Cancelling East West Link cost $780m, but saved $4b

This is old news, but it keeps coming up, and I never got around to blogging about it in detail, so…

The popular narrative of East West Link (from those who wanted it built) is that the Andrews Government paid $1 billion for nothing.

As you might expect, there’s a little more to it than that.

The Victorian Auditor General found:

The EWL project was terminated in June 2015 with more than $1.1 billion paid, or expected to be paid, by the state for little tangible benefit.

There’s your headline figure.

But VAGO also found that the cost will be:

partially offset by future proceeds from the sale of properties acquired for the project which the Department of Treasury & Finance estimates at $320 million.

So that makes it $780 million. Which is still a lot of money.

Before supporters of the road (and the 2010-2014 Liberal National Coalition) get too cocky, VAGO points out how badly the project was managed ahead of the 2014 election:

The audit found that the EWL business case did not provide a sound basis for the government’s decision to commit to the investment and that key decisions during the project planning, development and procurement phases were driven by an overriding sense of urgency to sign the contract before the November 2014 state election.

Advice to government in the lead up to signing the contract did not sufficiently assess the benefits of delaying contract signing to mitigate risks posed by the unresolved legal challenge to the project planning approval decision. Signing the contract in these circumstances was imprudent and exposed the state to significant cost and risk.

You’d expect the politicians involved to ignore the findings, but what’s unbelievable is that the government departments involved ignored VAGO’s recommendations. From the VAGO report introductory letter:

The report makes recommendations to DTF to provide guidance for development and delivery of major projects and for the Department of Premier & Cabinet to clarify requirements for frank and fearless advice from the public sector. Disappointingly, these departments have rejected the recommendations.

'Lies' #EWLink

Benefits of the project

Would it have brought benefits? Yes, but as we know from the Benefit Cost Ratio of only 0.8, benefits of less value than the money spent.

The BCR only crept higher than 1 if vague wibbly-wobbly Wider Economic Benefits were added.

VAGO had this to say on WEBs (in section 3.4.1), and it’s fair to say they weren’t supportive of the approach:

The EWL business case relied heavily on the inclusion of estimated WEBs to support the assertion that the project was of net benefit. Without these WEBs, which were unusually high as a proportion of total benefits, the project cost was clearly higher than the expected benefits.

WEBs relate to economic benefits that are not typically captured in traditional cost-benefit analysis. Commonly considered WEBs include:

  • ‘agglomeration’ impact (an increase in productivity due to improved proximity to suppliers and labour markets)
  • the impact of transport on increasing competition
  • competition related user benefits.

The DTF guidelines indicate that WEBs are most relevant to transport and other large infrastructure projects. However, guidance materials from DTF and Infrastructure Australia indicate caution should be exercised when estimating and considering WEBs as part of the economic assessment of projects.

In other words, Department of Treasury and Finance were only too happy to rely strongly on WEBs, even though their own guidance materials said you need to be very careful with them, and not rely on them too much.

DTF’s guidelines note that the extent to which WEBs exist over and above benefits counted in the standard economic evaluation is not yet clear, and will depend on the nature of the project under consideration.

The June 2013 business case does not provide sufficient information to explain the basis for the significant change in WEBs from the March 2013 business case.

This was important, because an external peer review of the March 2013 business case raised significant issues with the plausibility of the level of WEBs claimed for the project in terms of their ratio to total benefits.

So after the March 2013 business case showed the project was very shaky economically, a revision just two months later showed it to be stronger, but contained no justification for the change.

For VAGO to have “significant issues with the plausibility of the level of WEBs claimed” is really damning stuff.

East West Link: eastern section, western gateway in Royal Park

The cost if it had been built

But back to the cost. $780 million went out the door. But what would it have cost if the project had gone ahead? VAGO spells it out in section 2.4: $12.8 billion (nominal terms), with $3.3 billion of that being State funds, $1.5 billion from the Commonwealth, and $8 billion from tolls.

