Driving Blind – Fuel Levy vs Tolls

Jeremy Cronin’s response to calls for the Fuel Levy to be used to fund Gauteng roads provides no evidence for the existing of thinking.

He’s been horribly mis-represented, is a horrible communicator, or is a horrible policymaker – and I don’t even necessarily believe the Fuel Levy is the appropriate way of funding Gauteng roads.

Good reasons FOR an increased Fuel Levy

The argument for using the Fuel Levy to fund road infrastructure goes something like this:

  1. This is an existing tax (levy) with an existing mechanism to administer it – no additional infrastructure / toll-booths / retail billing systems / enforcement issues exist. Huge plus. Huge.
  2. There is a direct link between (i) distance travelled and weight of the vehicle and (ii) fuel consumed. Distance travelled and weight of vehicle are primary factors affecting wear on roads so it’s actually a good match for the cost incurred by vehicles on our roads, on average.
  3. The Fuel Levy adds costs to all forms of road trip, removing incentives to use back-roads and appropriately charging road-transport vehicles for use of all roads since all roads require maintenance.
  4. Driving during peak traffic times will give rise to poorer fuel consumption given the stop-start nature of traffic so the Fuel Levy acts in a small way as a type of “congestion charge”.
  5. A distortion to this is that more fuel efficient vehicles will pay less tax without necessarily weighing less and therefore without necessarily doing less damage to the roads. This distortion is probably overwhelming positive in that it provides a natural incentive to use more fuel efficient vehicles with foreign currency purchases of oil advantages and pollution advantages.

You can’t really overestimate the value of point 1 above, with points 2 and 3 saying why it’s broadly a useful idea and 4 being a subtle benefit. 5 is more of an “interesting point”.

Good reasons AGAINST an increased Fuel Levy

The good reasons against using the Fuel Levy, in my view, are the following:

  1. The Fuel Levy is levied nationally but will be used to fund roads in specific areas, there is a temporal mismatch between who pays (everyone) and who benefits today (for now, primarily Gauteng). This reduces the match of making those who incur the expenses pay for the expenses.  I’m not entirely convinced that this is a very strong reason when we take a longer view. We’ve spent money on Gauteng roads recently, but will hopefully be spending money in other areas at other times. Over time, this should even out. It’s not perfect though – Gauteng drivers (and Western Cape too!) get fabulous new roads with good surfaces, many lanes and few potholes whereas drivers in other areas may never get the same quality roads. This “on average over times” argument only truly works if road quality around the country, in the long run, is average. Provincial governments may have different priorities so Eastern Cape drivers may pay a fuel levy but never get good roads.
  2. If the Fuel Levy is not levied nationally, it will give rise to increased arbitrage opportunities near boundary regions.  (I’m not sure who prevalent this is with current coastal / inland pricing.)
  3. The relationship between fuel consumption and damage to road surfaces is not necessarily linear. There is no guarantee that this provides the correct costing whereas a Toll could be specified by vehicle mass and number of tyres allowing for a more exact match.
  4. Tolls can also be used more carefully to influence public transport choices through reducing or even waiving Tolls for certain types of public transport (buses, even taxis) which are better for the country overall.
  5. The move towards plug-in electrical vehicles or other vehicles using electrical power off the grid rather than petrol won’t pay the Fuel Levy even though they incur road maintenance costs. At the moment, this is hardly an issue for South Africa, but in terms  of setting appropriate policies, it may require further attention in future. This could be viewed  positively as an incentive to use electric vehicles rather than fossil-fuel burning vehicles, but as this is co-incidental there’s no reason to believe this is the “correct” subsidy now or in future.
  6. Congestion charging is more specifically dealt with by charging different toll rates based on time of day. This probably won’t be a popular option nor will it make the toll charges more simple, but it needs to be mentioned if we are going to have this in the “plus column” for the Fuel Levy approach.
  7. It’s more difficult to enter public-private partnerships where a private company builds the roads and as a result wins the concession to collect tolls from the road. Similarly, it’s more difficult for government to issue structured debt instruments with repayments directly out of the tolls received on the projects. These instruments are quite attractive to pension funds and life insurers looking for good quality, long-term but not necessarily liquid investments. By limiting funding options, the ultimate cost of the project for the country may be higher.

