It’s often claimed by road lobbyists that the cost of roads is only a small percentage of what Australian motorists pay in fuel tax, registration and other fees. This is untrue: in fact, taxes and charges on motorists fail to cover the cost to the public of car use. The confusion arises because of the peculiarities of our federal system, where the Federal Government collects most of the tax revenue, but state and local governments are responsible for most of the spending.
Headline-grabbing figures for fuel tax are easy to obtain simply because there are well over 10 million motorists in Australia, so if each motorist spends just one dollar there’s suddenly over $10 million changing hands. To take just one example, in June 2006 the Herald-Sun ran large headlines about the $2 billion a year ‘stolen’ from Australian motorists in GST on petrol. This sounds impressive until you realise it represents less than $200 a year per motorist – a fraction of what it costs to fill a pothole in front of one’s house, let alone resurface a road or repair a bridge, and a similarly small fraction of the $5000 a year one will typically spend to own and run a car.
Tallying Road System Public Costs and Revenue
The tables below outline (to the nearest $100 million) just those costs of motor vehicle use that can be quantified in dollar figures, and the total revenue collected from Australian motorists each year. See the Appendix for more detail on how these figures are obtained.
(Our figures have been updated from time to time to ensure they remain relatively current. Earlier versions of this analysis appear in some PTUA documents. The bottom line, however, is little changed from when we first undertook this exercise in 2004, and found the road deficit was $15 billion per year.)
|Annual costs of motor vehicle use in Australia|
|Road construction and maintenance||26,200|
|Road trauma and damage||16,500|
|Tax concessions for car use||7,500|
|State fuel subsidies||0|
|‘Economic’ costs (externalities)|
|Land use cost||10,000|
|Urban air pollution||4,300|
|Annual revenue collected from Australian
|Petrol and diesel excise (net of rebates)||11,600|
|GST on fuel and vehicles||4,000|
|Vehicle registration fees||7,000|
Sources: Bureau of Infrastructure, Transport and Regional Economics, Infrastructure Statistics Yearbook, 2017. Connelly et al, “The economic costs of road traffic crashes: Australia, states and territories.” (Accident Analysis and Prevention, 2006.) Bus Industry Confederation, submission to the National Fuel Tax Enquiry, 2001. Australian Tax Office, Taxation Statistics, various years. Australian Prudential Regulatory Authority, Half Yearly General Insurance Bulletin, June 2009. Insurance Council of Australia, GI Industry Trends – Motor Vehicle Insurance, 2004-2018. Laird et al, Back on Track, UNSW Press, 2001. PTUA calculations (see Appendix).
We have been careful to include in the ‘revenue’ table all the taxes and charges motorists actually pay for the use of roads. To avoid any arguments over the ways in which motorists ‘pay their way’, we include many items that many would say are not really ‘user charges’ at all, such as GST on car purchase and maintenance as well as fuel, and premiums for all motor vehicle insurance (not just the compulsory third-party component).
On the other hand, figures in the ‘costs’ table have been deliberately kept on the conservative side, and should be regarded as minimum estimates. There are many important exclusions. For example, construction and maintenance costs for privately owned toll roads are not included, even though we count tolls as one of the charges motorists pay for roads. We have ignored the very large sums that were provided in government support for the car industry prior to its demise in the mid-2010s. And we also ignore the economic burden of the waste stream from private road transport, including the over 50 million used tyres generated each year which are notoriously difficult to either recycle or dispose of effectively.
We also leave out the considerable social costs of mass car use, such as reduced neighbourliness within communities, reduced independence of children, and the increase in obesity resulting from lack of ‘incidental exercise’ such as cycling to the shops or walking to the bus stop. These costs are still subject to a lot of uncertainty. According to the road lobby, they are so nebulous that they shouldn’t be counted at all. In any event, we haven’t counted them here.
As transport researcher Garry Glazebrook has suggested, transport costs to the general public can be divided into two categories. ‘Financial costs’ are those which show up directly on financial accounts, as subsidies to road users from governments and those outside the road transport system. On the other hand, ‘economic costs’ or ‘externalities’ represent the burden of the road transport system on the broader economy, including on non-users of roads, expressed in equivalent dollar terms. For example, when traffic noise forces someone to move house or install double glazing, the cost of this does not appear as a direct subsidy in transport budgets, but it should be considered nonetheless.
The table above is very conservative in accounting for ‘externalities’. As we say above, difficult-to-quantify burdens such as health and social effects are mostly ignored. One other that is very commonly cited, but that we have omitted, is the notional cost of traffic congestion: usually estimated as around $15-20 billion a year in Australia. As we explain elsewhere, many international transport experts dispute that the ‘cost of congestion’ is a meaningful notion. Traffic congestion is also often used as an argument for building more roads, despite the fact that this doesn’t actually work.
Of course the road system also has very substantial benefits, else we would not bother having it in the first place. The important point here is that all the benefits of the road system (like those of public transport) accrue privately, to individual motorists and to the purchasers of goods and services delivered by road. A person driving a car does not generate any ‘positive externality’ that benefits unrelated members of the public. There are only indirect ‘financial benefits’ to the general public, in the form of direct payments by road users to the public purse – tallied in the ‘Revenue’ table above.
