Tax the poor so the rich can drive faster

One idea raised at this week’s tax talkfest was road congestion charging.  Economists on both sides of the political spectrum seem to love it, citing London, Manchester and Stockholm as example of successfully implemented schemes.

What is most surprising is the silence surrounding economic arguments against congestion charging (although Gary Banks, Productivity Commission Chairman, warned about the use of taxes to change behaviour yesterday). It is high time something was done about this.

I’ll let Felix Salmon start the inquiry:

For me, the strongest argument against a congestion charge is that there’s a decent chance that it will be a very expensive way of achieving not very much.

Indeed, this appears to have been the case in London where traffic is back up to the level seen before the implementation of the congestion charge.  In economic speak, the long-term price elasticity of demand for road space appears to be zero.

Taxing with the intention to change behaviour works when there are close substitutes.  But if there are no close substitutes, such as the limited bus and rail network in Brisbane, the tax will raise revenue but not change behaviour.  Especially if the trains and buses are full in peak hour when the tax will be in force.

Of course, proponents of congestion charging suggest that revenue would be spent on improving alternative transport options.

The second problem relates to implementation, and can loosely be called a boundary problem.  Where do you draw the line in the city for a charge area boundary? One would expect, as has been the case in London, that land use patterns would be affected.  If the tax works, and people stop driving, retailers relying on passing traffic will be affected, while those just outside the boundary may benefit.   The boundary may very well create its own traffic problems as people drive alternative routes to avoid the charge.

Another criticism is based on the principle of rationing public goods through price (I use the term public goods a little loosely).  As one commenter notes:

The fee is a method of rationing, but why should it be in the form of monetary compensation to the state? There are many other methods of rationing scarce road space available. The reason for utilizing a fee is obviously because it transfers wealth to the state.

Why not charge an individual fee for faster fire and police service?

Another criticism is that congestion charging, while it may be progressive in terms of revenue, is regressive in terms of utility (the poor are made worse off by the tax).  As one researcher in the field recently noted:

A critical assumption in many of the previous analyses of congestion charges is that there only exists a single value of time. This is somewhat surprising since one of the main features of a congestion charge is that it sorts people related to their value of time, given the existence of feasible transport alternatives

This incidence of the tax burden is easily demonstrated in the table below.   In this example a number of commuters who value a trip and their time differently is shown.  In the top panel all commuters except B drive this route and with seven commuters the trip takes 30mins.  In the bottom panel, a $4 congestion charge is introduced, and the number of commuters drops to six, and the trip time drops to 20mins (the figures are not meant to reflect reality, simply to demonstrate the point).

What we find is that the commuters with the lowest value of their time are the ones with the greatest reduction in utility and are the ones who stop commuting due to the charge.  Commuter C, with a high value on their time and a low value for the trip actually starts to make this commute.

So yes, one could summarise this argument as – a congestion tax gets poor people off the roads so rich people can drive faster:

What we also notice from this example is that the tax reduces total social welfare (utility) from 108 to 99.  This means that this group is worse off after the tax (as you would expect).

Making a congestion charge a net benefit requires that government spend the revenue in a way that makes up this welfare gap.  The question is, will spending on alternative transport expansion be sufficient?   I have my doubts.

Also, I can’t quite understand why economists don’t believe that congestion is a good allocation mechanism for road space. Congestion itself is a cost to users that rations the space, and people who commute in peak hour must still be deriving a benefit from their trip.

My personal view is that if one wants to reform transport taxes the Dutch probably have a far better model to replicate.  It is:

Sort of the love child of a congestion pricing program and a gasoline tax, the scheme will use satellite technology* to track every vehicle in the country and charge them per-mile-driven according to a flexible rate schedule. Initially the program will cover just commercial trucks, expanding over time to all vehicles by 2018.

Their’s is also promised to be revenue neutral as it replaces other taxes.  It will be an interesting experiment for the congestion charging brigade to watch unfold.

Tips, suggestions, comments and requests to [email protected] + follow me on Twitter @rumplestatskin


  1. We wouldn’t need to contemplate a “congestion tax” except for the desire of government and business to keep increasing the population. Then they want everyone else to pay for it!

    • Sherlock,

      I agree that increasing the population brings with it problems, many of which I find personally annoying and distasteful.

      But I think the greater part of the problem is not so much increasing population, but increasing population density.

      There are many parts of Australia that would benefit greatly from an increase in population, and Australia in general would reap many rewards from such a move.

