Unscrambling the Toll Road Egg

That’s the title of a presentation I’ll be giving to a seminar run by the Institute for Sensible Transport in Sydney. Registrations are open until tomorrow (Tuesday 1 August) for those with a professional interest. For my readers, in general, here’s a link to the presentation. Please advise if the link works, or doesn’t, as it’s a newish feature in Dropbox I haven’t used much

26 thoughts on “Unscrambling the Toll Road Egg

  1. The link works.

    On your point about toll roads being under used. This is true for the Brisbane toll roads and traffic forecasters have been sued by unhappy investors. But it was only ever partially true for Sydney, though not any more, and never true for Melbourne.

  2. (perhaps I should have said “accurate” rather than “appropriate).

    Bringing the ubiquity of ANPR cameras into the toll debate is worthwhile, since the marginal loss of privacy from increased tolling is likely to be small but a lot of people don’t know (and other want them not to know). The Sydney tunnel debacle should definitely have been a learning experience, and I fear that the privatised rail link in Sydney will become another. Different trains, different tracks, different station layouts… that’s already proving problematic with the Sydney trams, bob help us when it hits our train system.

    Integrated planning… oh how I wish we had it.

  3. @1 I’d welcome some stats on this, as i focused on Qld. My quick search generally suggested that growth has been weak in most cases


    For example, and fairly typical “The following chart shows that Eastlink (Melbourne) actual traffic volumes have been fairly consistently around 60-65% of original (2004) forecast since tolling began. It suggests the forecasts were good at estimating the ramp-up shape, but not so much the overall traffic volumes.”

  4. @John Quiggin

    Stats on EastLink are hard to get as it’s owned by a private company. But this


    Stats on CityLink should be easier to get because it’s owned by Transurban, a listed public company.

    In any case, the Brisbane roads are the stand outs for missed traffic forecasts. The subsequent litigation means the big engineering companies are now extremely reluctant to put their necks on the chopping block and forecast traffic for these roads.

  5. @John Quiggin

    The pricing on Eastlink is not congestion-related but (as with every toll road in Australia except for the Harbour Bridge) uniform through time. Don’t have data but drive on it every week. There are quite heavy morning and afternoon peaks with less (but still considerable) traffic during the day. Traffic seems to have increased substantially over the last 4-5 years at all times during the day.

    The nits who run these roads still believe they are “congestion pricing” even though they levy a uniform charge including at times where there is no congestion. I have never worked out why the State Government cannot devise a contract revision – congestion price and, should there be any revenue shortfall, we will compensate you. I don’t think there would be any shortfall. A price discriminating monopolist generally does better than one charging a uniform price. The nits in Transurban et al should welcome a form of congestion pricing.

  6. @hc

    East link is owned by Connect East, not Transurban.

    As to why the State government won’t allow congestion pricing, the last time the Labor party was in government in Victoria they copped a lot of stick for allowing any tolls on Eastlink after unwisely having said there wouldn’t be any. No way are they opening up that old wound by allowing congestion pricing. This is where the way economists like you and John think about these things and the way everybody else thinks about these things are poles apart. Economists think it is obvious that tolls should be higher when everyone wants to use the road. Everybody else thinks it is outrageous exploitation that tolls should be higher when everyone needs to use the road.

  7. n this case monopoly pricing via price discrimination means charging a higher price when demand is high which is approximately what congestion pricing seeks to do. There is a literature on when the two approaches coincide. The idea is simply to smooth out traffic over the peak periods so there is less socially-wasteful waiting. The gains can be particularly high if there is a bottleneck congestion when you get a morning and afternoon queue.

    Smith’s comment that most people see this arrangement as unjust has an element of truth to it. The people with a high value of travel time (the wealthy) get most advantage from the pricing whereas the “tolled off” poorer motorists lose out most unless they are compensated. Compensations might include better public transport etc but it is difficult to make them distributionally fair. But another way of thinking about it is that even poor drivers might want to make a few very high valued trips – they too, in this case, might benefit from pricing even if most of the time they are “tolled off” losers.