Things that this conversation is but shouldn’t be.

Overall, the question about how best to raise funds to pay for roads should not involve the cost of the roads themselves. Cost and efficiency of levying the tax sure, but not the cost of the roads themselves.

Problem is, most people don’t like paying more for something. So inevitably an increase in costs will give rise to complaints. The question of whether the correct number of roads were built / repaired, whether the correct roads were built / repaired and whether they were built / repaired cost effectively is an incredibly important question, just not relating to how to charge for those costs.

(Given that Airports Company SA, “ACSA”  now has some of the highest fees  in the world resulting from their desperate desire to recoup the huge expenses incurred upgrading on our airports recently, it’s fair to ask the question “were the right roads built at the right cost and who benefited from it all really?”)

Given that nobody likes paying more, it’s unsurprising that Gauteng road users are up in arms over the proposed Tolls.  Care to guess what would have / will happen if Tolls are replaced with a hike in Fuel Levy? All the roads users outside of Gauteng will be up in arms.

Cronin’s bizarre response to the Fuel Levy proposal

I’m quoting heavily from the link News24 / SAPA article.

“The National Treasury is opposed to ring-fencing tax revenues because (it) brings about inefficiencies in government spending over time as the lack of transparency means that spending agencies lose the accountability of the budget process for how effectively they apply the funds.”

Which of course is a complete red herring. How does compartmentalising revenues DECREASE transparency? Surely all plans to use funds would still need to be approved and be in line with developmental country goals? The idea that we just blindly give the Fuel Levy funds to SANRAL and say “make hay” is a straw man that Cronin is itching to knock down. But it’s an irrelevant point. Given the high costs of ACSA developments, Arms Deal catastrophes, Coega-styled White Elephants, it’s clear we have major problems in this area already. Fuel Levies won’t make this worse and the problem is infinitely solvable through appropriate policies, reporting, accountability and governance frameworks.

Cronin explained that the actual amount allocated out of the budget for road infrastructure was more than the amount raised from the fuel levy.

“In the 2009/10 financial year, for instance, the amount that would have been available for road infrastructure from the fuel levy (had it been ring-fenced) was R22bn whereas the actual amount allocated to road infrastructure was R29.2bn – a shortfall of R7.2bn.”

This point will require some attention, but one doesn’t need to be too imaginative to find a range of solutions. Firstly though, let’s highlight Cronin’s smoke and mirrors: 2009/2010 had enormous expenditure because of the World Cup, amongst other things. The fact that government got its forecasts wrong is a problem of forecasting, not method of revenue collection. Cronin is creating artificial problems by taking a very narrow read of “ring fencing”.

A few starting points to deal with the problem of the Fuel Levy not being sufficient for road infrastructure spend.

  • Decrease general taxes while increasing the fuel levy.
  • Having a combined fund for road transport arising from Fuel Levies and general taxes
  • Having the allocation from general taxes in a particular year or on a three-year cycle depend on the exact magnitude of planned investment over that period.
  • Allow notional pre-funding and loans within the government finances
  • Allow explicit, external borrowing for road infrastructure projects to smooth the long-term capital nature of some investments.

These aren’t full solutions, but they are useful starting points for an intelligent discussion of the issues.

Cronin also questioned whether finance raised from the fuel levy should only be invested in road infrastructure when other transport modes such as rail were also in need of financing.

Clearly Cronin and his policy cronies haven’t considered that maybe this is an explicitly Good Thing.  The smart reason to use a Fuel Levy rather than a general tax is that it matches use.  Rail should be brought up to standard by general taxes and then financing on an ongoing, sustainable basis via appropriate charge for use on commercial or near-commercial terms. It’s this sort of comment that makes me wonder whether our country’s finances are run by anyone with economic policy and development aims in mind at all.