The Road Deficit
The difference between the cost of motor vehicle use and the revenue collected from motorists is the ‘road deficit’. The conservative figures above result in a road deficit of $23.8 billion a year.
About $5.4 billion of this – the difference between purely financial costs and revenue – is the ‘financial deficit’, the amount by which the general public (whether motorists or not) directly subsidises the use of cars and the consequences of car use. The other $18.4 billion represents the ‘economic deficit’, the economic burden on the general public arising from the use of cars and trucks.
Currently, fuel tax raises around $11.6 billion a year. Thus, if the $23.8 billion road deficit were to be recovered by raising fuel tax, it would have to more than triple, from 41 cents per litre to about $1.25. But as we have explained, this figure for the road deficit is only a minimum. If all of the various externalities were included, and the revenue more carefully analysed, the difference would be much greater. Petrol would have to be priced at over $3 per litre if the full cost were charged to motorists.
The fact that public transport’s operating deficit is frequently castigated in the media and the road deficit is not – despite the fact they work out to about the same cost per passenger – and that motorists are instead led to assume they’re being ripped off, suggests a double standard is at work. In a 2012 blog post, urbanist Alan Davies tries to rationalise this by noting that the financial component of the road deficit is quite small relative to the economic component, so much of it isn’t a direct financial subsidy like that to public transport. But as an explanation this still doesn’t satisfy, because the financial component at over $5 billion is still large in absolute terms. If just that subsidy were eliminated, the extra $5 billion could, over a decade, build at least two extra train lines in every Australian capital city; alternatively, it would suffice to double the service on every urban bus route in the country. (And going on the results of past improvements, it would also generate extra patronage and revenue sufficient to cover much of the cost.)
In truth, the double standard persists because Australia is still a car-dependent society, and planners reinforce this in their decision making despite popular feeling to the contrary. Regardless of whether the cost burden is ‘financial’ or merely ‘economic’, as long as it’s borne by government rather than our own hip pockets, it remains invisible unless it’s pointed out to people. And when it is, it’s usually by powerful defenders of the road lobby status quo, who can appeal to the fact that cars are, after all, quite costly to own and operate, which makes rip-off yarns so believable even if they’re not true.
But even if there were no road deficit, it still wouldn’t necessarily be true that more expenditure on roads or cuts to fuel tax are justified. By buying petrol and registering cars people are not necessarily revealing a preference for road transport. There’s a substantial captive market of people who are forced to buy petrol and pay fuel tax and car registration despite wanting an alternative. They may prefer that the revenue be spent on public transport. (So it should also be obvious that by pointing out the costs of car use we are not trying to ‘blame’ motorists for these costs. It is not the public’s fault that the alternatives to driving are so poorly developed in Australia.)
In the end, it’s not important whether fuel tax and other fees pay for roads or not. They’re just another kind of tax and can be used for whatever we as a democracy want it to be used for. Indeed, we know that improvements to public transport have greater benefits for traffic congestion than building freeways, so public transport improvements can be just as good a use for petrol tax revenue as road works.
Linking fuel excise to road funding is a furphy….It is no more relevant to say that falling excise revenues will put road funding under pressure than it is to say this will put pressure on health spending or the age pension.
—Marion Terrill and Owain Emslie (Grattan Institute), The Conversation, September 2016
Public transport is a State Government responsibility….People pay taxes, petrol excise and speed camera fines to the State Government to fund public transport.
—Cr Ben Clissold, City of Casey, September 2005
Appendix: Detailed Analysis of Road Costs and Revenue
Official figures are notorious for taking years to generate reliably. At the time of writing, the latest period for which all the required detailed figures are available is the 2015-16 financial year. To get a reliable picture of general trends, and correct for extraneous factors such as the Global Financial Crisis, we have also noted equivalent figures for previous years where appropriate.
Generally, since the commencement of the former Howard Government’s AusLink programme back in 2005 there has been a steep increase in road spending by both Federal and State Governments. On the other hand, growth in car travel has been slowed by a number of factors, including increases in petrol prices, generational shifts in travel habits, and sluggish economic conditions that have persisted to some degree since the Global Financial Crisis. This means the extra spending on roads has not been met with as large an increase in revenue from motorists.
Previously published figures for the road deficit (such as our own earlier figure of $16 billion for 2004-05) therefore give a fairly conservative picture of the present situation. This is despite the various small, fairly cosmetic steps that have been taken to reduce subsidies to road transport, such as the abolition of direct fuel subsidies by State governments (see below).
Road Construction and Maintenance
The road lobby devotes much attention in its PR material to the money spent on road construction (which is always inadequate, despite still exceeding by a factor of four the money spent on new public transport infrastructure). Instead of drawing attention to all the costs that roads and cars impose on the community, road lobbyists simply point to a figure for road construction and maintenance, and claim that this is only some small percentage (say 15 or 25 per cent) of what motorists actually pay in fuel tax. The message that emerges is that motorists are being ripped off.
This canard dates back to the Kennett Government’s Better Roads Levy in the 1990s, which was a 3-cent-a-litre surcharge on petrol targeted toward Victorian road construction. When it was introduced in 1993, Federal fuel excise was around 30 cents a litre, so the Kennett surcharge represented around 10 per cent of the total tax on petrol. This made it possible for the road lobby to claim that “only 10 per cent of your petrol tax is dedicated to Victorian roads” – despite the fact that Federal funding of Victorian roads (via State and local assistance grants) far exceeded what the Kennett tax was raising.
