Washington State vetoes all-EV law due to road usage requirement, but it's not so hard


Governor Inslee of Washington has refused to sign a bill he says he supports, which bans new fossil cars by 2030. He refused to sign it because it ties it to creating a road use tax system for EVs, which he says he also supports, but not in the same bill.

He might be right, but the reality is that having a road use tax system is a pretty trivial thing for the cars of 2030. In fact the Teslas of 2018 could do it with a software update.

Read Washington State vetoes all-EV law due to road usage requirement, but it's not so hard


Just ask people what their odometer is every year when they renew their registration. Randomly audit a small percentage, and hit them with hefty fines and criminal penalties if they intentionally lied about it.

That works, though most people will want their car to just send the numbers in for them -- at their command, with the option to do it manually. However, to break down your miles by the states you drove in, you will need to rely on your car for that unless you want to do a lot of work. And if the car doesn't maintain such an odometer, there is no way to audit that. Now, there is not a lot of need for audit, unless different states charge wildly different taxes. For example, if one state charges no tax I would want to lie and say all my driving was in that state. With gasoline that was very difficult because while one could visit that state to buy gasoline if you lived near the border, generally there is broad correlation between where gasoline is purchased and where it is used.

I'm not sure how much people care about going through the enormous trouble of writing down a five or six digit number once a year. Sure, it could be a very very very minor perk that new car manufacturers could offer, but maintaining all the infrastructure to support it would be pretty expensive on the part of the state, and for virtually zero benefit.

I didn't say anything about tracking it by state. Trying to track that is letting the perfect be the enemy of the good enough. We're talking about a few hundred dollars a year of tax on average. It's replacing a tax that is much less accurate. This doesn't need to be so complicated.

Retrofitting all cars in the state is obviously not cost effective. So you've gotta have something for owners of older cars to use, and just writing down the odometer reading once a year is the best candidate for that.

Oh, I agree with you that tracking by state need not be on the must-have list. Sadly, the states see a big loss of revenue, so they will fight for this. Some states, anyway, particularly states with large populations on the border or who see a large number of out of state cars on their roads. For any one driver, it's not a big deal. For the state, they can imagine it would be a lot of money. Even if they just imagine it they will push it. Anyway, a per-state odometer is pretty easy to do in any modern car, and since every modern car, certainly by 2030, has an app, it can be no work for the car owner as well.

The states also "lose revenue" when people buy gasoline in a neighboring state. Oh well. It's hard to imagine a per-state odometer being mandated on all new vehicles, and even if it was, it would be many years before cars made after that date made up even a majority of the ones on the road.

I don't think they need this. But a lot of them want far more, they want a complete tracklog of every car so they can not just tax by the state but by the road -- higher taxes per mile for certain roads, automatic toll paying etc. We don't want them to have that, but per-state might placate them and doesn't reveal much private data.

People can buy gas in other states, but it usually balances out and in the past they had no solution to it. If you drive cross-country, you are going to pay taxes to the states you drive through. All those boring states people just drive through want their taxes.

Presumably you're talking about a federal mandate, because states that I'm passing through while driving cross-country have no right to require me to have a "smart odometer" or whatever you want to call it. How would they even collect that? I'm going to get 10 bills in the mail, for $5 here and $10 there, after I make a cross-country trip?

I could see them doing it for commercial vehicles. They already have truck stops for them, and this would be *less burdensome (though they'll probably still have truck stops too). But for regular passenger vehicles passing through a state infrequently? No way.

I'm opposed to it. I think it's a dumb idea that'll cost way more than the tiny benefit it provides. And I also think it'd probably be a slippery slope that leads directly to the kinds of systems you're saying you oppose, once it becomes clear that a crippled system that can only report which state you drive in is fairly useless.

Would be in the new cars for the states that require it to sell the car there. It's a software feature it has no cost. States would do a treaty collecting and sharng. Or maybe they wouldn't. The alternative is to tax the vehicles registered in their state, or to tax the electricity in the charging stations.

If it's only in new cars that require it to sell the car there (1), then it doesn't solve the problem of cars driving through a state. And what about all the cars that aren't new?