So while some would argue that scrapping the project wasted $780 million, another way to look at it is that it saved $4 billion of taxpayer funds, plus billions more in motorist tolls.

Despite the best efforts of the outgoing government, the incoming government fulfilled one of its major promises in scrapping the road.

Scrapping it, apart from saving money overall, also prevented the trashing of Royal Park, covering Clifton Hill in massive flyovers, further entrenching car dependency and producing more traffic — and long term, it would have made little difference to congestion on a corridor where traffic wasn’t increasing anyway.

* * *

Further reading:

EWLink zombie is dead, buried, cremated – some thoughts

East West Link is now that it’s dead, buried and cremated (to coin a phrase). Though I’m not sure that’s how you destroy zombies.


Some closing thoughts on the project…

The $339m payout is less than a single year of the expected $345m annual Availability Payments that would have been paid if it had been built — let alone the billions in other costs. So yes it’s a lot of money, but given the <1 Benefit Cost Ratio, we the taxpayers of Victoria really dodged a bullet here.

Part of the problem is that for the huge cost, the tollway would have provided some traffic relief… but only for a short time. Studies by the Linking Melbourne Authority showed traffic on Alexandra Parade would have returned to the previous levels by 2031, just 12 years after opening.

Given it won’t be built, the Eastern Freeway will continue to finish at Hoddle Street. It was planned that way. In 1973, Premier Hamer cancelled the forerunner to the East West Link, an inner-north surface freeway across, and but pushed ahead with building the Eastern, knowing it would finish at a dead-end, which the road lobby happily accepted at the time. So the current situation is by design.

Being a zombie, it came back. Salami tactics.

Where traffic from the Eastern Freeway goes

East West Link wouldn’t have helped unclog the Eastern and Hoddle Street, because most cars are headed for the CBD and inner city, not across to Flemington and the airport. You can actually see this when driving east to west: the traffic moves faster in the 1-2 lane sections from College Crescent to Elliot Avenue than it does in the 4-5 lane sections further east.

In fact, current traffic levels are falling. Vicroads data shows traffic fell up to 15% on Alexandra Parade (depending on where you measure it) between 2002 and 2012.

Of course, more traffic would head east-west if there was a motorway provided to do it. This is why the road wouldn’t have provided more than a few years’ relief. More traffic would be induced.

We’ve spent seventy years trying to solve traffic congestion by building more road space for cars. Sooner or later we’re going to have to accept that it doesn’t work.

Broadly, transport is supply-led. You build more roads, you get more cars. You provide more (usable) public transport, you get more people using it.

But the scale of PT capacity is vastly bigger, because each person isn’t bringing 2 tonnes of rubber, metal and plastic with them, so it takes a lot longer to fill up again.

Want to shift 2000 more people per hour along a road? You’ll have to widen the road all the way along and add a lane. Example: the M1 widening, 2007-2009, cost $1.39 billion. And now it’s full again.

Want to shift 2000 more people by train? That’s just 2-3 extra trains; about 10% of a rail line’s capacity.

So, more capacity can often be squeezed out of existing lines, but even if we have the expense of a new rail line, it will take decades to fill up again.

Despite red herrings like “tradies!!”, most people can use PT if it’s provided and competitive with driving.

So now Labor’s pushing ahead with the metro rail tunnel instead, which (along with level crossing removals, signal upgrades, fleet upgrades, and the one thing they haven’t announced yet: duplication of single track) enables a lot more people to be moved through our busy city. Now they just need to figure out how to pay for it, given the Feds under Abbott won’t.

The political implications of all this are obvious. The Coalition went into the 2010 election and won it on the back of public transport issues — reflecting survey after survey that said people wanted PT ahead of big roads.

(A few surveys conclude people want motorways, but only when they’re not offered a choice. But money is limited. Choices have to be made. We can’t afford every multi-billion dollar project that’s thought up, especially not the ones with poor returns on investment.)

Then in late-2011 the Coalition completely changed tack and tried to ram through the East West Link ahead of the 2014 election.