The way the road lobby has commonly misrepresented the facts since then is to report only direct Federal spending on roads, which comes to about 25 to 30 per cent of the money raised from fuel tax – yet as will be seen, represents only about 20 per cent of overall public spending on roads.
Motorists pay 38.143 cents per litre in fuel excise alongside state charges such as registration and stamp duty – but they do not see anywhere near that level of income going back into the transport network. At current levels of expenditure, the Commonwealth invests less than one-third of the tax take back into roads and the network.
—“Make motoring more affordable”, Australian Automobile Association, August 2010
Motorists are already hugely overtaxed with a fuel excise take of 38 cents a litre plus GST. Critically, the Federal Government only returns around 12 cents a litre to improve transport, so Australians are being short-changed by 26 cents a litre to subsidise other demands on the budget.
—RACV President Peter Chandler, RoyalAuto, May 2011
Motorists are paying too much tax already; they’re paying 38 cents a litre fuel excise as it stands, and we’re not seeing that money going fully back into investment in roads and land transport infrastructure, so we’d be very concerned…. currently, less than a quarter of what the Federal Government raises in revenue from fuel excise has been spent back on roads and other land transport infrastructure.
—Andrew McKellar (Chief Executive, Australian Automobile Association), ABC radio interview, 8 May 2014
What these statements never admit is that the Federal Government itself only funds a small number of roads: those formerly classed as the “AusLink National Network”, “National Highways” or “Roads of National Importance”. Most expenditure on roads is incurred by the states – even more so in recent years – and local government also accounts for a large chunk. Think of all the little streets in your local area: they’re entirely Council funded (from rates and developer contributions, not motorist charges), with the odd State or Federal grant chipping in. But Federal operating grants to the states, or to local government, are often not classed as road expenditure despite the fact that a lot of the money gets spent on roads.
|Annual spending on roads by Wyndham City Council|
|New road construction||$16,000,000|
|Rebuilding existing roads||$8,500,000|
|Routine road maintenance||$3,500,000|
|Road planning and management||$1,800,000|
|Annual road funding from Grants Commission||$830,000|
Source: City of Wyndham, 2005
It’s virtually impossible to trace a direct money trail from petrol tax to road funding, but that’s not the way government finances work. It’s better to think of all the money collected by all three tiers of government going into a big pot called ‘public revenue’, and all the money spent by governments coming out of the same pot. In the case of roads, petrol tax is collected by the Federal government, but roads are funded by all three tiers of government. In-between are various transfers of money between governments to ensure that everything balances.
The table below shows the offical BITRE figures for expenditure on Australia’s roads by the three tiers of government. Importantly, this excludes road expenditure on behalf of the private sector, which the BITRE estimates at $4,493 million in 2015-16.
|Road Funding Totals ($million)|
Source: Bureau of Infrastructure, Transport and Regional Economics, Road-Related Revenue and Expenditure (2011 update); Infrastructure Statistics Yearbook, 2017.
Figures are in nominal dollars; refer to the BITRE Yearbook for historical figures adjusted for inflation.
The figure for State spending includes expenditure by ‘public non-financial corporations’.
As the table shows, road expenditure by State governments more than doubled in nominal terms just between 2008-09 and 2015-16. But even in 2008-09, the total public spending of just under $15 billion already exceeded the amount actually raised from fuel tax in that year. This excess of spending over revenue has only increased in the years since.
Note that although private-sector road expenditure has been omitted here, its effect on government balance sheets is not actually zero, thanks to ‘hidden costs’ in tax concessions and off-budget deals (such as the one with Transurban bringing forward Citylink payments to cover the widening of the Monash Freeway, the cost of which will be borne chiefly by future taxpayers).
Road Trauma and Damage
The Australian road toll is around 1,400 deaths a year. In addition, over 30,000 people each year are hospitalised as a result of road crashes, and every few weeks the news gives us pictures of houses half-demolished by errant vehicles. As a proportion of population our road toll compares favourably with other Western nations; on the other hand, we score poorly compared to other Western nations on the rate of pedestrian and cyclist injuries due to road crashes. It is likely that the number of people hospitalised each year by cars is even greater than the official figures suggest, as not all road incidents are reported as such.
The Bureau of Transport and Regional Economics in 2000 estimated the direct cost of road trauma in 1996 as around $15 billion. In 2006, an update to this analysis put the cost of road trauma and property damage in 2003 at $17.3 billion, broken down as follows:
|Road Crash Totals and Costs, 2003|
|Fatalities||Serious Injuries||Minor Injuries||Property
Source: Connelly et al. “The economic costs of road traffic crashes: Australia, states and territories.” Accident Analysis and Prevention, vol.38 (2006). Figures for ‘property damage only’ crashes are inferred from the authors’ results.
This ‘direct cost’ figure is the one on which our accounting is based: it is a readily measurable minimum that accounts for the cost of treating crash victims, the cost of repairing property damage, and the victims’ lost income. To be conservative, we have adjusted the ‘fatality’ component of the cost to reflect a 25 per cent reduction in the road toll since 2003, which reduces the total figure from $17.3 billion to $16.5 billion, but we have not indexed the latter figure for inflation. Accordingly, this is still likely to underestimate the current figure – especially for property damage from car crashes.
