Benefits in user-pays driving
About 5% of Australia’s GDP is spent transporting goods and people. However, many aspects of road transport cause considerable concern such as the negative externalities of congestion for commuters, carbon emissions and the external costs of traffic accidents. Australian roads are also a large user of public resources - in 2005/06 $11b was spent on them. They are also subject to considerable taxation – fuel excises in 2005 were $10b and registration charges were considerable.
The policy issue we discuss in our report* is whether replacing current road use taxes by targeted user charges (congestion tolls, distance-mass charges) would yield better economic and environmental outcomes. Charging user costs would produce more efficient use of existing roads as well as cost-recovery. Furthermore, a better road system may be constructed using data on user charge revenues just as sales (or lack thereof) guide manufacturers about which products to produce. Implementing user charges is technologically feasible due to recent improvements in telematic devices (in-vehicle boxes with GPS capabilities). These can provide for transport the gains that mobile phones have provided in other areas. The question is whether we are willing to pay by different means for driving in a more environmentally sensitive way on less congested and safer roads.
The largest current road use taxes are the fuel excise and motor vehicle registration charges. Currently most road costs are more than met by these charges. However in the main, these taxes, do not act to affect the negative externalities from road use. Registration charges - while scaled to reflect road damages – do not reflect distances travelled on roads of differing durability. Fuel excises do not significantly reflect congestion or most environmental effects or road damages because such costs are location, distance and vehicle mass-related. Except for carbon emissions, fuel excises do not proxy well for user charges. Furthermore, a carbon tax specifically on fuel is not needed once the ETS is introduced. In any event a $20/tonne CO2 charge, twice that proposed by the government, implies only a 5 cent /litre charge on unleaded petrol.
Congestion externalities are not well captured by fuel excise since such externalities are time and location dependent. Their implied costs are huge – $3.5b in Sydney and $3b in Melbourne annually. But they are small in Ballarat and Dubbo – yet the fuel excise is applied to fuel consumed everywhere.
In moving to user charges to deal with congestion and environmental externalities, there are arguments for comprehensive reforms (electronic pricing based on GPS) and partial reforms (cordon pricing of CBDs plus pricing of major ring-roads and arterials). In our view comprehensive reforms make sense partly because of their increased effectiveness, the substantial costs of partial reform and the adaptability of comprehensive reform technologies. Partial reforms are not necessarily cheap as shown by the high costs of the London pricing cordon (70% of revenues). These are likely to be even higher in Australia where traffic volumes are lower and city structures more complex.
However, with a lack of political will there might be a transitional case for using inexpensive partial reforms (e.g. parking policies, limited cordons, pricing major roads) but for moving quickly towards comprehensive reform. Parking policies are a useful short-term surrogate for comprehensive urban reforms because drivers are ‘used to’ them. Though they are imposed only on terminating traffic, anyone considering a trip to the city knows they affect decisions on when and whether to drive. Parking should be priced so that markets for parking clear with supply equal to demand so anyone, anywhere can park almost immediately provided they pay charges. Subsidised on-street parking should end.
But comprehensive reform is a preferred alternative as it is now practical to use telematic devices to implement road user charges that specifically address externalities such as road damage and those from traffic accidents.
Our estimate is that road damages cost Australia $4b annually. These costs almost entirely arise from heavy vehicles. Currently damage costs are recouped via registration charges. While heavy vehicles pay higher charges, these don’t depend on actual loads, location or distance travelled apart from outright bans on certain heavy vehicles using low durability roads.
It has been proposed to use mass-distance-location pricing so to charge heavy vehicles by the actual weight (using weight censors) carried on the roads actually used (using telematics). Such user charging should be introduced. Only conservatism prevents this from happening. At a minimum the cautious ‘incremental pricing’ policy – allowing use of low durability roads when extra fees are paid – should be implemented.
Telematics can also be used to reduce the negative externalities from traffic accidents. In the US (no good estimates exist for Australia) around 70% of traffic accidents involve another vehicle. If the average damage per vehicle is D social damage is 1.7D, an externality since drivers only account for average damages D. The scale of such externalities is vast – internalising them in the US has been estimated to require a 200-400% premium on insurance policies.
A feasible alternative way of internalising traffic accident externalities is to charge insurance using driver characteristics and distances travelled (using telematics). This is a commercial reality with many companies offering such options. For the average motorist insurance costs would be 2-6 cents per km implying substantial savings for good short-distance travellers and hefty premiums for drivers with poor records who drive a lot. This is how vehicle insurance should operate.
Note, though, because it is driving rather than fuel consumption that creates congestion and increases the likelihood of traffic accidents, while there may be grounds for reduced excises on alternative fuels on environmental grounds, these discounts should not be large.
Introducing the user charges discussed so far would improve the use of existing roads. However, they also generate information that could be used to produce a better designed road network. Current road planning in Australia is driven excessively by engineering/safety considerations. However, Kenneth Small and Clifford Winston have shown that, in the US, switching from current to economically optimal road durability on interstate highways cuts maintenance costs by 40% with, if anything, enhanced safety.
What are sought are road designs in terms of capacity and durability which optimise the present value of roads as capital assets. This is complex because CSO’s and indivisibilities related to minimum road size need to be addressed. In addition, roads involve significant network uncertainties. Furthermore, roads are local monopolies so there are regulatory issues.
The methodology is to forecast efficient user charges given projected demands by user type. Ideal road investments should return costs with appropriate competitive return on capital. The information collected via telematics would considerably improve on the information currently available. Then, in principle, road services can be delivered as efficiently managed capital assets.
Though we support introducing user charges, this does not mean automatically removing the fuel excise. The traditional argument for fuel excises is as Australian fuel demand is very price inelastic, the excise is not very distortionary. Furthermore, these taxes are relatively stable, cheap to collect and difficult to evade.
Recent work in economics suggests the validity of the traditional arguments turns on whether fuel is a tax base alternative to untaxed leisure. This is untrue if fuel and leisure are substitutes. Evidence on this is mixed with some work suggesting fuel is complementary to leisure i.e. we tend to drive places for leisure.
We have estimated optimal fuel excises from the perspective both of revenue yield and their ability to approximately internalize congestion and other externalities. Our estimates considerably exceed the current tax (which has remained unchanged for several years) of 38 cents. An excise solely designed to deal with externalities is estimated to be about 48 cents. However, as we argue earlier externalities are best dealt with user charges. There is a case though for levying substantial fuel excises irrespective of environmental proxy arguments.
However, there are less convincing arguments for some smaller taxes relating to road transport. There is no economic reason for the luxury car tax – why isolate this specific luxury? Equity considerations are better targeted through the income tax/transfer system. Furthermore, the 10% tax on imported cars should also be removed as there are no strong economic arguments in singling out imported cars for taxation either.
To conclude technology has improved such that transport taxes should become more like user charges that reduce the negative externalities from driving. The information can be used to assist redesigning the road network so to yield further efficiency gains. In addition, there remain strong arguments for retaining fuel excise for raising general revenue rather than more distortionary taxes on labour.
Australia is a big country with a dispersed, small population. There will be productivity gains for industry and lifestyle conveniences if road transport operates more efficiently.
Harry Clarke and David Prentice were commissioned by the Australia's Future Tax System Review, to provide a Conceptual Framework for the Reform of Taxes Related to Roads and Transport.