Tony Abbott declared the 2014 election to be a referendum on the East West Link. Well if that’s true, the people voted it down.

Will this zombie come back? Perhaps.

But the lesson here is clear. If you’re determined to embark on the largest infrastructure project in the state’s history, seek a mandate for it.

PS. When I put in the Shaun Of The Dead references, it was just for a cheap laugh. But that specific point about destroying the brain is a reminder that these issues about how we want our city to develop, and whether we want to pursue policies prioritising private cars or mass transport are not primarily about engineering, but ideas.

More reading:

The #EWLink billion dollar booby trap

The previously secret East West Link “side letter”, offering compensation even if the contract was deemed invalid, seems to underscore just how desperate the State Coalition was to build the road — despite it being a project that:

Here’s an excerpt from the document, which perhaps surprisingly, was released by the Coalition themselves yesterday:

East West Link side letter

It’s a bit like a contractual version of Pauline Hanson’s assassination video: Fellow Victorians, if you are seeing this now, it means East West Link has been killed.

Obviously it was designed to ensure the project went ahead, even if the Coalition was voted out, and Labor tried to cancel it. A billion dollar booby trap.

East West Link: eastern section, western gateway in Royal Park

I think this is far worse than simply signing a bad contract. It goes much further, by offering a kind of insurance against its later cancellation. (Remember, governments regularly cancel contracts).

This is veering into the territory of… how do I put this? It’s lacking respect for the electorate so much that you would force your decision onto them, even if they specifically voted against it. Remember, numerous people from Tony Abbott down declared the November 2014 election a referendum on the project.

This kind of letter says “We want this built. We don’t have a mandate for it, but we don’t care what you think. And if you vote us out, we’re going to try and force the other side to build it anyway.”

My notes from a quick skim of the #EWLink business case

Late last night, the Herald Sun unexpectedly published the entire East West Link business case, ahead of its official release today.

Some notes from me from a quick flick through:

p12 makes various high-level claims, particularly faster trips for motorists — but as we know, this benefit never lasts because traffic increases.

EWLink: Proposed tolls

p17 flags the toll prices used in the modelling: (2012 pricing) cars $5.50 in peak, $4.40 off-peak. Light commercial vehicles $8.80 peak, $7.04 off-peak. Heavy commercial vehicles $16.50 peak, $13.20 off-peak. I wonder what regular motorists (especially those with commercial vehicles) make of these toll levels?

By comparison, bypassing the city along the Bolte Bridge or the Domain/Burnley tunnels (not both) currently costs $7.06 in a car, or $8.15 for both sections. It’s only marginally more expensive for both because there’s a cap… I assume it’s unknown if a similar cap could exist where adjoining motorways are run by different operators.

p17 says the funding gap between the toll revenue and the cost of construction is $5.3 to 5.8 billion.

p39 says north-south public transport is being degraded by traffic congestion, which may be the case, but that’s because authorities have allowed it to happen by failing to provide tram/bus priority through busy intersections such as Alexandra Parade. They continue to prioritise large numbers of vehicles (single-occupant cars) over large numbers of people. It’s important to recognise that while the greater East West Link project includes tram priority measures, these can be implemented without building a big road tunnel.

p41-42 appears to be cherry-picking statistics to try and claim there’s a lot of demand for cross-city traffic. For instance the diagram at the top of page 42 implies lots of cross-city traffic, but it’s mapping out in percentage terms the demand from different directions heading to the Eastern Freeway in the AM peak — in other words, feeding into the freeway in the counter-peak direction, as if counter-peak travel is where the congestion problem is.

A diagram on page 41 does look at AM peak from the Eastern Freeway, and like previous studies shows little traffic heading to the west of the city — 2% to the south-west (eg Newport area), 6% due west to Footscray and beyond, 7% north-west to around Essendon and beyond. The vast majority of traffic is heading to the CBD and inner north.