What are not included are the very substantial indirect health costs of car use. Car dependence is accepted to be a big factor in Australia’s ‘obesity epidemic’ and the increasing incidence of heart disease, diabetes and other obesity-related diseases. Pollution from cars is a major factor in respiratory illness in urban areas. And while successive Federal governments have been preoccupied with the challenge an ageing population poses to the national budget (in large part because of the ‘obesity epidemic’), some health economists now think a greater challenge is posed by lifestyle-related illnesses such as depression and certain types of cancer that affect young and middle-aged Australians. The incidence of these diseases is closely linked to whether the urban environment is conducive to moderate levels of physical activity and to the existence of healthy social networks.
Cars make us sick, sad and dead – how cars are managed is critical to public health – and much more important than we previously realised – we need to repopulate the streets. Car centred suburbs are ‘obesogenic’ (fattening) and foster depression and isolation by discouraging social interaction, walking and cycling.
—Dr Rob Moodie, Chief Executive Officer, VicHealth. Presentation to Planning Institute of Australia, 2003.
The Heart Foundation, Diabetes Australia and Cancer Council have identified heavy car use as a key factor in the overweight and obesity epidemic in Victoria. The health groups’ recommended responses include:
- addressing urban planning to support walkable communities, and
- addressing public transport planning to reduce car dependency.
—Diabetes Australia (Victoria), Heart Foundation (Victorian Division) and The Cancer Council Victoria, media release, 2004
The adverse health effects resulting from climate change, road-traffic crashes, physical inactivity, urban air pollution, energy insecurity, and environmental degradation are linked via their common antecedent of fossil-fuel energy use in transport.
Public policies that encourage a transition to a low-carbon low-energy transportation system have the potential to bring substantial public-health benefits by affecting all these health pathways.
Increased active transport (walking and cycling), with public transport for longer journeys, could have a substantial role in meeting targets for greenhouse-gas emissions and would result in major public-health benefits.
—James Woodcock et al., Energy and Transport, The Lancet, September 2007.
The data show clearly that those living in areas of increasing growth reported high hospital admissions for respiratory problems…. Previous literature shows a strong association between increased respiratory conditions and air pollution.
—VicHealth, Submission to Victorian Parliamentary Inquiry into environmental design and public health, 2012
Our study has shown that not using cars for commuting may be protective against weight gain. This highlights the need for promoting active commuting through public health, urban planning, and transportation initiatives.
—Takemi Sugiyama, Baker IDI Heart and Diabetes Institute, October 2012
Daily active commuting to work is important for reaching recommended levels of daily physical activity. Public transportation users are walking and cycling to stations or stops and between intermediary destinations. Availability of public transport is therefore a determinant of active transportation.
—Sune Djurhuus, Research Centre for Prevention and Health (Denmark), October 2012
The indirect cost to society of all these illnesses is estimated at around $60 billion a year, the largest component of this figure being pollution-related respiratory illness at around $18 billion a year. For the sake of this exercise we have not counted most of these indirect costs, except where others have identified them as specific consequences of car and truck use as in the ‘noise’ and ‘pollution’ figures below.
The Australian tax system provides generous deductions for the use of cars, and to a lesser extent for the purchase or leasing of cars. In most cases equivalent deductions are not available for use of public transport, even for business activities. The tax treatment of car expenses in Australia is so generous that it has induced people to use cars more often than they otherwise would. Until 2011, the most notorious of these perverse incentives concerned Fringe Benefits Tax (FBT), as financial journalists liked to point out:
The most commonly ‘packaged’ item is a car. This can be attractive no matter what your salary. For a start, cars and car parking are subject to a discounted rate of FBT. For example, using the statutory formula method to calculate the taxable value of the car, the FBT rate depends on how far you travel each year. It ranges from 26 per cent for less than 15,000 kilometres to 7 per cent for more than 40,000 kilometres per year. So the further you travel, the lower the rate.
—Marilyn Smith, The Age (Business), 23 March 2002.
Though the former Howard Government continually maintained it didn’t give tax deductions for private expenditure, the Tax Office made clear that both business and private travel counted toward the kilometres used to determine the statutory tax rate under the old rules. That substantial public funds have been used to subsidise personal car use was confirmed in a 2009 report from the Victorian Auditor-General, which investigated the use of ratepayer-funded vehicles in three local council car fleets. In two cases, at least half of all use was found to be for personal rather than business purposes; in the third case, no data were available to distinguish business from private use.
So it shouldn’t surprise anyone that anecdotes abounded of company-car owners going on gratuitous driving expeditions so as to ‘clock up’ the necessary 25,000 or 40,000 kilometres to claim the FBT tax break. A 2007 Senate enquiry found that half the cars on Australian city streets in peak hour were company cars benefiting from this loophole.
It just seemed to us ridiculous when you are facing oil depletion and climate change that you should have in place a fringe benefit related to employer-provided cars that requires people to achieve a certain odometer reading before they can get the next level of benefit.