"Treaty" (interstate compact?) collecting and sharing can easily happen through the simple "report your odometer reading once a year" method plus some statistical analysis and/or self-reporting, though the state with the short end of the stick isn't likely to agree.

Software doesn't write itself for free. I don't understand how you can imply that.

(1) which probably still isn't constitutional. California has a special exemption that allows it to have its own emissions standards, but for the most part states can't impose rules like this on car manufacturers. States can also regulate dealerships, but regulating dealerships to require a feature to be in the car is probably not feasible, and it's easy to get around such a regulation anyway by taking possession in another state. States can, of course, pass lots of rules on their own state residents. See below for that.

The alternative is to tax the vehicles registered in their state,

Yep. That was my suggestion that started this thread. Just ask people their odometer reading once a year, when they renew their registration. All the complications of having smart odometers would have little benefit even if they worked perfectly, and they won't work perfectly or anywhere near perfectly until nearly every car in the country has one.

or to tax the electricity in the charging stations.

Much less accurate than the simple method I provided in my first comment.

Correct these regulations are all being written with phrases like "All cars sold after 2030 must..."

So it's only new (or far future) cars.

That should read, "If it's only in new cars sold in states that require it to sell the car there (1), then it doesn't solve the problem of cars driving through a state."

But yeah, the fact that it's only new cars means this is useless for decades (unless you want to treat new car owners differently, which I don't). And in a few decades we'll likely have much better solutions.

We could switch to self-reporting odometer readings and taxing based on that within a few years. Maybe keep the tax on diesel, and give diesel drivers a different calculation or an exemption, in order to cover the majority of road damage caused by vehicles not registered in the state.

It's not at all clear to me that paying for roads by the mile makes sense in the first place. For small roads with primarily only local traffic, they should probably be paid for with property taxes, at a fixed (non-ad-valorem) rate regardless of use. For large motorways, tolls work okay, and toll collection is getting more and more streamlined over time.

Maybe it makes sense to charge based on usage for the many roads that fall in-between the local road and the limited access highway, though it's not clear that this is all that much better than just paying for it with property and sales/income taxes. (Alternatively, you could just build toll collection into the numerous traffic cams that are already all over these roads, if you really wanted to have people pay-per-mile. I don't like the cams, but they're already there.)

Building roads is usually not paid for by the use taxes, but out of general funds. Use taxes are for maintenance, which is correlated with use (mostly with weight and use.)

Maintenance is correlated with use (mostly weight and use), therefore we should tax drivers by the mile? I'm not convinced.

I think the owner of every lot in my subdivision should pay the same amount for the maintenance of roads in the subdivision, regardless of use (so long as use is reasonable, which restrictions against things like running a trucking company out of your garage help ensure).

It would be a mix. You could price it many ways, and indeed, the mere existence of the road is of value to those on the network, and they certainly should pay for its construction. Wear and tear however is based somewhat on time (which again, all who use it at all should pay) but mostly on use, and then mostly on weight. Crazy on weight, in that road wear is proportional roughly to the 4th power of weight.

Of course, if I don't use the roads at all because I have a VTOL and take no deliveries or mail, perhaps I need not pay.

I don't see why it matters what wear and tear is based on.

If you don't use the roads at all, directly or indirectly, and you don't care if they continue to be maintained or not, sure, you shouldn't have to pay. In an ideal world, all roads would be privately owned, and you wouldn't have to pay for anything that you don't use or agree to pay for. But I really don't see how that has anything to do with whether or not roads should be paid for by the mile.

I definitely will charge more to people who are driving 40 ton trucks and tearing it up than I charge to people on motorcycles, and the more miles those trucks drive, the more I will charge them. Why would I do otherwise?

Because it's not feasible.

As I said above, the people who live in my neighborhood don't run trucking businesses out of their garages. The only trucks on these roads are the ones making deliveries. The roads get "torn up" maybe once a decade. Most don't even get "torn up" for much longer than that. You want to charge truck drivers $0.05 per delivery to cover the $4 per residence per year spent filling potholes?

