subscribe Support our award-winning journalism. The Premium package (digital only) is R30 for the first month and thereafter you pay R129 p/m now ad-free for all subscribers.
Subscribe now

Despite finance minister Enoch Godongwana’s confirmation two weeks ago that the e-toll scheme would finally be scrapped, there are some who still argue that this was the best way to finance the Gauteng Freeway Improvement Project (GFIP).

This shows a clear misunderstanding on two fronts: first, the cost of collecting e-tolls; and second, the unworkable way in which this scheme was to be administered. From the start, there were always two distinct arms to the project — the first being the costs of building the roads, and the second being the way in which this construction would be financed. On the issue of needing to upgrade the Gauteng freeway network, there was never any doubt. But when it came to how to pay off the bonds used to finance the upgrade, the e-toll scheme was the worst, and most expensive, option by far.

How much should the GFIP have cost society?

If you follow the numbers from the South African National Roads Agency Ltd (Sanral), it’s clear why the final bill of R17.8bn paid to upgrade the 200km of roads was grossly overpriced. The following timeline, provided by Sanral and the National Treasury, paints that picture eloquently:

  • In 2005, Sanral’s own “statement of intent” document presented a price tag of R4.6bn for a 340km freeway upgrade and development project. Of this, the 200km for phase 1 of the GFIP was about R2.8bn.
  • However, in January 2008, then transport minister Jeff Radebe greenlit the GFIP at a massively increased price tag of R11.8bn. A project which, just a few years earlier had a price tag indication at under R3bn, had shot up by about 400%.
  • But it got even worse. Despite a grossly inflated price tag of R11.8bn, which was approved in the same year that the GFIP construction began, the final construction cost of the project came in at R17.8bn — another 50% on top of the already inflated price by the time the construction was finished a few years later.

The answer about why the construction costs increased so massively depends who you speak to.

The construction companies and Sanral argued that the 2010 Soccer World Cup had changed the dynamics, making the environment a lot more expensive. Perhaps — but definitely not by almost 400%, on Sanral’s estimated pricing in 2005.

In 2013, the Competition Commission indicated that the construction firms had indeed colluded to drive up prices, but this was only part of the problem. The bigger issue is that Sanral allowed this to happen on its watch — it had all the authority to contain costs, but instead looked the other way and allowed the corruption and collusive conduct to roll on unabated.

It chose the e-toll option, which would slap roughly R1bn per year onto society just to cover the costs of collecting the e-toll fees

What financing options for the GFIP were available to Sanral?

There were always two options for Sanral to finance the GFIP.

The first was to receive funds directly from the Treasury for this upgrade, as it does every year for its non-tolled portfolio of roads.

The second was to declare this 200km stretch a tolled road and then try to collect these funds from road users. The challenge was that conventional tolling (through a boom-down plaza) would cause more congestion on these freeways, so an electronic “drive now, pay later” mechanism would be preferable, as it would allow the traffic to flow.

This is where Sanral’s plans went awry. It chose the e-toll option, which would slap roughly R1bn per year onto society just to cover the costs of collecting the e-toll fees — and that’s before any money was allocated to settle the actual road construction bonds.

That administration cost of about R1bn per year was based on the original five-year tender of R6.22bn, which Sanral said had been secured by the Electronic Tolling Company (ETC). The actual contracted figure was massively inflated to R9.9bn, which escalated the annual e-toll operations services component to R1.64bn per year over five years. That seems iniquitous as a funding mechanism for a bond to cover an R11.8bn freeway upgrade project — which would work out to R811m per year over 20 years, at 10% interest. But even at the grossly inflated R17.8bn cost, the annual bond payments would average about R1.2bn per year for 20 years.

So, why on earth would Sanral want to double the cost to society by slapping administration or collection costs of R1.6bn per year (to be paid to the ETC) on top of the construction costs?

It wasn’t as if an extra R811m per year for the GFIP bonds would have broken the bank anyway — as far back as 2008, Sanral was already receiving R5bn a year from the Treasury for its non-tolled road portfolio. Especially as a Treasury allocation would come without the headache of the administration costs, or implementation of the scheme.

The unworkable e-toll headache

Back in 2008,  Sanral and Radebe erred when it came to applying their minds to the e-toll pitfalls that lay ahead.

A thorough and meaningful independent investigation could have provided more sound advice than their “paid for” consultants did. After all, this was a scheme that had never been tried and tested in South Africa before.

One requirement for success, for example, was overwhelming public support for the scheme — but this was clearly missing from Sanral’s strategy for the Gauteng project.

Furthermore, a thorough and independent adviser might have also informed them that the e-toll scheme’s functionality relied heavily on the country’s vehicle registry system (eNatis), as well as an efficient postal services system. Yet neither could even remotely be described as efficient.

And finally, Radebe ought to have questioned Sanral’s enforcement ideas — including the threat of withholding vehicle licence renewals if people didn’t pay their toll fees, or downgrading their credit ratings. Neither of these was possible.

With the end of e-tolls, confusion abounds as to who owes what

Since Godongwana’s announcement, it has become obvious that there is no clear plan on how to proceed with this alternative option, in the absence of e-tolls.

While it appears that the financing of the GFIP bonds will be covered by 70% allocations from the Treasury and 30% from the Gauteng provincial government, the amount due is unclear.

A few days ago, Gauteng premier Panyaza Lesufi said the province still needs to negotiate with the Treasury how much it will have to pay, but also how its portion will be financed and over what period. Back in 2012, court papers lodged by Sanral indicated that R20bn had been borrowed and ring-fenced for the GFIP bonds.

Furthermore, Sanral’s financial statements show that between 2012 and 2022, the Treasury had already allocated R22.4bn to the agency specifically for the GFIP debt. This should effectively leave very little outstanding GFIP debt, including interest — unless, of course, Sanral has been reckless and not attributed any of the Treasury allocations towards the GFIP debt.

Either way, this illustrates that what Gauteng residents, and society, now need is absolute clarity and transparency from Sanral and the Treasury on how they arrived at their numbers.

The truth is, the e-toll plan was intended to enrich a few organisations connected to the revenue flows. These include Sanral, which planned to generate around 30% of its total revenue from just 1% of the road network, had it achieved the 94% compliance levels it wanted.

Fortunately, the public refused to roll over and comply and the government has had to come to its senses and apply the right funding mechanism — using the allocations from the Treasury. Even with a needless 10-year delay, the Treasury allocation option is far cheaper than the grossly overpriced e-toll option.

* Duvenage is CEO of the Organisation Undoing Tax Abuse

subscribe Support our award-winning journalism. The Premium package (digital only) is R30 for the first month and thereafter you pay R129 p/m now ad-free for all subscribers.
Subscribe now

Would you like to comment on this article?
Sign up (it's quick and free) or sign in now.

Speech Bubbles

Please read our Comment Policy before commenting.