—Senator Christine Milne, February 2007
By contrast, it makes no sense (then or now) for employers to salary-package public transport fares on behalf of employees, because these incur the maximum FBT rate of 95 per cent.
As of 2011, this perverse FBT rule has been partially reformed so that a flat statutory rate of 20 per cent applies to new company car leases, irrespective of distance travelled. This may help put an end to extraneous personal car use that is driven by tax rules rather than people’s actual need to travel. Yet it remains a generous concession – the same as previously applied to company cars driven 15,000 to 25,000km per year – and it does not affect pre-2011 leases that are renewed. And the novated-leasing industry still loves to point out the value of this tax perk, such as when a threat to remove it was defeated by the change of government in 2013:
Tony wants to pay for your car
Well, not the whole car, he hasn’t hit Oprah status. However, Tony Abbott’s new job title means great things for those of us who need and drive a car. Salary packaging the cost of a car known more commonly as novated leasing is an opportunity to have the taxman pay for a fair chunk.
—Letter to Fincar leasing customers as reported by Crikey, November 2013
Perverse as they are, FBT concessions for car use are actually only a small part of the story. The following table shows the extent of tax deductions claimed for car use in recent years. It also indicates that the rate of growth of car-related deductions (nearly 4.5 per cent per year on average) has been greater than growth in incomes.
|Year||Personal||Business||Total Deductions||Revenue Forgone||LCT||FBT||Net Subsidy|
Source: Australian Taxation Office, Taxation Statistics, various years; BITRE, Road Related Revenue and Expenditure, 2011 update and Infrastructure Statistics Yearbook 2017 (FBT figures). All amounts in $million.
‘Year’ is the financial year ending in June of the year stated.
Note that individuals claim two kinds of deductions: personal ‘work related deductions’ against wage or salary income, and ‘business expenses’ against sole trader business income.
‘Business’ deductions include those claimed by sole traders, partnerships, trusts and companies but exclude ‘management expenses’ by superannuation funds, as figures are not available.
Revenue forgone up to 2009 is estimated as 36 per cent of total deductions claimed. For 2016 it is estimated as 36 per cent of personal deductions plus 28 per cent of business deductions.
‘Luxury Car Tax’ (LCT) prior to introduction of the GST in 2000-01 is reckoned as the amount of sales tax collected on new vehicles, less what would have been collected by a 10 per cent GST.
FBT for motor vehicles for 2001 is estimated from total FBT collections, as the ATO ceased to provide FBT statistics by type of benefit. From 2004 onward figures for motor vehicle FBT are provided by BITRE.
‘Net Subsidy’ is revenue forgone due to deductions, less the amount collected in Luxury Car Tax and FBT.
A table similar to this one appeared in the 2001 book Back on Track: Rethinking Transport Policy in Australia and New Zealand by Philip Laird et al. Our figures differ in two ways:
- Laird’s table omitted figures for expenses claimed by individuals against business income. Thus, the total deductions claimed in 1997 are somewhat higher than the figure given by Laird.
- We calculate the tax subsidy by deducting both FBT and Luxury Car Tax from the revenue cost of deductions. This recognises that the revenue lost through tax deductibility of car expenses is partially recovered through tax when the cars are purchased. To allow the effective subsidy to cars to be compared before and after the GST changes, we reckon the ‘Luxury Car Tax’ prior to 2000-01 as the total sales tax on new vehicles (then 22 to 45 per cent), less what would have been raised by a 10 per cent GST.
Public transport advocates are not the only ones to draw attention to the extent of these tax breaks. Even the right-wing Melbourne Institute, not known for being radical greenies, has questioned whether general taxpayers should be subsidising motorists to such an extent.
Deductions for motor cars have grown more than twice as fast as wages and salaries under the Howard Government — 88 per cent versus 38 per cent between 1995-6 and 2002-3. The car is proving to be even more lucrative than charity.
—Melbourne Institute findings as reported in The Australian, 30 August 2005
The tax subsidy to car use depends on movements in tax deductions, FBT collections and overall tax rates. The trend that can be discerned from the table is for deductions to increase strongly while FBT collections for cars and parking decline, so that the net subsidy would steadily increase but for the cuts in company tax rates in particular after 2009. For the 2015-16 financial year the estimated subsidy is $7.5 billion.
State Fuel Subsidies – some relief for taxpayers
It appears to be a little known fact that until 2007 all six states, and the Northern Territory, subsidised the consumption of petrol and diesel. In a rare move against other incentives for car use, all these fuel subsidies have now been withdrawn – the last jurisdiction to axe its subsidy was South Australia, in its 2010 Budget.
The largest of these subsidies by far, and perhaps the only one to be really noticeable to consumers, was the Queensland subsidy of a little over eight cents a litre (8.354 cents to be precise). This was eventually abolished in July 2009. But even in Victoria, petrol and diesel subsidies used to cost the government about $40 million a year – enough to grade-separate a significant level crossing every two years, or buy 100 new buses in one year.
These subsidies came about through historical accident. Prior to 1997, all states other than Queensland collected tax of around 8 cents per litre on fuel, in addition to the Federal fuel excise. In 1997 this was ruled by the High Court to be unconstitutional. To avoid the fiscal crisis that would have otherwise resulted (as the states relied on this revenue to help pay for roads), the Federal Government agreed to collect an extra 8 cents itself and pass it on to the states. As Queensland did not actually collect its own fuel tax, it introduced a Fuel Subsidy Scheme, whereby the extra 8 cents from the Federal Government was passed back to consumers. Other states applied smaller subsidies to cover the small difference between the Federal tax and their own taxes (which varied slightly from state to state, and sometimes from location to location within states).