      The problem is that every bugger wants to live in the cities.

  2. I agree that congestion taxes are disastrous, and that individual planning is much more effective than centrally planned road management.

    The only thing I disagree with is the headline. The so-called rich will be paying this tax, just like every other tax in Australia.

    • The rich will be paying it but it won’t deter them. The poor, however, will be forced off the roads — freeing up space for the rich. That’s the Australia of today.

    • Yes that’s right in terms of $. But my example shows that in terms of utility (if you believe in such a concept), the poorer (lower value of time) are made worse off by being excluded from using the road by the tax, while the rich (high value of time) are made better off by having their travel times decreased for a ‘relatively small’ fee.

      So depending on how the revenue is spent, we can’t be sure that the scheme in total is even improving welfare, nor the net distributional impacts (if revenue raised dramatically improves alternative transport options, then everyone can be made better off from the taxes from the rich vehicle commuters).

  3. Never doubt the ability of politicians to find new taxes to fund their excesses.

    I dropped out of city life a long time ago and made a living in smaller towns and cities. I’ll never go back even though there is more money to be made there. The lifestyle tradeoff is too horrendous to contemplate.

    I do some fairly long detours to avoid cities and the so called freeways nowadays. Adelaide and Hobart are the only livable capital cities for the working stiff.

    • SabreOfParadise

      Life is all about the trade offs. I rent right next to the Melbourne CBD. I’ve traded a mortgage and a backyard for cafes and commute on foot. Suits me just fine.

  4. It also occurred to me a long time ago that if gov’ts stop improving the roads, bridges and tunnels people will move away from the 7 cancerous sores that are our capital cities, to the great benefit of the rest of Australia.

    • Oopps hit enter.

      Anyways what Im trying to say is upgrade smaller towns like Newcastle to be a large city like Sydney and link with high speed railway.

      This will pull people away from Sydney and help spread the load.

      I know its a very simplistic view but Its the answer to Sydney’s problems.

      • You mean like also if Newcastle was an industrial city with lots of factories and industries so you would have a lot less commuting around anyway

        That’s soooo old fashioned 😉

  5. <>

    Yes, and all cars need to be painted yellow, have beaded seat covers and disabled indicators.

    A congestion tax would be rewarding govt for failing to deliver a workable public transport system.

  6. Congestion is a direct consequence of high immigration, reduce it to a more acceptable level (60k) and all will be fine quickly.Infrastructure/health/education cannot keep up with Big Australia

    • That’s right, it’s all those foreigners taking up OUR road space!

      If it wasn’t for them it would be a 20 min trip from the Blue Mountains to the Sydney CBD.

      Damn foreigners

      • Are you seriously saying than Australia is building one additional GoldCoast every two years ??? (UNi/Hospital/road/housing) in addition to the current infrastructure maintenance.

        Yeah keep joking.

        • I am not seriously saying anything.

          I am making a sarcastic and ridiculous statement to highlight the simplistic and totally unfounded statement of yours that congestion is caused by immigration over 60k per year.

          • yeah ridiculous I suppose that immigrants like myself leave their cars in the garage LOL

            Saying that high immigration do not create congestion (+saturated health care/childcare etc..) is the ridiculous statement.

  7. Rumplestatskin,
    If you have never read “Still Stuck in Traffic” by Anthony Downs (2004), you would love it.
    Congestion happens to be a cost on the other side of the ledger from the “benefit” of agglomeration efficiencies.
    – Santos and Rojey (2002): “Distributional impacts of Road Pricing: the truth behind the myth”; finds that the “discrimination” of road tolls against lower income groups actually depends heavily on where the lower income groups predominantly live and work. If lower income groups live in blighted suburbs close to their main sources of employment and higher income groups travel further, road tolls are not regressive in their impact.
    See also:
    – Kalinowska and Steininger (2009) “Distributional impacts of car road pricing: Settlement structures determine divergence across countries”
    Having said that, I disagree with road charges and believe that roads should be funded by “trip destination” property owners. The City of Austin, Texas, uses an excellent system along these lines, which partly substitutes for local property taxes. There are very sound economic arguments in favour of this.
    But there is a very convincing argument in favour of tolled “express” lanes, in the paper “Variation in the Value of Travel Time Savings and Its Impact on the Benefits of Managed Lanes” by Sunil Patil, Mark Burris, Douglass Shaw, and Sisinnio Concas. The main point is that there is “value” to be captured for the local economy IF a driver can pay extra to get a quicker trip, in a number of situations. The fee can be set quite high. It makes a lot of sense to allow access on this basis by car drivers, to dedicated bus/public transport/high occupancy vehicle lanes.