If the roads were privately owned by the neighborhood (which isn't completely unreasonable; this is true in some places), I'd expect that every property owner would pay for them. We wouldn't force truck drivers to install a tracking device or put a weigh station at the entrance, or anything like that.

Most residents benefit roughly equally from the roads. Those that don't like paying the same as their neighbors can move somewhere else. I see no reason to not charge every address an equal amount.

For the roads that do see a lot of through trucks, yes, we should charge trucks more. We still can't be exact, unless we weigh the trucks constantly, but we can get close enough. A lot of this can probably be done through self-reporting and truck stops (weigh stations). Truck stops (weigh stations) are getting less and less burdensome. The trucks can be weighed without even stopping.

For cars, it's such a small amount compared to trucks anyway, so we can be less exact. Maybe it's small enough that we don't even have to worry about it at all. I'd like to see the numbers.

For motorcycles, it should probably be free in terms of per-mile charges.

If you charge the trucks for heavy load damage, they pass it on to their customers. Those who get a lot of deliveries in big trucks pay more than their neighbours. This is reasonable. Almost all local roads forbid trucks except deliveries or moving.

Another thing that would happen is that lighter trucks would carry lighter loads. With robotic trucks, that is an easy win. Ten one-ton trucks do much less damage than one ten-ton truck. We want to encourage that.

Trucks know their weight. They can self-report, with random weigh stations to catch anybody cheating. Trucks charge their customers for freight by the pound. They know.

Even for cars, somebody driving a lightweight https://nimbus.green takes a lot less space and does less damage and emits less pollution than somebody in a hummer. We should like the Hummer driver to pay a larger share or be encouraged to switch to a smaller vehicle. People should pay for their externalities, and then the market will sort it out. Or do you prefer regulations?

You're ignoring overhead.

People should pay for their externalities, and then the market will sort it out. Or do you prefer regulations?

In the case we're talking about, how to pay for local residential roads, I prefer regulations against things like running a trucking company out of your residence and having and paying for the roads, including maintenance, with an equal payment from every property owner.

The difference in damage done between different property owners is too small to worry about it, especially after factoring in what portion of the cost of building and maintaining a local, residential road is fixing damage in the first place. I see potholes about once a decade per mile.

Moreover, if everyone drove a Hummer (maybe one of the up and coming electric ones), we could get away with much lower quality roads, which would be much less expensive to build and maintain.

Actually, I have noted that robocars, able to never hit a pothole, could indeed handle lower maintenance roads. So could cars with advanced computerized suspension tackle the more seriously bad roads. And such cars will be available for going off-pavement but it's a long time before we can insist on it for everybody. Certainly the roads used the most will still want to be in good condition, but they are also the ones that take the most damage.

I'm not sure what your point is.

If we're going to put something in all new cars, how about a way for cars to automatically pay money anonymously every time they pass through a toll intersection?

Anonymous isn't happening as long as cars have licence plates and broadcast MAC addresses everywhere they go.

However, I do think the idea of toll based road regulation is interesting. But you don't need to put infrastructure at every intersection to do that. Congestion charging tends to work by having a cordon around the central area.

I think you can do it with less infrastructure than that, just random checks to assure that self-reporting is correct. But that's something I am investigating for congestion management, not road use taxes.

It's up to us what we choose to do with the cameras. We don't have to record anything (for more than a few seconds) unless there's a toll violation. Most likely we'll want to record things for other purposes, but there's no reason we can't throw those recordings away after a few days. Obviously it's possible to track people. That's already true. But it doesn't mean we have to do it.

The payments should be anonymous. At least optionally. There should be no record matching the payer (or the car, etc.) with the payment. The theoretical ability for a group of people working in government with a high level of security access and a whole lot of computing power to secretly violate the law and break the anonymization is okay, though the security on this should be tight, and legal access to the recordings, or to prevent them from being automatically destroyed after a few days, should require a judge's signature and be based on a specific need.

I said every time they pass through "a toll intersection". Not every intersection should be a toll one.

Self-reporting? Sounds like you mean mandating tracking devices.

