Congestion pricing - November 2018
Proposition: "In general, using more congestion charges in crowded transportation networks — such as higher tolls during peak travel times in cities, and peak fees for airplane takeoff and landing slots — and using the proceeds to lower other taxes would make citizens on average better off."
Collaborator credits: we would like to thank Warwick Davis and Dr Leslie Martin for their expert overviews of the results. We also acknowledge the IGM Forum for the original poll question.
Background information
This month we put to the Economic Society of Australia's National Economic Panel the identical question on congestion pricing that was put by University of Chicago's IGM Forum to their US Economic Experts Panel in 2012 and to their European Economic Experts Panel in 2016. This provides a useful indicator of the level of economic consensus on this issue across the continents.
Overview of poll results by Warwick Davis
Three months ago Australia’s population ticked over 25 million – an increase of around 5 million in 8 years. Much of this growth is centred on major cities. The strain placed on transportation infrastructure has hardly gone unnoticed. Barely a week passes without new announcements of road and public transport investments to “bust congestion”. Yet governments of all persuasion are unwilling to make the most of the infrastructure that we already have by considering the introduction of congestion charges.
Overview of poll results by Dr Leslie Martin
The Bureau of Infrastructure, Transport and Regional Economics (BITRE) estimates that avoidable congestion costs Australian capital cities up to $16.5 billion per year, due primarily to private and business time costs. Infrastructure Victoria estimates that by 2030 traffic congestion will cost every resident of Melbourne $1700 per year.
Dr Leslie Martin
Responses (27)
Agree
8
Agree
5
Agree
8
Strongly agree
8
Strongly agree
8
Road congestion charges provide a useful incentive to drivers to either switch from driving to public transport, to switch to driving at a less congested time, or possibly to forego travel by, for example, working from home. Society benefits from reduced congestion and lesser carbon emissions etc. The same is true of plane travel charges. The proceeds could be used to lower other taxes but may be better spent on transport infrastructure, e.g. public transport, which would yield further public benefits.
Agree
7
Strongly agree
9
Disagree
8
This would tax some individuals and transfer to others. It's not clear that people are better off on average.
Strongly agree
10
An obviously sensible policy.
Strongly agree
9
Agree
8
In principle, congestion charges can be an effective means of reducing congestion at peak times. Congestion charges could be introduced in a way that is revenue neutral, replacing some combination of fuel taxes, motor vehicle registration fees, parking fees or public transport prices. Although this probably would make the "average citizen" better off, careful management would be required to identify winners and losers from the policy and ensure a fair transition into the policy. For example, low-wage workers are often afforded less flexibility in working hours and locations and may find it difficult to adjust their road or transport usage in response to a congestion charge. To address this, employers may be able to take simple steps such as setting shift changeover times outside the peak congestion periods. As a longer term response, employers might also relocate businesses in response to congestion charging.
Agree
9
I am not sure whether we discuss the right problem, as a statement this is a double dividend argument - we gain by getting more efficiency into using a scarce resource and have revenue to reduce taxes. The economics is sound. The problem I see is in the political economy of the question ... we do not have a price on carbon for reasons that we do not want "new taxes" - the same debate will start on congestion pricing. Whether the political economy stacks up is the real challenge and there I am much more skeptical.
Agree
8
Agree, but it really depends which taxes are being cut. To my mind a better policy would be for the revenue raised from such congestion charges to be rebated to households in a progressive manner (since the burden of such congestion charges falls disproportionately on those with low incomes who tend to have fewer margins of adjustment).
Agree
7
Agree
9
The question asks whether there is at least one existing tax that reduces welfare more than a congestion tax, a question to which the answer is almost surely yes. A congestion tax is likely to be progressive, as it will be paid disproportionately by those who must travel at peak hours (i.e., people with jobs), although some incidence may fall on low-wage workers who do not have the labour market power to escape the tax. For this reason we should be careful in implementation not to apply the tax to all methods of transport at peak hours - for example, to leave buses out of it, and/or car travel on secondary roads. As with sin taxes, the behavioral distortion induced by a congestion tax is likely to be pro-social: a reduction in costs paid by others due to the behavior that is reduced (though those others will in fact be the same people who pay the tax, rather than second-hand smokers who are not paying the tax). What taxes are likely worse for welfare than a congestion tax? Some likely candidates are stamp duty, personal income tax on any bracket below the top, the Medicare levy. A final caveat though that in implementation, tax revenue often gets earmarked for particular expenditure categories, and this political problem could make a congestion tax less appealing: for example, if the money raised by the tax gets ploughed into building a highway expansion that only the rich really need, rather than spending on welfare programs or extending the subway lines to under-served poorer suburbs.