  8. Besides the sad fact that present govt wastes revenue I support the CONCEPT of a congestion charge on roads. The London style method is appalling. The Singapore method is much fairer. They charge a few dollars for going past a few key points during peak hour. All other road use is free.
    I also strongly support the idea of creating well-designed small cities as an alternative to jamming everyone in OGCPS’s (One Giant City Per State)

    • I don’t think there’s any particular policy at the government or state level encouraging OGCPS – it’s just happened to become that way thanks to history and clustering of commerce. There’s no point building a car assembly plant if your supplier is 200KM away and you could just as easily build there, for example.

  9. wayne from st albans

    the reverse means testing can be averted
    through the roll out of separated
    bicycle lanes – access for all,
    at no cost to the user.
    Nothing fairer than what the Dutch
    have done to their cities.
    Groningen 7th largest city in the NED,
    it has the highest commercial rental
    yields in the country – 60% of
    all transport trips are by bike.
    A pop of 200,000 people, no cars
    in the city centre. I saw so many
    elderly dutch women riding bikes
    to do their shopping – unbeleivable!
    The local government says their reversal of the
    road hierachy is an economic program
    not an environmnetal one. It has
    been so succesful, that now, the traders
    lobby the council for more pedestrian/cycling infrastructure at the expense of motorised vehicles.

    The answers are to be found in the
    Netherlands – bring on your arguements
    against why it cant work here!

    urban designer/landscape architect

    • “the roll out of separated bicycle lanes – access for all, at no cost to the user.”

      Agreed. This is my first prioirty option for spending congestions charge revenue. I also believe that if the revenue is polled into a city transport fund, some rigourous rules about the cost/benefit ratio of spending should be implemented.

      In Brisbane there is a single train tunnel proposed for $7billion. But if I recall, the road tunnel that was estimated to cost $1.2billion actually cost $3billion, so I don’t hold out hope that this is an effective use of funds.

      Seperated bike lanes give high benefits per dollar, and things like rapid bus systems would be good (eg Curitiba
      Would give high bang for your buck and be versitile for upgrading to different fuel sources (easier to upgrade bus engines with hybrid tech, or gas etc).

    • “The answers are to be found in the
      Netherlands – bring on your arguements
      against why it cant work here!”

      Not a concrete argument, but a relevant difference between the NED and AUS is the terrain. The NED is so flat that self-powered transport only needs to overcome friction whereas in AUS significant effort is expended when cycling up hills, and increased danger (especially for the elderly) coming down the other side.

      Not a reason against it, just a difficulty in implementing a Dutch-inspired system here is Aus.

      • Can be fixed with hybrid electric bikes….

        And also getting rid of the silly helmet laws (Cameron has looked at this before on his old blog btw) – but that’s another issue

        As a former motorbike commuter, two wheeled commuting is FAR superior to trains and of course cars.

        Modern textiles keep you dry when it rains, so there is no argument there.

        • I used the back of my head to break the fall onto a concrete bike path. Can’t say I would look forward to doing the same trick again without a helmet.

          • “I used the back of my head to break the fall onto a concrete bike path”
            Would you care to elaborate on how did that happen? If I were to guess I would say too much alcohol or macho!
            I can often see lots of crazy dudes doing crazy things on the wheels. I don’t see a good reason to pay for their action and have to wear a helmet.
            The point is no one stops you wearing one if you want to. I just don’t like Mum and Dad Government pretending to love me and taking care of me.

          • Two questions for you, Jill,

            1. Do people fall off pushbikes for reasons other than alcohol and macho?

            2. If you were to fall off a pushbike for reasons other than above (e.g. chased by dog, striking foreign object, etc.) and had chosen not to wear a helmet, would you be prepared to forego all taxpayer-subsidised assistance for your immediate, and any ongoing, medical treatment and/or rehabilitation?

    • “bring on your arguements
      against why it cant work here!”

      The bogan. It votes, it voices an opinion.

      There is your answer.

  10. The Dutch model is interesting. It won’t take too long before the same technology will be used to issue ‘instantaneous’ speeding fines!

    • If fines are forever more seen as a revenue source, then yes it will be.

      However where we are at now, it is a very small step from GPS Nav’s / Tom Tom, to pilotless vehicles all together.