As part of the tax changes in 2000 this additional 8 cent tax was abolished completely, as the states now benefited from a 10% GST on all goods and services, including fuel. Logically then, the state subsidies for fuel consumption would also have ceased. However, the subsidies continued: the Queensland Government in particular retained its 8 cent subsidy on the recommendation of a ‘Fuel Taskforce’ dominated by representatives of the road lobby. (The alternative proposal, for a $150 reduction in annual registration fees, would likely have been of more help to the low income earners the scheme was supposed to assist. It was resisted by the road lobby because by shifting the cost burden from car ownership to car use, it might have created a disincentive to drive more.)
Except in Queensland, the fuel subsidies amounted to a fraction of a cent per litre and a few dollars a year per person, and their abolition, except in Queensland and northern NSW, went virtually unnoticed to the motorist at the bowser. Yet even these tiny subsidies added up to many millions of dollars a year that could be spent far more sustainably.
|State Expenditure on Fuel Subsidies, 2001-02|
|Fuel Use per Capita
Source: Commonwealth Grants Commission. Subsidies – Petroleum Products, Draft Assessment Paper CGC 2003/52, August 2003. Australian Bureau of Statistics, 2001 Census (population figures).
(This same source draws attention to the disturbing fact that – in 2002 at least – Victoria had the highest per capita fuel consumption of all states and territories, even taking into account the broad expanses of Queensland, WA and the Top End! This highlights the fact that Australia‘s unsustainable level of fuel consumption is actually due not to long-distance travel in the bush, but to routine travel in urban areas for which public transport substitutes more easily.)
According to the Queensland Budget Papers, payments under the Fuel Subsidy Scheme totalled $572 million in 2008-09, its last year of operation.
While the state fuel subsidies have now been abolished, this has often come with a ‘quid pro quo’ for the car lobby. In Victoria, the abolition of the fuel subsidy in 2007 was linked to a reduction in stamp duty on mid-priced cars, at a revenue cost greater than the removed subsidy. Similarly in Tasmania, abolition of fuel subsidies was used to partly fund a 21% reduction in motor vehicle tax. So while the states no longer subsidise petrol consumption, the net effect on the road deficit – other than in Queensland – has arguably been detrimental.
Land Use Cost
When you buy or rent a house, a large part of what you pay is to cover the cost of the land. When you purchase goods or services, you pay a certain amount to cover the proprietor’s business overheads, including the rent on the premises. On the other hand, land for the roads that cars drive on is provided for free, courtesy of the common weal. The cost of acquiring land to build roads is paid by all of us as taxpayers.
This is part of what makes ‘just-in-time’ (JIT) inventory management so attractive for businesses. To store goods in a warehouse you have to pay rent on the warehouse, but storing goods in trucks that never leave the open road incurs no cost beyond a relatively small amount for registration and fuel. This is a classic market distortion that results in excessive numbers of large trucks in our streets.
Just as the cost of roads includes the cost of building and maintaining the roads, it also has to factor in the cost of the land they use. In 1996, the National Institute of Economic and Industry Research (NIEIR) estimated the value of land under public roads to be between $100 and $120 billion. This figure doesn’t include roads and car parks on private land – urbanist David Engwicht has estimated that when you take into account all the space given over to driveways, carports, and commercial car parks (most of which sit idle for half the day), the typical family car has around three times as much space allocated to it as the typical family home. However, car space on private land is privately costed, so we don’t count it in the public cost of car use.
Allowing for inflation between 1996 and today, and the roads built during that time, makes an estimate of $200 billion – NIEIR’s upper estimate plus two-thirds – a very conservative estimate for the value of land under roads now, even after excluding footpaths, bike lanes, bus lanes and tramways. If the value were deemed to increase at the same rate as Australian house prices, it would now be around $400 billion even excluding the value of roads built since 1996. Nonetheless, to avoid accusations that we’re unnecessarily padding the figures, we have adjusted only for the Consumer Price Index, which increased 63 per cent between 1996 and 2016.
The annual rent on land typically sits at around 5 per cent of the site value. 5 per cent is also close to the ‘risk free rate of return’ that sets the minimum yield for low-risk investments such as government bonds. Thus, a conservative cost for public land use by cars is 5% of $200 billion, or $10 billion a year. This represents the ‘opportunity cost’ to the Australian public of turning over land to the more-or-less exclusive use of cars and trucks rather than other uses.
It should however be noted that the notional ‘Capital Asset Charge’ the Victorian Treasury applies to all public infrastructure (with the anomalous exception of roads) is based on an 8 per cent rate of return. Using this rate would put the land use cost of Australian roads at $16 billion a year. But in keeping with the conservatism of our figures, we use the lower $10 billion figure in our calculations.
Noise and Pollution
These figures are taken from a 2001 submission by the Bus Industry Confederation to the National Fuel Tax Enquiry. As explained at the outset, they represent ‘externalities’ associated with car and truck use: costs that are paid neither by motorists nor by transport authorities acting on behalf of motorists.
