Against the backdrop of high petrol prices, sales of electric vehicles (EVs) and hybrids reached 50 per cent of new car sales in Australia earlier this year. While this was welcome news for the climate transition and has helped some commuters avoid high petrol bills, it was less good news for federal government coffers.
As drivers switch to hybrids and EVs, they buy less fuel and pay less fuel excise. If the EV take-off continues, fuel excise – which currently collects 53.7c for every litre of fuel sold – could fall from just under 1 per cent of GDP today to effectively 0 per cent by 2050.
Policymakers are now weighing options to replace the excise. But it’s worth taking a step back and asking why we tax fuel use in the first place, and what that tells us about the right tool for the job.
How do we tax fuel use?
Since 1929, the Australian Government has imposed some form of fuel excise – a flat tax on every litre of fuel sold in Australia. The primary motivator for the excise has been that drivers should pay for the road they use. Driving more, or driving a heavier vehicle, ‘uses’ more road and, historically, it also meant you burnt more fuel and therefore paid more excise.
But this logic only holds if similar vehicles burn similar amounts of fuel per kilometre. The rise of hybrid vehicles and EVs calls that into question.
For example, a Tesla Model Y EV pays zero excise, while a Toyota RAV4 hybrid pays around 2.09c/km and a petrol-powered Mazda 3 pays 2.85c/km. Over a year, this means the petrol-powered Mazda 3 pays about $340 dollars more than the Tesla EV in the hands of the typical Australia driver.
The result of the take-off in EV sales is not only a potentially inequitable tax arrangement, but a decline in fuel excise revenue. If EV adoption progresses, we estimate that fuel excise revenue from passenger cars alone would fall to effectively zero by as early as 2045.
Is a road user charge the answer?
One of the most commonly floated alternatives to the fuel excise is a road user charge (RUC). In its simplest form, this would be a per kilometre charge applied to passenger vehicles (like the version proposed by NSW), coming closest to a simple like-for-like replacement of the fuel excise.
However, while the focus is often on how best to replace ‘lost’ fuel excise revenue, the bigger question is which costs of road use a new charge should cover.
A useful rule of thumb in policy design is to give each policy tool one goal (the so-called ‘Tinbergen Rule’). The issue with road use is that there are at least 2 broad categories of costs, and with many associated potential goals it is hard for just one tool to do all jobs.
The costs drivers impose on others
The first and often overlooked category of costs is the cost drivers impose on others. If you use a busy road, you add to congestion and that affects other drivers. Your car also burns fuel, which pollutes the air and contributes to global warming. Economists call these costs externalities, because the driver is not the only one who bears them.
Taxing these costs is important because it makes people consider the harm they are doing to others when they drive – it ‘internalises the externality’. But a simple per kilometre RUC is likely not the right tool for this job.
Congestion tends to happen in a particular place at a particular time, such as the M1 at 8am. A flat tax on road use is poorly suited to this job. Place and time-based charges, such as tolls that change through the day, would target this cost directly.
Carbon emissions and many air pollutants scale with the amount of fuel burned, which with the rise of EVs is no longer correlated with road use. Absent an economy-wide carbon price, a fuel excise that is maintained and recalibrated to the size of these costs would better address these externalities than a RUC applied to all vehicles.
These kinds of externalities are a large component of what’s called the marginal social cost of road use – the cost to all of society (not just the user) from driving an extra kilometre.
The cost of the road itself
The second category of costs is the cost of the road itself. Governments pay to build roads and then to maintain and repair them. The fuel excise was largely motivated by ensuring users of roads contributed to these costs. This is also the most commonly cited reason for why we need to introduce a RUC to ensure EV drivers contribute to these costs.
However, the issue with both the fuel excise and a simple RUC is that all cars do not contribute equally to road damage and maintenance costs. Repair costs depend on the type of road, how many vehicles use it and how heavy they are. A B-double truck does far more damage to a suburban road than a Toyota Corolla does to the Hume Highway. This road damage is another element of the marginal social cost of road use.
Generally speaking, road damage rises dramatically with a vehicle’s axle weight (roughly the 4th power). A vehicle with 10 times the axle weight does around 10,000 times the damage. Heavy vehicles already pay a separate RUC, partly for this reason, though it’s imperfectly levied through a charge per litre of fuel purchased.
By contrast, passenger cars and other light vehicles do very little damage to the roads they use. A RUC that predominantly focuses on passenger vehicles is therefore hard to justify on the cost of road damage alone.
However, the cost of building roads and their upkeep remain. Governments can fund it in two ways. Drivers can pay, or taxpayers can.
Should drivers or taxpayers pay for roads?
Under a user (or beneficiary) pays approach, a RUC apportions these costs per kilometre of road use, so those who benefit from using the roads pay for them. Transport for NSW estimates these costs at around 5.61c/km for light vehicles. A RUC at that rate would cost the average passenger car driver around $673 annually, with those who drive more paying more and those who drive less paying less.
The downside of this approach is that the cost to a driver is likely greater than the marginal cost of that additional road use (which is effectively zero). This would lead to less driving than is socially efficient. It may also have distributional effects, as those who need to travel further to access jobs and services will pay more.
To mitigate this, a RUC for passenger cars can be designed as a flat-rate annual access charge for all drivers so that marginal driving decisions are not affected, with a variable per kilometre amount retained for heavier vehicles tied to the marginal cost of their road damage.
The alternative is funding roads through general revenue (i.e. from all taxpayers). This would require raising other taxes to cover the costs.
For policymakers, in addition to the efficiency cost of raising different taxes, there is also an equity question of whether people who drive little (or not at all) should be asked to help pay for those who drive a lot.
Funding through general revenue suits ‘public goods’ like national defence well, but roads don’t neatly fit that definition. Providers can exclude drivers from roads, such as with tolls or through licencing, and on a busy road each extra car leaves less room for others.
How can policymakers proceed from here?
Each policy approach will have its trade-offs, and the most appropriate policy will depend on policymakers’ goals – whether they are targeting beneficiary pays, marginal social cost, or a combination.
One-off road access or kilometre-based charges are relatively easy to implement for passenger vehicles, such as through annual registration fees or odometer readings. This puts policy in the beneficiary pays world.
More dynamic charges, such as those that vary by place and time to account for costs of congestion, better target marginal social cost. However, these are likely politically harder, as road users see a kilometre-based charge as a closer proxy for the excise. It’s an open question whether kilometre-based charging can be a stepping stone to more dynamic pricing.
For all the weaknesses of the fuel excise, if appropriately set it does effectively give a carbon price for land transport, suggesting it has an ongoing role to play.
The High Court’s previous striking down of Victoria’s RUC in 2023 adds a further degree of difficulty for policymakers, mixing in potential federal-state relations considerations to any RUC design.
In designing future road tax policy, we shouldn’t let perfect be the enemy of the good. But governments should at least be clear about what externalities or costs they’re targeting and can feasibly address.
Amy Tramontozzi is a Pre-doctorate Economist at the e61 Institute, with research interests in housing policy and the economics of climate change. Amy completed a Master’s in Applied Data Science from the University of Chicago and a First Class Honours degree in Economics from the University of Edinburgh.
Lachlan is a Research Manager at the e61 Institute. He leads work at the intersection of economics, data, and public policy, with a current focus on immigration, productivity, and firm dynamics.
Lachlan’s work experience spans the private and public sectors, both in Australia and internationally, across a range of policy areas, including social policy, financial markets, trade, and labour markets.
Lachlan holds a Honours in Economics from the University of Melbourne. He is currently undertaking a PhD (Economics) at the Australian National University.