Efficiency Can Be Redistributive

Many economists advocate a “congestion tax” because it improves allocative efficiency. People who value the use of a road less than the total costs (including external congestion effects on other drivers) “should” put off their use of the road.

But moving from one regime (ration using time) to another (ration using price) has distributional consequences.  It’s tempting to think we are talking Coase Theorem, but in fact we are talking Kaldor-Hicks-Scitovsky, which is another thing entirely.

Should Manhattan charge a congestion tax, or just charge Uber?

Published on:
Author: Mike Munger
  • JW Ogden

    I say do it anyway and put the proceeds toward increasing capacity.

  • We already have a congestion tax. It’s called a gas tax. The more time you spend driving, the more of it you pay. The more time you spend in congested traffic, the more of it you pay. Adding a new congestion tax on top of the one we already have is exactly the same thing as raising the existing fuel tax.

    • ThaomasH

      The gasoline tax is poorly correlated with congestion and is better suited as a road use charge and a charge on CO2 release into the atmosphere.

      • Can you help me understand how the correlation can be anything other than direct and positive?

        • ThaomasH

          Some gasoline is used in non-congestion driving and there is not a perfect correspondence between the type of vehicle and the congestion it causes. Why settle for “direct and positive” when we can have perfect (as possible) correspondence between the harm to other drivers and the tax?

          • Thanks. That helps me see what you mean. I still completely disagree, and I still think a gas tax already covers this. Even if not perfectly, the surveillance infrastructure required to implement a congestion tax would be too costly to enable any efficiency gains. I just don’t see the point.

          • ThaomasH

            I do assume that the costs of the sensors would be small relative to the costs of congestion which I think are quite large. But certainly that would need to go into the cost benefit analysis of this are any other congestion reducing measure including a higher gasoline tax, if that is what you are suggesting. If the current gasoline tax is higher that it needs to be as a road use charge and charge on CO2 release on congestion amelioration grounds, then it could be reduced on the introduction of an explicit congestion tax.

          • Actually, I was suggesting that because a congestion tax already exists in the form of a gas tax, and because gas taxes don’t alleviate congestion, it is therefore unlikely that a congestion tax would work, assuming the goal is to reduce congestion.

            I also think it’s rich of city planners to utterly fail at their jobs and then propose that this failure be offset by a new tax burden shifted onto the shoulders a general public who typically has no say or only limited say when it comes to city planning.

          • Sean II

            That calls to mind another important point: in the US at least, a ton of traffic arises from the negative geographic correlation between good jobs and good schools.

            Good jobs are in cities. Good schools are typically not. Leaves parents two choices: 1) live near work and commute for school, or 2) live near school and commute for work.

            As I recall this was brought to us by another neat-sounding idea about how taxes might be used to produce a positive externality.

            How eager are we to give the same sort of people a crack at traffic?

          • ThaomasH

            My contention is that a tax on gasoline in not very well aligned with congestion so a gasoline tax is only minimally a “congestion tax.” by minimally aligned I mean how much would the gasoline tax revenues go up if congestion went up by 10%? Some, sure because the stop and go and more engine time would use more gasoline and so gasoline taxes would go up, but by 10%?

          • But this says nothing about the impact of gas taxes or congestion taxes on traffic congestion.

    • Sean II

      One key point that seems missed here: traffic is already taxing, in a way that goes beyond mere financial penalties. Stop and go driving is a miserable experience. If there were good ways to avoid it, everyone would use them.

      Why don’t they? Because most people don’t choose when to drive. Their work schedule choose for them, to the tune of ten trips per week.

      Any effective congestion tax would thus be highly regressive. It would strike at people who can’t modify their own schedules. Service workers, hourly workers, lower human capital folks who are closely supervised in their comings and goings. (Workers on the high side of human capital already use their clout and discretion to drive off peak._

      If the aim is to reduce congestion, we’d probably get better results offering a tax cut to businesses in proportion as they shift a percentage of their staff to flex time and/or remote work arrangements.

      • This is a great set of points.

        • Sean II

          Thanks. I’m well rested lately.

      • ThaomasH

        Congestion is harmful to divers but a driver can hardly take into consideration the harm he is causing others. That is exactly the reason that we should have an explicit tax on it, to give everyone the incentive to change their behavior. Overall, I’m not sure that it would be regressive, but we could raise the EITC to offset it for low income workers.

        • Sean II

          But drivers aren’t causing the congestion, and most can’t change their behavior.

          That’s why they haven’t responded to the powerful incentives already on the table.

          Tell me: how does a suburban mom with a 6:30am day care drop-off and a 7:00am work report time change her driving behavior?

          When else would you have her get on the road? What’s she supposed to do, hit the day care parking lot at 5:30am and then sit in the car for an hour waiting for the place to open?

          • ThaomasH

            If all the suburban moms and pops were subject to a congestion tax, maybe schools and businesses would have different start times. And over time maybe people would live nearer to schools and work. Who knows?

            Suppose there were a really large tax on bread, so much so that very little bread was consumed, just mainly rice. (A few rich people eat bread.) Now some starry eyed economist comes along and suggest removing the bread tax. There could be many objections such as that it would be regressive (the price of rice would probably rise), but one objection that would be pretty silly is to ask the economist who is going to produce all the extra bread, how will it be marketed and distributed, how much more bread will be consumed and how much less rice? General equilibrium is really hard to do

  • ThaomasH

    A first best solution would be a congestion tax. The way to do this would be equip all vehicles with sensors that record the presence of other sensors and record it. Being near a large number of slow moving or stopped vehicles would be “congestion” and the time spent in the situation and the number of other vehicles involved would be the base for the tax. The tax could be calibrated to include a road use charge, leaving the gasoline tax a a pure CO2 accumulation charge.

    The tax would be optimal because all vehicles would be charged in proportion to the harm done to the free flow of other vehicles. The cost per person would be smaller for buses and carpools than for single occupant vehicles giving people the incentive to shift means of transportation and the time of use and choice of routes. If the charge were equal to the harm inflicted, there would be no reason to use revenues to subsidize mass transit, but could be returned to the owners of the streets as a tax credit.

    If the tax cannot be extended to other vehicles, then it should at least apply to both taxis and Uber equally to prevent substitution between the two modes.