      To calibrate these to include congestion, road conditions and the known turn-off points of other vehicles, it would be a instant argument for implementation due to safety.

      I.e no one would speed, no one would indicate wrongly, no tailgating, etc etc.

      But it may incur a loss of government revenue.

  11. Dutch system seems too complicated.
    I neither support or not support the congestion tax.

    On one hand I support it as it should convert many to public transport but if the complaints about Sydney’s train system are anything to go by, it wont be able to handle it.

    On the other hand I do not support it as everyone (as in most people) need to get to and from work in peak hours. So it is a regressive tax (of course) with no foreseeable benefit.

    These are just initial thoughts and I am amenable to discussion on the issue.

    • Good points. This I why I found it so odd that the debate amongst economics and policy people seems to be so one-sided. So I wrote about the other side.

      I am a little indifferent myself, but would say the key to success (maximise welfare for all) is they way revenues are spent to maximise alternative transport options.

      • Agree with you both (Senexx and Rumple). Another quality piece, well articulated and really like your preparedness to present an alternate view.

      • Just remembered I wrote here and came back to see if there were any replies and Cam (shorter to type, “way revenues are spent to maximise alternative transport options” is the key to why a congestion tax is a good idea.

        Additionally as a rural resident that only uses the Sydney train system once in a blue moon to get to and fro, I think the complaints are excessive.

        • The trouble with the positive view of the congestion tax is it requires expenditure on public transport fixes before any of the congestion tax is collected.

  12. In Beijing they have an interesting system to deal with congestion. Everyday, certain numbers (last number of your registration plate) are not allowed into Beijing. Not sure if its efficient but it sure is fair hehe.

      • LOL yes that would be a solution. Though perhaps paying the fines might be more cost-effective.

        If only they could use this congestion tax, and others to provide free public transport. With so much fare evasion, and so much money burned by creating our ticketing systems, one really must wonder how much better such an option would be!

  13. I’m not sure of this, but I wonder if its related to dimensions of view. Drawing a line would be a barrier in a 2D world, but in a 3D world you’d step over it.

  14. The Dutch system of satellite tracking will never pass with the Australian public (I hope), because it will simply be abused. They’ll promise it won’t be, but a few years down the track and it’ll be used to issue speeding fines remotely, or remotely limit speeds. It’s a bad idea unfortunately.

    What they really need to do, is invest heavily in public transport, BIG changes to the train system is what we really need. Light rail throughout sydney and a real subway.

    Then, we need to have extremely harsh testing standards for car licences with rigorous training. Such as in Germany. Re-test all current licence holders at their own cost over the next 5 years, under new standards that involve defensive and advanced driving techniques. Fail 20-30% of drivers who currently have a licence. Force them to use the upgraded train networks until they take approved driver training courses at their own expense and are competent drivers.

    That solves both congestion and the safety issues on the roads.

    • Rumplestastskin’s suggestion earlier, of rapid bus transit, is a far better idea than anything that is limited by rails. Rails really are 19th century.
      Curitiba sort of works, with its less-flexible busway network, because massive “compulsory acquisition” of property for “transit oriented development”.
      There is a real problem surrounding free markets and private property rights, wherever any “fixed” route public transport “investments” are concerned. Wealth transfers to the property owners is the inevitable result. Compulsory acquisition of land, as in Curitiba (and in the Netherlands) allows for the maximum possible number of people to “buy in” to the planners favoured location.

      • From one of Colin Clark’s seminal books, “Regional and Urban Location”.

        (Page 231)”….If rail and subway services to the centres of large cities were charged at full cost, including interest and depreciation, two consequences would follow. The employers of the lower paid service workers in the city would have to raise their wages, in some cases reduce the services offered, or move to suburban locations (for example, some of the retail businesses still carried on in city centres). Meanwhile the higher paid salaried and businessmen, who in most cases could not change their workplace, would have an incentive to move their residences closer to the centre (while at the same time having less incentive to reside close to railway or subway stations).

        These movements would have their reflections in the price of urban land. They would reduce the demand for and the price of business land in the city centre, and of residential land in the outer susburbs, particularly land now high priced because of its proximity to railway stations. There would be countervailing movements raising the price of business land in the suburbs and of residential land nearer the city centre; but these would probably be of a lesser order of magnitude. In net effect, the subsidies on rail and subway suburban transport are subsidies to the owners of certain types of land – for which there is no social justification….”

        In case you are wondering who Colin Clark was:

        Investment in roads, (and buses) on the other hand, spreads the “capital gains” over a much wider range of property owners.