The value of $1000 million for noise costs is the BIC’s ‘lower’ estimate of $700 million adjusted for 45 per cent CPI growth from 2001 to 2016. The ‘upper’ estimate is close to $2.8 billion in real terms. Uncertainty arises because the most recent comprehensive studies of noise damage appeared in the early 1990s, and changes in costs since then have to be estimated. The BIC also notes that their figures assume that noise damage does not occur in non-urban areas, an assumption that many rural Australians might dispute.
The figure for pollution takes into account some of the indirect health costs due to respiratory illness mentioned above. However, the approach taken is more conservative, setting aside possible effects about which there is still much uncertainty. So again, this is a minimum credible estimate rather than a most probable estimate. The figure has not been time-adjusted, as on past experience any marginal improvements in pollution performance since 2001 wlll have been more than cancelled out by the increase in car and truck travel and by cost inflation.
Our figure for climate change is also based on the BIC submission to the National Fuel Tax enquiry, but has been updated to reflect trends in motor vehicle emissions since 2000. The figure has also been verified against later estimates both here and overseas, notably the Stern Review released in the UK in 2006, and the Garnaut Review here.
Following the BIC methodology, climate change costs are estimated based on the short-term marginal effect of greenhouse gas emissions. The marginal cost is estimated at $40 per tonne of CO2 equivalent, which was close to the minimum estimated cost (before the Global Financial Crisis and subsequent recession deflated prices) for the European Union to meet its target under the Kyoto protocol, and is also roughly equivalent to the fixed price for ‘small-scale technology certificates’ under the Australian renewable energy obligation. As the BIC states, “it is stressed this value is only relevant for the short-term: the marginal damage costs and the marginal abatement costs will increase dramatically in future years”.
We note that the UK Department for Transport gave a lower bound cost (assuming a ‘business as usual’ scenario) of £42.44 (about AU$70) per tonne of CO2 in 2006, rising to £98.32 (about AU$160) per tonne in 2060. The Stern Review on the economics of climate change gave a lower estimate of US$30 per tonne of CO2 (about AU$32), assuming that action to stabilise greenhouse emissions commenced in 2006. With action delayed, the cost is US$85 (about AU$90) per tonne, similar to the UK DfT estimate. In all cases the costs rise further over time.
In December 2007, British Prime Minister Gordon Brown instituted a national ‘carbon price’ to be used in the economic assessment of government decisions. The price started at £25.50 per tonne (about AU$42) in 2007 and rises annually, to reach £59.60 per tonne (about AU$100) in 2050. The Gillard Government famously nominated $23 per tonne as its starting carbon price based on a modest 5% reduction target in emissions by 2020. This is below the lowest cost figure from the Stern Review (now almost certainly out of date) and would very likely not have sufficed to achieve the 26-28% reduction from 2005 to 2030 that Australia has now committed to under the Paris Agreement.
According to the BITRE Infrastructure Statistics Yearbook, the aggregate emissions from road vehicles (excluding buses) reached 78.5 million tonnes of CO2 equivalent in 2016, and are steadily increasing (though with a brief pause between 2004 and 2009). Costing these emissions at the conservative $40 per tonne gives the $3.1 billion figure used here. Costed at the UK ‘business as usual’ figure of $70 per tonne, the figure would be about $5.5 billion, rising in future years.
Petrol and Diesel Excise
According to the Australian Taxation Office, fuel excise raised $11,645 million in 2015-16. The total excise collected on all petroleum products was $17.8 billion, but around $6.2 billion of this was rebated. Rebates are available for use of diesel and some other fuels in trains, ships, large trucks and buses, farm machinery, mining operations, electricity generation and health care. At various times, rebates have also been available for general fuel sales in rural areas, and for designated ‘clean’ fuels such as biodiesel.
Once again, we have been deliberately conservative in using the raw Tax Office figures for excise and rebates. This gives figures for net excise that are about $600 million higher than those reported by BITRE in their ‘Road Related Revenue’ summary. This may be due to our inclusion of certain other fuels and products unrelated to road transport (such as aviation fuel and condensate) which BITRE may omit from their figures. However, restricting to the ATO figures for petrol and diesel only reduces the revenue figure by about $2 billion, which undershoots the BITRE figure substantially. Again, to avoid any doubt about the good deal motorists get under present-day arrangements, our accounting aims to ensure that any foreseeable errors are on the side of overestimating revenue and underestimating expenditure.
|Petrol/Diesel excise and rebates ($million)|
|Net fuel excise revenue||10,525||10,332||11,645|
|Total excise collections||14,246||15,543||17,804|
|Rebates (off-road uses)*||2,552||5,065||6,082|
|Rebates (on-road uses)*||890||4||0|
|Rural sales rebate||261||0||0|
|Cleaner fuels grants||2||102||16|
Source: Australian Taxation Office, Taxation Statistics, various years.
* In 2006 a new Fuel Tax Credits Scheme replaced the earlier rebate schemes, with the effect that ATO figures no longer discriminate between ‘off-road’ and ‘on-road’ uses. The full scheme amount has been assigned to ‘off-road uses’ above as this still accounts for the majority of rebates.