  8. @hc

    Low and middle income people are more likely to make two trips per day in congested traffic, once going to work, and once going home. They can’t avoid the congestion tolls by changing their travel times and most likely live in places that are badly serviced by public transport. The argument that congestion charges efficiently sort out those drivers who place a high value on their time from those who don’t assume substitution options that mostly don’t exist. The argument that you can create alternatives by having better public transport assumes that governments can waive a magic wand and make public transport appear. In practice it takes 10 years to build a train line from planning to completion.

  9. Smith :
    The argument that you can create alternatives by having better public transport assumes that governments can waive a magic wand and make public transport appear. In practice it takes 10 years to build a train line from planning to completion.

    Major roads take place years to plan before construction, the new Sydney Northconnex is part of a plan that was drawn up over 30 years ago.

  10. Link doesn’t work in Microsoft Edge – just get circle going round and round at Dropbox address, but presentation never materializes (gave up after 10 minutes). Works in Google Chrome. (I didn’t try anything else).

  11. @Smith

    I think this is not right. The substitution possibilities involve small levels of switching – around 5% reductions will usually be enough. You can see this yourself when the 7% of journeys taking kids to school don’t occur because of school holidays – the roads clear.

    In addition most trips in the city are not radial trips to the centre to work – they are cross town trips to schools, shopping or to visit friends. These can be rescheduled. Trips to work can be made before the peak and trips home can be delayed until after the peak.

    The argument that substitution possibilities are inadequate or that we need public transport beforehand to accommodate shifts are mainly wrong.

  12. @Smith

    I am with you on this one. That is, one cannot exclude the possibility that under very specific conditions, congestion pricing does improve allocative efficiency (based on individuals’ wealth constrained preferences defined over a large number of physical things and activities). But the converse also cannot be excluded. Hence, whether or nor congestion pricing is a ‘good thing’ is an empirical question that requires detailed location (city, region….) specific studies.

  13. The marginal costs of road use include wear and air pollution as well as congestion. The former is created almost entirely by heavy vehicles: some combination of license fees (trucks, unlike luxury cars, do not spend a lot of time slumbering in garages) and fuel taxes will get close if an electronic spy in the cab logging actual miles is too difficult politically. The more interesting and important omission is air pollution. The costs here are enormous: $>3trn annually worldwide according to the OECD. The damage is done in urbanised areas, not open country, so the social cost will track congestion fairly well. Unlike congestion, the problem is limited in time as most vehicles will be electric in 20 years. How fast the transition happens depends a lot on whether public policy recognizes the social costs and imposes differential pollution charges, as in London.

  14. is this right?

    public money used in partnership with for profit entities is legally censored by “commercial-
    i n-confidence”conditions by mutual agreement and the data necessary to plan public infrastructure is unavailable to planners in the public service?

  15. Smith,
    Toll road forecasting is a problem not confined to Brisbane. Multiple examples in Sydney and Melbourne have also fallen well short of forecast, typically around 40-50% short. See this paper by Robert Bain. Note the graph. Toll road tunnels in Sydney nd Brisbane both went bankrupt.

    Click to access Toll%20Roads.pdf

    The toll roads that were not tunnels did not go bankript, not because the forecsting was better, but because since tunnels are five times the cost per km to build of surface roads, the non-tunnels are less marginal projects, and can “afford” a shortfall.

    IMO toll roads as a “solution” to urban transport have passed their point of no return. Our urban freeways are now so expensive that we should not be building them. Rail or light rail are cheaper, have higher capacity, and fewer impacts. Yet we do not build them because we delude ourselves that the roads we build instead “pay for themselves”. They do not. In % GDP terms Australia has been spending more on urban transport infrastructure than any other country in the OECD. Yet it is not working. We need to manage demand, build the correct type of infrastructure, and use rational means of funding it. Our current policy does not do any of those well.

  16. @Smith
    Many toll roads are both congested and underused – congested during peak hours but underused outside peak hours because traffic is pushed onto ‘free’ local roads.

  17. @hc
    I like to think of ‘congestion charges’ as being a way for those who need or chose to use a road during peak hour to say thank you to those who choose to travel in off-peak, or not at all.

  18. @Ben Aveling

    The difficulty is the realise such compensations. How do you compensate the “tolled-off”? Roughly by providing better public transport but not very precise.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s