Taxing congestion: Is it helpful?

A number of fine authors have come in behind taxing congestion today – namely Greg Mankiw and Stephen Dubner.

The justification for “taxing congestion” appears to be Pigovian – someone clogging up the road has a negative externality on everyone else, and so we should tax that externality. However, I feel that this is just half the story.

In the case of congestion, everyone else on the road is also holding up that one person. In fact, on average, one person on the road is suffering the same negative externality as they are providing. As a result, doesn’t the existence of congestion effectively cancel itself out? Sure putting a toll on will reduce congestion – but if we already have the optimal solution why would we want to introduce a tax on top of it. Note: It may be efficient to actually have some congestion, as the goal of policy is to maximise welfare – not minimise congestion.

Now I have made the argument for an externality in the past (here and here) – my thinking was that the externality fell outside of the drivers and on other areas with which driving was a means to (eg work). Of course, I can’t think of a single situation where there isn’t a “price” mechanism to sort this out (eg with work people wages will adjust to sort out the optimal labour market solution in the case of the “externality”).

I would like to hear if anyone has an externality justification for toll roads – bonus points for using the term “non-linearity”.

  • Pingback: Taxing congestion: Is it helpful? | TVHE | dankles.com()

  • steve

    I wouldn’t have thought it is a question of the externality at all, because of exactly what you describe, that it cancels out. However if we think about it differently in terms of the market for road use, then additional quantity may justify an increase in the charge for road use (beyond standard petrol taxes and road user charges). In this case the supplier of road services provides the additional quantity at a cost.

    for example the toll on the northern motorway extension, or on transmision gully in Wellington. This is how a toll should work and I would forget about the pigovian approach and externalities because people already pay for these when they experience congestion themselves.

    Tolls in NZ though will be severely distorted because of the govt which as said it can only go to paying for certain things untill the road is paid off. It ignores the fact that there is still some investment there, and it still has an economic value and therefore all roads can justify a toll.

  • “However if we think about it differently in terms of the market for road use, then additional quantity may justify an increase in the charge for road use”

    100% agreed – as a way of financing a road, or road maintenance, a toll makes good sense.

    However, my goal here was solely to explore the use of a toll as an externality correcting device – as those other authors have suggested.

  • Pingback: Taxing congestion: Is it helpful? | TVHE | helpfulname.com()

  • moz

    The AJAX editor she borken!

    Spaces come up as %20 afterwards

    Agnitio: hmm I just used it and it works, wonder what is going on

  • moz

    What about the costs imposed on non-motorists? Too often the pretence is that only motorists matter. You’re also ignoring the existing road toll (it’s lives and they’re not dollars so perhaps they don’t matter?)

    Congestion charging will benefit (for example) some schoolchildren who will now not be driven to school and others who will now not be killed by motorists. It will also reduce pollution intensity especially around the inner city and that benefits anyone in the area. The health system will benefit from reduced demand – initially through fewer motorists crashing into people but later through a healthier population (less car use means more active transport). Arguably there will be an even later cost as fitter – healthier people live longer in the medically intensive part of their old age. People who already use active transport and public transport will benefit from less motor traffic – congestion%2C pollution and amateur drivers. And so on…

    I’m using “benefit” here meaning “an imposed cost is removed”.

    (oooh, I do like the “five minutes to edit your post” thingy. That’s excellent. Well, if it worked it would be excellent. Bugger.

  • “Congestion charging will benefit (for example) some schoolchildren who will now not be driven to school and others who will now not be killed by motorists”

    Does congestion cause accidents? If so I’m not sure that congestion related accidents are particularly costly – as they must be involve very slow speeds.

    “Arguably there will be an even later cost as fitter, healthier people live longer in the medically intensive part of their old age. People who already use active transport and public transport will benefit from less motor traffic – congestion, pollution and amateur drivers”

    Firstly, some of these “externalities” are already internalised – a persons decision to be fit benefits themselves, not some uninvolved third party. However, I definitely agree with you that there may well be other externalities from the use of motor vehicles – but the congestion charge has been suggested as a way of dealing with a congestion externality. However, this specific externality may not actually exist – and so can’t really be used to justify a toll.