Strongly agree
10
In the short run and for available capacity, setting congestion charges would better allocate scarce capacity from less valued to higher valued uses for a net efficiency gain. In the longer run, the information on congestion prices and usage would better signal the choice of national productivity investments to create additional capacity. For the case of roads, current special taxes on motor vehicles, including Commonwealth excise on petroleum products, and state taxes, including registration, transfer duty and city parking levies, are poorly correlated with vehicle use of government funded roads, pollution and congestion. As proposed by the Henry Review, and many others, an aggregate revenue neutral reform package would replace these taxes with a road user fee, a pollution tax and a congestion fee.
Agree
6
The key to this is whether the behavioural implications of congestion charging are more beneficial than that of other taxes. In expectation, one would expect congestion charging to reduce congestion and its associated problems (air pollution, frustration). Some other taxes are even better (like status good taxes or financial transaction cost taxes) and some are quite a bit worse (taper rates on benefits, income taxes). On balance, I think the non-congestion charges generally have less favourable behavioural externalities.
Agree
8
In general, targeting the externality directly by taxing it would be better. In some cases, one may need to think of equity. Can richer people access public transport more than poorer people, who will then have to pay higher tolls to commute to work, for example. Would their general taxes be lowered enough to compensate them?
Agree
7
Yes, the case for congestion pricing comes straight out of Econ101.
Strongly agree
8
Congestion pricing effectively involves imposing a Pigovian-type tax on the negative externality of traffic jams, and seems to work well in reducing peak hour traffic flows for example in London, Singapore and Stockholm. The Bureau of Infrastructure, Transport and Regional Economics has estimated the social costs of metropolitan congestion in Australia will rise to around $30 billion by 2030 which suggests it should be adopted in all capital cities. Estimating how high the charges should be, and how much revenue would be raised, ultimately depends on estimating demand elasticities. On what to do with the revenue raised, it would depend on which level of government is responsible for implementing and administering the policy. For instance, if it's city councils, the extra revenue could offset public transport subsidies; if it's state governments, then the revenue could be used to lower motor vehicle registration.
Strongly agree
9
Congestion taxes would be expected to drive a number of behavioural changes that would serve to reduce congestion including switching time of travel, ride sharing, increased use of public transport, changing transport mode to cycling, long distance rail, working from home and employer initiated changes in starting and finishing times, video conferencing rather than face to face meetings...…..The inclusion with the policy of returning any proceeds to the community through tax relief supports the acceptance of the policy and effectively helps compensate those making adjustments which result in gains through lower congestion faced by those who cannot (or will not) change.
Strongly agree
8
Strongly agree
10
The answer is a no-brainer. The interesting point is why congestion prices (particularly for CBD access) have met so much resistance when most people would rarely if ever pay them. In my view, the answer is that most people in the political and business elite work in the CBD and mix with others who do, thereby getting a misleading sense of public feeling.
Strongly agree
9
The potential welfare gains from congestion pricing are very large – counting both the certain gains from de-congestion and the gains available if the resulting revenues are used to reduce the level of welfare-minimising taxes and/or raise the level of welfare-maximising investments. Having led several multi-country studies on the topic and having tabled the arguments and evidence to persuade successively the European Commission, the European Transport Ministers, the Transport and Environment Ministers of the OECD member-countries, et al., to endorse the principle of congestion pricing, I have no difficulty in agreeing with the above proposition with a high level of confidence. But having witnessed very little follow-up by way of decisions to introduce congestion pricing in practice, I have some difficulty in finding any useful words to add to the debate. Rather, let me invite readers to consult my short summary of the argument and evidence in the 13-page document, Roy, R. (2007), A Politician’s Guide to Efficient Pricing, ECMT, Paris, and my fuller treatment thereof in the 31-page document, Roy, R. (2008), “Mind-forg’d Manacles – The Constraints to Optimising Urban Transport Policy”, OECD, Paris.
Agree
7
Agree
8
I believe that higher tolls to reduce traffic congestion is also an instrument to change behaviour. This can be used to improve productivity as new technologies may reduce the need to travel.
Strongly agree
10
Price is a better rationing device for these types of commercial goods and services.