There are certain approaches that are so efficient and useful that while I would generally agree, I can see some virtue in certain privacy risks, though I broadly agree we should be able to travel the roads without being tracked. But I also think that each square-foot-second of road use (plus second-ton^4 per axle) should be paid for with a market price, as this would eliminate road congestion among other things.

The most privacy protecting way to do that might be to have anonymous payment transponders and measurement devices, but it's not clear that isn't actually a larger privacy risk as anonymity is difficult. It's also a lot of infrastructure.

So another architecture I have been exploring is that you (or rather your computer, this is all pretty invisible to you) buy a route through the road network for your trip (or a collection of route options which can be adjusted with time.) When driving the route you may go through random checkpoints, and only at these checkpoints would you do a transaction to show you bought the right to be there at that time, and if you don't do that, your licence plate is photographed and ticketed. The records of the routes would get destroyed once expired. This can be done with a smartphone, and the infrastructure is a modest number of portable checkpoints. It could also be done without the cell phone but with ALPR for all cars, which is privacy invading if we don't assure that information is destroyed if the photographed car is shown to have purchased the route rights.

I also think that each square-foot-second of road use (plus second-ton^4 per axle) should be paid for with a market price

You can have a free market, or you can have that convoluted way of paying for roads. Not both.

I agree that roads are complex, there is no easy fix for that. Driving across the only bridge over the river at 8:45AM is going to be a more valuable thing to do than driving on a quiet suburban street at 2am. But within the context of those values (which the market should set) I think the sq-ft-second is a pretty simple way to make it into a fairly scalar number. A motorcycle, car and bus will use different amounts of square foot seconds as they cross. The F2S is of course not simply the area your vehicle occupies, but the space it precludes others from occupying both in front of you and behind you, and to the side. Vehicles that can safety travel with less headway need fewer F2S and should pay less. Vehicles that can drive precisely in narrower lanes can pay less if they can find other vehicles to match. Speed is interesting -- the faster you go the fewer seconds you need the space, but the larger headways you need. Traffic engineers study this to find the optimum speeds at which road capacity is highest, and this is based on how many F2S each vehicle needs at this speed.

If the road manager sells to vehicles that need fewer F2S, they can get more cars on the road, so their price will be lower than vehicles that need more. Of course a bus, which needs a lot of F2S, shares that cost among the many passengers so the F2S per passenger can actually be lowest.

All this is what we want. What alternative is better?

I suppose you are also a fan of airlines charging extra for carry-on items, and wish that they'd charge for it by the pound and also charge passengers by the pound (passengers can be weighed on the luggage scales prior to departure).

What alternative is better? Just about anything that's simpler. The system we have is better, even without accounting for electric vehicles, but taxing vehicle owners per mile using their odometer readings self-reported once a year would be even better, as it removes the electric car loophole (and partial hybrid car loophole). I also think that local roads (and local road maintenance) should be paid for with property taxes. I'm not sure if that's how it's done currently or not. Depends on the area, I guess.

I wonder, how should we pay for the workers who come around periodically to trim the trees encroaching on the roadways? This is about the only "road maintenance" that's regularly required on the low-trafgic local roads where I live, and it's not a per-mile thing. I guess you'd like us to base fees for this on the size, type, and number of trees on a property? Same question with regard to snow plowing and street cleaning, and all the other forms of maintenance that have nothing to do with miles driven.

It'd be great if the market would set values, and then we wouldn't have to worry about any of this. Road owners who impose rates based on permanently installing specialized equipment in cars in order to make complicated calculations would simply go out of business as no one would want to use them. But roads arent privately owned, so we just can't do that.

And so no, you don't nickle and dime the small stuff. But this is the big stuff. And in those cases, you see if computers can handle it all, so its not complicated to the user. The user just knows it will be cheaper to ride in the small pod than the big SUV. They probably don't do a detailed analysis. Their phone just tells them what the two choices will cost. The market does the rest.

The big stuff is the eighteen wheelers traveling on major interstates. A typical SUV driving around the neighborhood is nickles and dimes.

