Gotta Pay The Toll

The Obama administration wants to give states more ability to charge tolls on existing Interstates:

In a major shift for how governments fund transportation projects, the administration wants to let states charge tolls on interstate highways. A federal ban currently bars states from doing so in most places, but the latest White House push could change that.

Tucked into the GROW AMERICA Act, the White House’s $302 billion transportation bill, is a toll provision that calls for eliminating “the prohibition on tolling existing free Interstate highways, subject to the approval of the Secretary, for purposes of reconstruction.”

It also allows states more flexibility to use toll revenue for repairs “on all components of their highway systems.”

The proposal reflects the growing need for new sources of funding to maintain the nation’s aging transportation infrastructure. But it’s also a slippery slope — any driver knows that once a toll is in place, they become a handy tool for milking motorists. Tolls, for instance, just increased on I-95 and elsewhere in Maryland last year.

One might point out that we wouldn’t have to engage in slippery slope Interstate funding if we more properly funded such construction and upkeep. This is, as much as anything, a biproduct of our unwillingness to consider higher gasoline taxes.

Arguably, though, there is no inherently best way to fund roads among the various mechanisms we have used and some which have been suggested.

The gasoline tax is historically a big way through which we have done it. With each passing year, though, our collections don’t rise sufficiently to cover the spread. It’s also an increasing complaint that these gasoline taxes do not sufficiently take into account the fact that wear some vehicles take on the roads is disproportionate to the amount of gas they consume. One subject of such complaints are hybrids. Which puts us in the odd position of subsidizing them with our taxes and then complaining that those who utilize them do not pay their way and coming up with alternative methods to make sure that they pay their share. Another case is freight trucks, who do pay a lot in fuel taxes and other charges, but many say that are still getting a de facto subsidy because of the sheer weight (literally and figuratively) they put on the Interstate system. d

Gasoline taxes are also imperfect insofar as they are imprecise. They’re not a bad way to fund big projects like Interstates (excluding the below exceptions), but when it comes to individual jurisdictions people can pass through locales without refilling in ways that skew the averages. A lot of people, including myself, complain about the excessive toll-taking in Delaware. However, such toll-taking is at least partially justified by the fact that a disproportionate number of people pass through the state without having had to refill. This is much more the case when it comes to more locally-funded roads. Adding to the fact, of course, that local governments don’t generally collect gasoline taxes and if they tried to levy higher rates people would simply move on to the next town and fill up there, using up the roads along the way.

This is why a lot of our road work is funded by general taxes. The downside to using general taxes is that it’s not the user-fee basis that many people would prefer. The upside is that while most user-based funding would be regressive in nature, general taxes at the federal level typically aren’t. At the state and local level, though, they are. This is, however, less immutable than the use-based taxes which would be much more difficult to make progressive. If nothing else, we could turn more road construction over to the federal government or states could change to more progressive systems of taxation. Not that they would, or should, but at least they could. It would become incredibly complicated trying to index user fees and income.

A more recent proposal is a Vehicle Miles Traveled (VMT) tax. If we trusted the government enough to allow them to put GPS’s in our cars, we could pinpoint almost exactly where each car’s taxes should go! Think of it like a micro-toll! Further, this wouldn’t be very hard to index to the type of car someone is driving so if we wanted to we could charge vehicles by weight. Of course, a lot of us don’t so trust the government, and that’s one problem. A lot of folks on the left like VMT’s because it opens up another revenue stream while a lot of folks on the right dislike it for the same reason. The more revenue streams, the easier it is to tax more. For my own part, I would be open to this (with strict limits on what the GPS can collect) if it were a replacement of rather than in addition to gasoline taxes.

The VMT plan would almost certainly require more federal and less local funding of roads, which opens itself up to the political process in all of its inglory. We could study traffic patterns on the major roads and fund accordingly, but we can do that now and there are still no end of complaints from liberals and conservatives alike as to how the funds are distributed. Further, even Interstates are a local (and state) matter as well as federal and once you get states involved in funding, it’s hard to draw a line that cleanly delineates federal obligations with more local wants, especially when you throw in the (overblown, in my view, but ever-present) concept of supply-induced demand that would prevent a clean look at expansion funding directly following traffic patterns.

All of which brings us to tolls. Tolls are a funding mechanism that, at least on paper, economic conservatives and libertarians (should) generally support. It’s very user-based and makes no attempt at being progressive. The “social engineering” involved is generally pretty minimal. People tend to have mixed feelings about tolls, which makes them more popular than gasoline taxes which seemingly everybody outside of some segments of the left oppose. Which is, of course, why we get more tolls. Tolls have a regressive component built in, of course, and are a general pain-in-the-arse. Yet they thrive most on the blue east coast corridor. Owing in part to the Delaware problem.

Delaware, of course, being the other downside to tolls. Yes, they are justified in collecting some money. No, they’re not justified in collecting that much. But toll booths favor smaller locales with the opportunity to tax outsiders, and that’s the risk they run. The counter is federal oversight, which hasn’t had much impact on Delaware and which the government appears to be increasingly abandoning.

Which, to be honest, I don’t strongly object to. All of the above funding mechanisms have problems, we’ve decided we hate the gas tax, and roads have to be paid for. What does concern me a bit is the part of the proposal that involves variable price tolling. Jonathan Last discusses peak pricing in his article on High-Occupancy Toll lanes:

At first the economists fixated on “peak pricing,” that is, charging a toll during rush hour. But flat tolls were a crude mechanism. What they longed for was a dynamic system that would always reflect the “true” cost of usage. In 1993, two economists at the Reason Foundation, Gordon Fielding and Daniel Klein, proposed a regime of variable pricing: When traffic was light, the toll might be 50 cents; when traffic was heavy, it might jump to $8. Dynamic pricing would force drivers to pay a true price to avoid traffic. The market would then cause driver economicus to regulate his behavior in the most efficient manner.

The creation of cheap, passive Radio Frequency Identification transponders in the early 1990s made dynamic pricing possible. Drivers registered for transponders (such as the E-ZPass system in the northeast, or SunPass in Florida) that were tied to a credit card. Tolls could be collected electronically while the car was moving. With the problem of collection solved, adjusting prices on the fly was easy. All that was needed was a system of sensors at on-ramps and exits to track the movement of vehicles within the network and a computer algorithm that could raise or lower prices so that traffic volume in the HOT lanes was kept moving at some predetermined minimum speed, say, 50 mph. The first HOT lanes in America, on SR-91 in Orange County, California, opened in 1995.

I support variable pricing at least in theory, but start having a problem with it when it’s as opaque as described here. Not that I mistrust whatever formula they’re using, but that it’s hard to use prices to nudge people when they don’t know at the outset how much it’s going to cost before you leave. Raising prices from the hours of 7-10 AM and them from 4-7 PM on Monday through Friday to nudge people to modify is fair and predictable. Adding costs because as it turns out on 2pm on a particular Tuesday there are a lot of cars on the road on that day is adding insult to injury as their blood pressure is rising and they are stuck in traffic. There’s not much nudging to be done when they’re already out in traffic. This sort of thing might be appropriate in cases where there are alternative routes readily available, but in order to map that out effectively you would have to compress a lot of information into quick decision-making process. Marketeer that I am in many respects, I find myself skeptical of that.