Why we should raise gas taxes now, but implement road pricing soon.

Ed Kohler @The Deets doesn’t seem to like tolls, he asks: What Should be Done if Gas Taxes Aren’t Raising Enough Revenue? : (He is responding to my interview on Jim Pethokoukis’s podcast.)

“One thing that confuses me about transportation funding is the idea that raising gas taxes is politically impossible, but creating new forms of transportation taxes like a tax on every mile you drive would be feasible. To me, the raising gas taxes seems far more efficient since all of the systems are already in place to collect the taxes. “

Why we should raise gas taxes now:

  1. Road quality is not where we would like it and will only get worse if insufficient revenues are raised to maintain and reconstruct existing infrastructure, as existing roads continue to age and deteriorate.
  2. Raising gas taxes is more administratively efficient in the short run than implementing tolls, especially tolls requiring lots of infrastructure. It presently has lower collection costs. See [1], [2]. There is no guarantee that will remain true as new technologies change the cost of collection.
  3. The gas tax is an ok environmental tax. There are better environmental taxes (but at $0.11/gallon for the carbon price of $43/tC, this seems too small to matter from a demand perspective). The gas tax is a terrible congestion tax however, as it does not differentiate by place, or more importantly, time of day. Yes idling consumes slightly more fuel than engine off, but newer engine technologies are making idling more efficient.
  4. Currently roads are only partially funded with gas taxes, general revenue is a major source. This charges non-users as well as users, and sends no signal about the appropriate amount of roads that should be built or how scarce road space should be allocated.
  5. So, states can (and should) raise the gas tax before using property taxes or general revenue to pay for roads. (For marketing purposes, the cynical politician should of course call it a wholesale tax, like Virginia just did. Thus they are taxing refiners instead of consumers, right? (However making the tax a percentage of sales price rather than of volume increases volatility of revenue).

We do need to be careful that the money gets spent on maintaining the valuable parts of the existing system, not building wasteful new facilities. There is a serious lack of trust in existing institutions.

Why we should implement road pricing soon:

  1. Congestion is a problem. It is not getting measurably worse, but it is not obviously getting better. Even if reduces in the aggregate, it won’t disappear to zero anytime soon. All of it is unnecessary. But behold, there is a theoretically sound and empirically proven solution: congestion pricing.  Almost all economists like congestion pricing. The arguments for it are fairly straight-forward. Congestion delay is a dead-weight loss. If instead of charging time, you charged the monetary equivalent of the marginal cost of travel (the delay you impose on others), that loss could be eliminated. If you took the  toll revenue raised by allocating demand in an efficient way, and used it  to maintain existing infrastructure, and returned the surplus to people (e.g. through lowering some other tax or providing an annual road dividend, (make the rebate as progressive as you like, it doesn’t matter from a transportation perspective)), society as a whole would be better off. There is a huge literature on this. (Sadly, congestion pricing cannot be achieved with only gas taxes (unless we required really small gas tanks in cars that lasted only 15 minutes or so … a truly bad idea). Something that charges by time and place is required. That is a toll of some kind.)
  2. Emissions continue with the Internal Combustion Engine. If gas taxes are replaced with something that is more effective in reducing road use (for an equivalent amount of revenue), it will result in lower overall VMT and thus less gasoline consumption and lower emissions. (And if it results in less congestion, less fuel sales from the time spent idling or driving inefficiently in stop-and-go conditions). Nevertheless, the marginal effect of the gas tax on demand is relatively small. We can see with the huge swing in gas prices in the last decade, VMT was essentially flat. Anything that made the costs of travel more salient would reduce demand. This depends on the design of the system, invisible systems have low saliency. An in-vehicle taximeter would be highly salient.
  3. Electrification is coming. At some point, electric vehicles will become non-ignorable share of free-riders on the road network. EVs do not require less road pavement or cause less congestion than similarly sized ICE-vehicles. See [3]. The rate of electrification is unclear [4], but given federal fuel efficiency requirements, it is not implausible that most new vehicles will be EVs or EV-ICE hybrids in a 10 year time frame.
  4. Road pricing would facilitate charging more for scarce resources (weak bridges, thin pavements, travel at peak times) and more for more damaging users (e.g. heavy trucks on few axles).

The politics:

  1. In most countries with higher gas taxes, those high gas taxes are not hypothecated, i.e. they are not spent on transportation the way they are in the US (e.g. Australia). So the high gas taxes are independent of road maintenance. Some countries (France, Italy, Japan, China) toll their motorways. Some don’t. Some toll trucks (e.g. Germany). The gas taxes in these countries go to general revenue. The tie between usage and payment in the US is to be commended and should be extended, not weakened. It sends better signals if roads (and other transportation) are paid for directly by beneficiaries. (Feel free to tax other externalities separately
  2.  Some organizations in favor of tolling/pricing in addition to conservative and libertarian think tanks include the decidedly un-libertarian  Environmental Defense and Sierra Club, while the politically neutral GAO says pricing could be more equitable and efficient. The argument will be about the appropriate toll to set, but the same conservative road pricing-favoring organizations that are condemned for sponsorship by oil companies also publish reports that advocate road privatization, and we know what will happen to prices if roads are private — they will go up, which will lower travel (and fuel) demand. (That is, profit-maximing tolls are higher than welfare maximizing tolls.) (Whether this is good or bad is another debate).
  3. Will government run, politically driven agencies be able to implement tolling? The jury is of course out, but the evidence is that it is hard, or we would see it more widely. Similarly if raising gas taxes were politically easy, they would already be higher. Minnesota’s gas tax was raised in 1988 and again in 2008, about once every 20 years. The federal gas tax was last raised in 1993 (20 years ago). Congestion pricing in general is more popular after it is implemented than before. There are many experiments going on, trying to figure out how to achieve acceptability. I like the road utility myself.

I have discussed some of the difficulties of implementing pricing before. But eventually, we will have electrified the fleet enough, we will be annoyed enough with congestion, the damage wrought by heavy vehicles, and so on, that we will do something else beyond (or in addition to) the gas tax. We will also have given up the last vestiges of concerns about privacy, and the cost of the technology to implement such a system will have dropped.  Hopefully we will phase it in gradually, to avoid the potential catastrophe of a Big Bang rollout.

We need to think not only about what will (or should) happen tomorrow, but over a two-decade period.

3 thoughts on “Why we should raise gas taxes now, but implement road pricing soon.

  1. Great follow-up. Minor quibble, if the EPA is to be believed on CO2 per gallon of gas, a $43/tC tax would be about $0.38/gal. Using general gas price elasticities and current prices this would mean a ~3-4% drop in demand short term and potentially ~7% drop in a year or more. Not all of this would be reduced travel by personal vehicle, obviously – a lot could come from purchasing more fuel efficient vehicles or EVs, but a decent drop in demand IMO.


  2. Let’s not neglect the congestion-reducing role of public transportation, which is even more starved for funding than road maintenance. Tolls paid by Golden Gate Bridge users pay for a first-class bus and ferry system that has extended the useful life of a 1937 bridge into the 21st Century. Why not use road tolls more generally to make sure people have reliable, convenient, and robust transit options during rush hours?


Comments are closed.