Between 2001 and 2014 the rate of fuel excise was frozen so that it didn’t rise with inflation (unlike the price of everything else, including public transport). This means growth in revenue from fuel excise over this period was almost entirely due to growth in car and truck use, which roughly coincided with growth in GDP. Without adjustment for inflation, excise is an unsustainable revenue source: given an economic downturn or further rises in fuel prices, traffic levels and hence revenue from fuel excise can be expected to decline, while many of the costs of car use (such as the cost of maintaining roads) will continue to rise with inflation.
From the table above, and that for road expenditure, this effect can be seen between 2005 and 2009. While expenditure on roads steadily grew, excise revenue was stagnant or even declined slightly. So while until about 2006 it was possible to say fuel excise revenue sufficed to cover the raw cost of road construction in Australia (leaving aside other costs such as road trauma and pollution) this is no longer possible. Even as of 2009 there was a shortfall of at least $4.6 billion between what was collected in fuel excise and what was spent on roads. As of 2016 the shortfall has blown out to over $14 billion, with more than twice as much spent on roads as is collected from fuel tax.
GST on Fuel and Vehicles
GST is charged on the pump price of fuel, and on the purchase and maintenance costs of vehicles, just as it is on public transport fares and all consumer goods other than basic food. Because it’s levied equally on all forms of transport it is questionable whether the GST should be counted as a tax on road use, but we include it anyway for the sake of argument. It’s important to realise that the government doesn’t actually get all of the GST money collected, since any vehicle use for business purposes has the GST refunded as an ‘input tax credit’. The GST revenue figure is thus equal to total GST collections less input tax credits refunded.
The GST is unlike many other charges in that it fluctuates with retail prices, so estimates will vary from year to year. BITRE uses Bureau of Statistics figures to estimate GST on the purchase, maintenance and fuelling of motor vehicles as $4,016 million in 2015-16. This figure has changed little in real terms since 2010.
It’s worth noting that even at today’s high prices, the overall tax on petrol is still less than it would have been under pre-GST arrangements. The excise now set at 41 cents per litre (and frozen at 38 cents between 2001 and 2014) would by now have risen to 65 cents per litre had the Howard Government not cut and frozen the rate. So unless average pump prices rise to $2.64 a litre (when the GST collected is 24 cents), the motorist is still getting a better deal than previously. (This is also confirmed by the ABS Household Expenditure Survey, which shows expenditure on petrol decreasing as a proportion of household income.)
Vehicle Registration Fees
Car and truck registration fees levied by State governments raised $7,011 million in 2015-16, according to the BITRE. This amount has steadily increased in real terms over the past decade, after being relatively flat across the turn of the century.
|Year||Vehicle registration revenue ($million)|
While much is made of the money State Governments collect from registration fees – and there has certainly been a substantial increase in nominal-dollar terms in the past decade – the total is still less than what the Federal Government effectively gives back through vehicle-related tax concessions.
Until 2009 the Australian Prudential Regulatory Authority (APRA) published annual statistics on the Australian insurance industry, which included the total annual premiums paid for various classes of insurance. From these statistics it was possible to determine the total insurance premiums received from motorists, both for compulsory third party (CTP) insurance (paid to both public-sector organisations like the TAC and to private-sector companies) and for other kinds of car insurance (almost all to private-sector insurance companies).
Unfortunately, APRA has ceased to publish statistics in this form, and so total premium revenue must be estimated. For this exercise we have relied on the published figures for the period prior to 2009, together with BITRE figures for the size of the vehicle fleet, and figures published by the Insurance Council of Australia that index the average premium for comprehensive motor insurance between 2004 and the present day. These latter figures have been used to adjust the 2008-09 aggregate premium to provide an estimate of total premiums collected in 2015-16.
|Motor vehicle insurance premiums ($million)|
|Fleet size (million)||13.9||15.7||18.4|
Source: Australian Prudential Regulatory Authority. Half Yearly General Insurance Bulletin, June 2005 and June 2009. BITRE, Infrastructure Statistics Yearbook, 2017. Insurance Council of Australia, GI Industry Trends – Motor Vehicle Insurance.
The estimated total of $15.6 billion for 2015-16, adjusted from $12.1 billion in 2008-09, has been used in our accounting. Note again that this figure is for all car insurance, not just that which covers damage due to crashes. And while premiums do show a growing trend, the amount is still a good deal less than the estimated financial cost of road trauma and damage.
Figures for toll revenue are provided periodically by BITRE. The figures tabulated here are in nominal dollars, one may also consult the BITRE Yearbooks for inflation-adjusted figures calibrated to every year.
|Year||Toll revenue ($million)|
Aside from registration fees, State governments also collect revenue from motorists in the form of stamp duty on registration transfers, driver licence fees, and sundry road transport and maintenance taxes. These various fees totalled $3,205 million in 2015-16, with stamp duty accounting for the largest share.
To this figure must be added the small amounts collected by the Federal Government under the Federal Interstate Registration Scheme, and the residual import tariffs on motor vehicles. These totalled $682.8 million in 2015-16.
Until 2000, State Governments collected around $1.5 billion a year in franchise fees from fuel retailers. This fee was abolished with the introduction of the GST, so any reference to it in road lobby literature may safely be ignored.
Last modified: 5 January 2019