Of course, Cronin does have to deal with these comments:

Organisations which initially opposed tolling reacted, saying that it would still pose a financial burden on already overtaxed road users.

Bold statements of being “overtaxed” because, nobody likes paying more for something.  Even if it costs more. Even if by more appropriately matching the price of using roads to the costs incurred by our country in building roads we move heavy long-haul transport to trains and sea (which would be a good thing). Of course, our rail network also needs work to ensure it is a viable, efficient and reliable transport mechanism, but at the moment is competing against unfairly “cheap” roads where users are paying the direct cost of using the road.

In summary, Cronin’s response is generally irrelevant and doesn’t cover the range of very real reasons to be wary of an increase in the Fuel Levy.  I really would like to contribute to policy on these issues, but it’s not clear government is even interested in openly considering policy issues.

Conversations that aren’t yet happening enough

This post has primarily looked at the efficiency of collection of government revenue and how well the collection mechanism matches the cost. These are good principles most of the time.

However, any policy at all in our country that doesn’t keep structural unemployment in mind is flawed. A driver of structural unemployment are the high costs of transport, particularly for low-income workers and the currently unemployed. Along with all this “matching” of revenue collection and incurring road use expenses for the country, we also need to consider adjustments to promote (and as a starting point, not deter) employment.  A transport subsidy may be the right answer.


Published by David Kirk

The opinions expressed on this site are those of the author and other commenters and are not necessarily those of his employer or any other organisation. David Kirk runs Milliman’s actuarial consulting practice in Africa. He is an actuary and is the creator of New Business Margin on Revenue. He specialises in risk and capital management, regulatory change and insurance strategy . He also has extensive experience in embedded value reporting, insurance-related IFRS and share option valuation.

Join the conversation


  1. I seem to think that we have all lost the plot in respect of Roads, what improvements have been made to road infrastructure across South Africa since 1994? In general absolutely none, unless of course we take into account the numerous highways that were simply privatised and turned into Toll Roads. To travel from Malelane to Pretoria will cost in the region of R 150.oo on a road that was built and paid for with tax payers money. In Guateng there have too been minimal upgrading of roads over the past 15 years – WHERE HAS ALL THIS MONEY GONE. If we are not seriously not generating enough funds for the maintenance and development of our road network why not introduce the same system as what is found in Austria and Switzerland but for a few exceptions all national roads are covered with a once off annual fee in the form of a “Vignette” which is stuck onto the windscreen of the Vehicle. The cost of this varies dependant on Vehicle size and weight. This “Sticker” needs only to be bought if the vehicle in question will use any of the National Road network. Vehicle caught using the National road network are given spot fines that by far exceed the purchase price of the road tax “Sticker”.

    This system insures that the roads are paid for by all those who use them, even our neighbours in Swaziland have learned this and charge all foreign vehicles entering a Road Tax.

    If I am not mistaken there are app. 15 000 000 vehicles registered in South Africa of which 1/3 are transport vehicles.

    If a annual Road tax of R 500.00 was charged for passenger Vehicle and R 2 000,oo charged for all commercial vehicles this would give the country an additional15 Billion for roads.

    This way each person is burdened for their own roads and no one should have to pay any tolls.


    1. Thanks Shaun. The thing is, we already have a fuel levy and vehicle licence fee, so the mechanisms to charge other than tolls already exist. The advantages and disadvantages of tolls remain of course. I’m not close enough to the detail to know whether money is being well spent or wasted, but some of the reports of the costs of the system to enforce the tolls do seem pretty high.

  2. Another big disadvantage of fuel levy is that fuel consumption on urban roads is higher than fuel consumption on freeways. In other words, on a per kilometre basis, people who don’t use the freeway will be subsidising those who do.

Leave a comment

Your email address will not be published.

This site uses Akismet to reduce spam. Learn how your comment data is processed.