  • moz

    Um, congestion definitely causes accidents. You’re probably thinking “stuck in traffic doing 5kph” when much congestion is more “battling through crowds to drop the kids off at school”. In NSW they’ve been running a campaign on and off trying to persuade parents not to park in the no stopping zone outside the school, drive up onto the grass verge, make illegal U turns and so on. See, *their* kid benefits from being delivered right to the gate because that way they’re less likely to be run over by another parent. Admittedly only a couple of kids a year die that way but that’s apparently enough to worry some parents.

    The congestion externality per se seems to me to be an outcome-oriented argument: congestion charges have been seen to reduce congestion, therefore if we want to reduce congestion that is one proven way to do so. It’s better than the pave-the-earth model of building more roads, which is proven not to work but is still strangely popular (look up “induced traffic” for one reason why).

    The motorist lobby irritate me with their lies. All the “build more roads where there are already roads” reasons fall apart if you look at them. It has almost nothing to do with effective or efficient transport and everything to do with their self-image as motorists and umwillingness to see that driving is often a poor choice[1]. Otherwise they’d be all for things like public transport and freight rail instead of vehemently opposed. I mean, every person who walks to work is one less person gumming up the roads, so you’d expect active transport campaigns to be at least considered on a cost-benefit basis compared to building more roads… but that doesn’t seem to happen. Likewise, the “congestion cost” never seems to get applied to public transport – in Melbun there’s been vehement arguments for building more motorways because of “congestion costs” but the same people are equally vehemently opposed to that analysis being done for public transport. Their reasons in both cases make the same amount of sense when applied against them 🙂

    [1] one of the classic disproofs of man as rational economic actor is the cost of motoring. For instance, many people would find it cheaper to use taxis than to own a car.

  • moz

    > a persons decision to be fit benefits themselves, not some uninvolved third party

    Well yes, by definition. However, a person’s decision to be fit also benefits a host of involved third parties, like all the taxpayers in the country they live in who pay lower taxes as a result of their diminished health care costs (which of course reduces GDP in the short term and thus should not be encouraged – economic growth at all costs).

  • My non-thought-through intuition is that it is to do with sorting as well as externalities. Externalities because more cars on the road means more congestion for users. But sorting is where it gets interesting.

    For some people a congestion free road is very valuable, but for some people its not so valuable (using the un-tolled (congested) road is not very valuable at all, but still more valuable than 0 – so they use it when it is un-tolled and help clog things up). However if they didn’t use it they wouldn’t be much worse off, but the people who really value using the road in an uncongested state would be much much happier (and would happily pay the toll in exchange for the road being much less congested). Imposing a toll gets the people who value the road not much off it, and gives the people who really value the uncongested road the ability to realise that value. The net benefit is positive and society is better off as a result?

    Feel free to impose some rigour to this stream of consciousness rambling….

  • “Um, congestion definitely causes accidents. You’re probably thinking “stuck in traffic doing 5kph” when much congestion is more “battling through crowds to drop the kids off at school””

    I think the type of congestion they are looking at taxing is the “stuck in traffic doing 5kph” congestion – they are looking at speeding up travel, not slowing it down.

    “The congestion externality per se seems to me to be an outcome-oriented argument: congestion charges have been seen to reduce congestion, therefore if we want to reduce congestion that is one proven way to do so. It’s better than the pave-the-earth model of building more roads, which is proven not to work but is still strangely popular”

    As I said before – there are other externalities that we can hit. Usually we hit these with a petrol tax. However, the argument for a direct congestion externality is a bit more tenuous – we need a “better argument” before we can come up for this type of justification for tolls.

    “However, a person’s decision to be fit also benefits a host of involved third parties, like all the taxpayers in the country they live in who pay lower taxes as a result of their diminished health care costs”

    The types of externalities that stem from lower taxes is a difficult one – namely because we have to ask “is society paying it through taxes because they associate some arbitrary social value from it”. Before we answer that question it is hard to call lower health care bills an externality (although I often have myself 🙂 ).

  • Hi Dant03,

    If there is some cost to the level of congestion then people who do not receive a benefit in excess of that cost will not be on the road in the first place – we agree here.

    However, you are suggesting that, if we have a substitute road and agents that value different roads differently (which is the essential assumptions), there may be multiple equilibrium between the allocation of traffic. In this case a toll may help push us towards a pareto superior outcome.

    I can see the possibility – but I would need to see the example a bit more explicitly before I feel like I understand it 😉