Maybe one day no one will own cars, and we'll all ride around in taxis. I hope not, but maybe. At that point then we can discuss your vision of how to pay taxes for the roads. Until then, I'm going to continue to hold that this perfectionist dream is not feasible.

How much in road damage is caused by a single Hummer being driven 5 miles a day, 300 days a year, on local, neighborhood roads? Are we talking about pennies a year? Even if it's a few dollars a year, and I doubt it's that much, that's not going to be enough to cause someone to change their behavior.

I don't think non-commercial vehicles on neighborhood roads are the big stuff. I don't even think the small amount of commercial traffic that travels over neighborhood roads is the big stuff. I think the vast majority of the cost of these local, neighborhood roads is going to be owed whether anyone drives on them or not.

But if I'm wrong, show me the numbers.

I see a report in 2014 of about $230B on road maintenance. (The USA under spends -- one report suggested that there is $500 per car per year in damage from potholes and poorly maintained roads.)

$230B would be 7.6 cents/mile driven or about $920 per year per car, with the average car driving 12K miles/year. The math I don't have is how to apportion it. As I said, big trucks do much more damage, though they drive a much more limited set of roads. But of the share for cars, yes, that Hummer will be doing an order of magnitude larger share than a light car. Definitely not pennies per year. Easily hundreds of dollars. Who drives only 5 miles/day?

My question was about the impact of "non-commercial vehicles on neighborhood roads."

Lots of people drive less than 5 miles/day on these roads. And a very large portion of the maintenance costs of these local roads would be required whether any non-commercial vehicles drove on them or not.

So you didn't even begin to answer the question.

Again, I've seen a pothole in a local, neighborhood road exactly once in over a decade of living here. (And I'm pretty sure it was caused by a neighbor who was running a business involving large trucks out of his house. So even that one pothole a decade likely wasn't caused by something as small as a Hummer.)

Maintenance on local, neighborhood roads that are not open to through trucks should be paid for with flat-rate-per-parcel assessments on local properties. Getting more fancy than that is nitpicking for pennies a day.

As for other roads, I understand that you don't have enough information to calculate the numbers, and I think that if you don't have that information, you can't say what the best system is for paying for the road maintenance.

Tolls work fine for major motorways, and as technology improves, they become feasible for smaller and smaller roadways. For local, neighborhood roads, I've discussed above. So it's really the in-between roads that are the issue, and I think a tax based on vehicle weight and total mileage driven in a year would be easy to implement and reasonably fair. I'll continue to be opposed to mandating special odometers on new non-commercial vehicles unless someone shows me the numbers as to why that's necessary to solve a significant issue.

I have not seen a source of that information because few roads are barred to heavy trucks entirely. They all allow delivery trucks but many forbid through traffic.

I am not saying that flat rate could not be suitable for some roads and yes, it's simpler. In fact, you may be misunderstanding me -- of course it's much simpler and simpler is better. I wrote this in the context of governments who want to track everything we do, and how to keep them happy without actually tracking all we do.

Governments will never be satisfied without actually tracking all we do.

Replacing fuel taxes with distance-based user fees is being tested and tried in the U.S. and overseas.

New Zealand and Germany have a distance-based fee for trucks, and Australia and several European countries are testing their application to passenger vehicles.

In the U.S., states have been testing mileage-based user fees for over a decade. And six years ago, the FAST Act created the Surface Transportation System Funding Alternatives (STSFA) Program, which has given federal grants to states to accelerate testing of mileage-based fees. Since then:

  • Fourteen states have received grants for 37 MBUF projects;
  • Two large regional pilots have been carried out, one by the Eastern Transportation Coalition and involving seven states and the District of Columbia, and the other by the Western Road Usage Charge Consortium (RUC West), which includes British Columbia, testing multi-state and international issues;
  • Two states, Oregon and Utah, established actual MBUF programs with a slow and long-term rollout across their fleets.
  • A third state, Virginia, is planning to launch one in the next year.

The pilot projects have begun building an extensive base of knowledge about mileage-based fees and what it would take to implement them. This ranges from technical viability to interoperability and public acceptance, but also equity and privacy concerns, among many others.

Add new comment