      • Fair enough, but the fact is busses must rely on the roads network, are less comfortable than trains, slower, unreliable and cannot carry as many passengers.

        Buses are public transport, trains are mass transit.

        Buses relying on the roads network also means they are subjected to the effects of congestion, which are well known the increase over time. No matter what measures are put in place to counter-act congestion in the short term. London illustrates this.

        Most european cities have an excellent train network, one of my favourite countries i’ve been to is Germany, excellent trains, light rail, and the bus network integrated together. They have shared cycleways and by far the best roads in europe.

        • I agree with all you say. Which is why Curitiba has articulated buses carrying 130+ passengers on separate road space (to avoid traffic), with well separated bus stops and automated ticketing (you can’t buy a ticket on the bus).

          They get pretty good volumes of commutes for far less money that underground rail.

          Brisbane has a few separated busways now which work very well, and we have taken ticket buying off the bus in many routes, so that has help dramatically.

          • Interesting, I’d still be concerned though with removing potential road space for a dedicated bus line.

            If you look at a city like Sydney, it wouldn’t involve widening the roads to create a new bus lane, it would involve closing a road to traffic to “fit in” a bus lane.

            That’s a bit of a concern to me.

          • The main point I am trying to convey against rails and fixed guideways, is that urban development is expected to conform to THEM, which in a free market with private property rights, will result in capital gains to a few, a wealth transfer from the many, AND sub-optimum participation in “location efficiencies” by households and business (because of rising land prices at the “concentrated” locations).
            Buses, on the other hand, can “FOLLOW” whatever is happening in urban development. Even the provision of dedicated busways does not change the fact that buses can do pickups all over the suburbs before driving onto the dedicated busways. To replicate this in a rail based system, requires “transfers”.
            The passionate dedication of advocates to rail based systems in this day and age, even to the extent of opposing far more relevant bus services and starving them of funding, is truly retrograde by every rational standard.
            It would actually make a lot of sense in MANY specific situations in our cities, to rip up the train tracks and turn the strips of land into dedicated busways (with access allowed to private vehicles prepared to pay a HEFTY toll – ref my 8.36AM comment).

      • From a passenger perspective, I much prefer trains purely for comfort reasons. They are more spacious and have a smoother ride, and aren’t affected by other traffic.

        But I can see that busses give much better bang for buck, but i agree with comments below, they need dedicated seperate lanes to make them decently quick. Otherwise, they are the worst of both worlds, you’re stuck in traffic on public transport.

        One thing I personally really love about public transport, is being able to go the pub after work, have a few beers and not worry about drink driving 🙂

  15. wayne from st albans

    disagree somewhat to a previous comment –

    when you have flat terrain – the winds
    can be much worse than the hills.
    NED is well known for its strong winds.
    (thats where the derailleur comes in handy – with regards to hilly terrain)

    Melbourne is flat as a pancake in every
    direction except
    for the Ivanhoe/Rosanna/Eltham wedge.

    fact – Los Angeles has the same pop as
    NED and 6 times the density.
    NED has 30,000km of bike paths,
    LA has 3,000km. The density arguments
    against cycling dont stack up either.
    After writing a report for local government
    documentating the bicycle infrastructure
    of the NED, the only difference between us and them is political will.
    They tend to have a bi partisan approach
    over there to implement good (left)
    public policy.

    The best thing about the humble bicycle
    is that you can be as poor as a church
    mouse and yet you can still have access.

    I even saw the disabled (physical and
    mentally) riding a pushy.

    • “when you have flat terrain – the winds
      can be much worse than the hills.”

      I agree with this completely. In the absence of hills, the wind is a far bigger problem than hills

  16. regarding the Dutch satellite approach what effects would you see on inner city vs outskirt property prices, guess depends on what the fee is..

  17. Rumplestatskin, I know you often appreciate my reading recommendations. Check out Prof. Alex Anas of the State University of New York at Buffalo. This guy and his colleagues are streets ahead of everyone else on modelling urban form and policy outcomes.
    Check out especially,
    2006. ANAS and RHEE. “Curbing Excess Sprawl with Congestion Tolls and Urban Boundaries”
    2007. ANAS and RHEE. “When are Urban Growth Boundaries not a Second Best to Congestion Tolls”
    AND, on THIS page:
    Check out
    2011. ANAS. “Decentralization and the Stability of Travel Time”
    2011. ANAS and HIRAMATSU “The Economics of Cordon Tolling”