In comparison, here’s the screendump from VicRoads traffic status web site this morning (8:16am, peak hour). The camera image shows counter-peak the Eastern Freeway seems to be free-flowing. It also shows free-flowing traffic most of the way across to the west (in both directions), again underscoring that the east-west route isn’t the main problem; as per the page 41 diagram, it’s traffic going into the CBD and inner suburbs.

EWLink: Realtime traffic on Eastern Freeway and Alexandra Parade, 15/12/2014 8:16am

p100 forecasts traffic rampup to % of steady state volume: 91% by month 6. 96.5% by month 12. 100% by month 22. I wonder: Is this in line with recent experience?

A NSW Auditor-General report on Sydney’s Cross City Tunnel (see page 32) found that projections of 80% initially, and 88% after a year were about double the traffic levels that actually eventuated. Brisbane’s Clem7 and Airportlink tollways, and Melbourne’s EastLink had similar problems.

Note that in East West Link’s case the taxpayer bears the risk.

p165 Whoa! The construction cost is much much higher than theoretical revenue of $112 million/year (56x) relative to Citylink (8x) or Eastlink (20x). The average construction cost is also much higher per kilometre than those projects.

p168 The assumed tolling period is 40 years.

p176 Benefit Cost Ratio of stage 1 is 0.8 (eg it costs more than it makes) when “Wider Economic Benefits” (WEBs) are excluded. Including WEBs is 1.3-1.4.

Update: The earlier estimate, using the methodology preferred by Infrastructure Australia, came out at just 0.45. In later versions of the document, the methodology changed and the estimate rose to 0.8. The version released by the Herald Sun has the higher figure, and it’s been speculated that someone supportive of the project dropped that version to them deliberately to pre-empt reporting of the lower figure. Josh Gordon at The Age has some nice analysis of how the figure grew from 0.45 to 0.8 with some WEBs, and then to 1.4 by including other projects such as the Tullamarine Freeway widening, and even Wider WEBs.

WEBs are notoriously wibbly-wobbly in their calculation, and often controversial. For instance it’s not clear how they claim $2153m in agglomeration economies (specifically “growth in Melbourne’s competitive central core”) when the tollway doesn’t directly serve Melbourne’s central core.

It also claims a lot of benefits from travel time savings, but as I’ve already noted, we know these never last.

Compared to the 1.4 the road gets with WEBs, the metro rail tunnel (which is also an incredibly expensive project) apparently got 1.9. And compared to the 0.8 for EWL without WEBs, the metro rail tunnel got 1.17 — so at least it isn’t loss-making when evaluated without possibly dodgy WEBs.

p193. If they built the road elevated rather than underground, the BCR (excluding WEBs) would still only be 0.9. It’s only by building it as a surface road (eg a ground-level motorway, thus obliterating large areas of the inner-northern suburbs) that you can get a BCR above 1: 2.6 to be precise.

EWLink costs and revenues

p209 summarises the revenues and outlays, and if I’m reading this right, seems to show toll revenue of about $200m per year against availability service payments from the government to the operator of about $345m each year. I assume by June 2023 that’s the “steady state”.

If the toll revenue doesn’t get that high, then taxpayers foot the larger bill. And remember this is only stage 1.

p211 ponders the state privatising the road later — that is, selling the toll revenue stream, presumably to offload the taxpayer risk in case revenue flops in the future.

I’ll keep dipping into the document as I get time in the next day or two, and may add some points as I find them.

Hopefully when there’s an official release in the next day or two, the PDFs available will be searchable — it’ll make finding things a lot easier!

And presumably there’s more detail coming as well — for me one thing that stands out is the courageous predictions of quick growth in tollway traffic and revenue, in the face of recent experiences with other Australian tollroads.

And I’d love to see detail on the modelling assumptions that show how well the traffic would flow if the revenue targets are met. It still strikes me that these massive tollroad projects can be profitable, or provide for free-flowing traffic, but not do both.

Update: The official release of documents has now occurred.