater New York City -- to get evidence that a spouse was cheating.
Eckersley and others have said that any VMT tax has to be designed with privacy in mind -- and that current technology can do that.
Big Brother installed under your seat?
In the pilot studies, researchers have tested that argument. Last decade, Oregon installed devices in about 300 cars and measured their mileage in-state, out of state, and in the Portland area during rush hour.
To pay the VMT, drivers would pull up to a gas station. As their fuel was pumped, the device would communicate with a central computer. The Oregon fuel tax would get deducted and replaced by the VMT tax.
Officials considered two types of devices. The first was a "thin" one that used GPS to track the driver, then sent these coordinates to the central computer, which calculated the tax.
The second was a "thick" device that used GPS to track driving, but also did the VMT math. When the driver pulled up to the gas station, it wouldn't learn where the driver had motored; it would only know how much to charge him or her. Then the data would be erased. Oregon officials ended up choosing the thick device for the pilot, not the thin one.
In both cases, the GPS was programmed to only record whether the driver was in-state or out of state. It didn't follow specific driving paths.
At the end of the study, only three people had qualms about privacy. But when word got out to the public, suspicions flared.
"It wasn't the privacy issue so much," said Jim Whitty, who directed the VMT program for the Oregon Department of Transportation. "It was more the government-mandated specific device. Because no one can for sure know what that box does. They can imagine whatever they want to about what it could do."
Now, Whitty said, the state's scrapping the GPS idea -- it's going to attach a device to the odometer instead. Of course, that means all miles -- whether in Oregon or not, whether at rush hour or not -- will be equally taxed.
Another study, by the University of Iowa, let people choose their level of privacy. The study spanned 48 states, and drivers had a device that tracked whenever they crossed into a jurisdiction with a different tax rate.
Some people chose the highest level of privacy: They just got a bill, and their driving patterns would escape into the ether. But others chose to see where they were driving and how much they would be charged -- which meant, of course, that this information had to be recorded.
Would lower insurance fees ease privacy qualms?
Paul Hanley, who directs transportation policy research at the University of Iowa's Public Policy Center and helped run the study, said it's possible to erase all record of one's driving patterns, but then it becomes impossible to challenge the government if one thinks the tax was calculated incorrectly.
"So you trade off a little bit of privacy on that, but you gain auditability," he said.
During the study, two participants' cars were stolen. When they -- and the police -- asked Hanley if the cars could be tracked, he said they couldn't.
The study concluded in August. When looking at the results, Hanley found participants "trusted the system, they trusted us that they were not being tracked, they had faith in the system itself ... the participants said, 'We trust you, but how would we trust anybody else?'"
Those butterflies have convinced advocates of the VMT tax that if they're going to persuade the public, they have to do it right the first time.
Paul Sorensen, an operations researcher at the RAND Corp. who has focused on implementing the VMT, thinks the public will only be convinced through credible practice runs that win people over.
The people most skeptical of the government, he said, won't be convinced by pilots that are government-run. So some of the large-scale trials could be run by independent organizations that bring credibility to the project.
Moreover, not everyone would have to get the same privacy level. Some people could choose to surrender some information, to get lower rates on auto insurance or to pay lower per-mile rates.
The most suspicious drivers, though, could have their own special path: Once a year, they could check into a gas station and have their odometer read.
Correction: The Oregon study used "thick" devices in its pilot study to help calculate VMT. An earlier version incorrectly stated that both "thin" and "thick" devices were used in the study.