The Response to Submissions on the New M5 EIS has been released.

It is a mix of unjustified assertions, claims that are reversed in some other section, misleading statements, and statements and calculations that are outright wrong. Many objections to the M5 are not treated seriously, or are acknowledged but not discussed, or are not even acknowledged.

There are however two important admissions made, even if those admissions are not followed to their inevitable rational conclusion.

The major sources of concern for Alexandria residents are:

1. the ill thought through plan to turn Euston Road into a busier route than Victoria Road, and about the rat-running in local streets that will inevitably occur.
2. Residents are concerned about the loss of amenity in Sydney Park, and along Euston Road.

There is also the question as to whether putting the planned volume of cars on what is already a heavily congested road is even possible.

The Response to Submissions acknowledges these concerns, but dismisses them:

“The widening of Campbell Street and Euston Road would create a more efficient through route adjacent to the industrial development to the park’s south and east, and would divert traffic from the Princes Highway and Sydney Park Road”

This is patently false on numerous grounds:

1. Traffic at the north end of the Princes Highway (King St) is generated by areas to the north, not the east or south. Euston Road does not offer an alternative route.
2. Widening of Campbell Street and Euston Road does not (in theory)extend beyond the immediate vicinity of Sydney Park. The roads these roads feed are already congested, and congestion already banks back on Campbell and Euston Roads. Widening them will not make them more efficient. This will still be true even if there is some undisclosed plan to widen Euston Road for its entire length (which presumably there is, given the existence of road reservations. If so, these plans need to be disclosed now).
3. Even if Campbell Street and Euston Road could be made more efficient, the law of induced traffic predicts that this will increase traffic by a quantity equal to the extra capacity, meaning no reduction in the quantity of traffic on Princes Highway and Sydney Park Road, just an increase in traffic on Campbell and Euston Roads. Sydney Park would go from being bordered by heavy traffic on the north and north west, to being bordered by heavy traffic on the north, north west, west, south and east.

Similarly fallacious logic is also on show in the following claim:

“Quicker, more reliable trips between Western Sydney and the Sydney Airport/Port Botany precinct to support Sydney’s urban freight movements, through a doubling of capacity in the M5 Motorway corridor”.

This claim is worryingly misinformed:

1. The New M5 does not connect Western Sydney to Sydney Airport, Port Botany. It does not increase capacity in the existing M5 corridor.
2. Until the M4-M5 link is built, the extra capacity created in the new M5 corridor will not relieve demand in the existing corridor because it does not serve the same destinations as the existing corridor.
3. Until the M4-M5 link is built, there actually won’t be an effective increase in capacity in the new M5 corridor because the roads around the St Peters interchange are already bottlenecked and will continue to be bottlenecked. (In the words of the local member of parliament, “the New M5 is a tunnel to a traffic jam”.)
4. Even once the M4-M5 link is built, it will not be the extra capacity created in the new M5 corridor that will, perhaps, relieve demand in the existing corridor, because any temporary drop in traffic volume would be replaced by induced traffic, if it were not for the tolls.

What is not acknowledged in the Response to Submissions is that many of the claimed benefits to Alexandria and nearby areas arise not from the New M5 itself, but from tacked on activities.
For example:

• “Extend Campbell Street west over Alexandra Canal to Bourke Road”,
• “dedicated pedestrian and cycle facilities” and
• “upgrades of the drainage and flooding provisions”

Without arguing the merits of these projects here, they can be done without building the New M5.

Also separable from the New M5 works is the claim that it will result in:

“A reduction (almost by half) in average travel times … along with a doubling of average travel speeds”

The EIS itself shows that the forecast increase in congestion is attributable not to the extra capacity but to the reintroduction and expansion of tolls.  Without arguing the merits of tolls here, tolls may be imposed without building the New M5.

What is not separable from the New M5 is the M4-M5 link. Full credit to the authors of the Response to Submissions, it is acknowledged that the benefits of the New M5 depend on the delivery of (at least) the M4-M5 link, or some “equivalent alternative project”:

“The traffic impact assessment carried out for the project (Chapter 9 and Appendix G of the EIS) identified that additional road network augmentation beyond completion of only the New M5 would be required by around 2031. At this time, traffic forecasting indicates unacceptably low levels of road network performance based on available network capacity (including the extra capacity generated through completion of the New M5) and from continued growth in traffic demand. Should the M4-M5 Link not proceed, traffic forecasts indicate that an equivalent alternative project would be required by around 2031 to avoid unacceptable deterioration in road network performance.”

However, the Response to Submissions does not follow this to its logical conclusion: because the benefits for the New M5 depend on the delivery of the M4-M5, the two can only properly be considered together. Only by comparing the combined benefits with the combined costs can the overall merits be properly assessed.

Otherwise there is a risk that the net benefits of the New M5 (if any) will be more than offset by the net cost of the M4-M5. In such a case,it would be irresponsible to build the New M5. Alternately, it may be that neither the M4-M5 nor New M5 are, on their own, of net benefit, but that collectively they are. (It seems unlikely, but let us consider the possibility.) In such a case, it would be a missed opportunity to consider the New M5 and the M4-M5 separately. Only if the New M5 makes sense on its own should it be considered on its own, and the Response to Submission acknowledges that it is not viable as a standalone project.

What should however be considered separately are the benefits attributed to the different projects that compose the WestConnex. Prima facie, the greatest benefits will come from the widening of the M4 and M5, which are also the lowest cost parts of the whole project, while the New M5and M4-M5 are simultaneously the most expensive, highest risk, and prima facie, least beneficial components of the whole project.

Separating the benefits by sub-project may well reveal that only part of the whole project has a net benefit, and it would irresponsible to not consider this option, both because it may be that a subset of the whole project would have more value than the whole project, but also because it may be that while the whole may not be justifiable, some subset of the whole might be.

The Response to Submissions claims that the Strategic Business case described the project as “economically viable”:

“The economic analysis undertaken demonstrates that the WestConnex program of works is economically viable and would return $1.71 (2015 dollars) for every dollar invested.”

Duncan Gay has made a similar statement on 2GB:

“The state will get its money back.”

This a fundamental misrepresentation of the Strategic Business Case. Or perhaps, this is a fundamental misunderstanding of the Strategic Business Case. The Strategic Business Case does not predict a financial return of $1.71. It does not predict the state will get its money back. It predicts that the project will create $1.71 of benefits. Benefits are not revenue, and to suggest that they are, or that they are readily fungible, is wrong.

Benefits can be assigned an estimated financial value but this is only an estimate. With a very few exceptions, the benefits of this project are predominantly non-financial. In particular, the single largest portion of the benefit is the value of time saved.  This is notoriously hard to predict, both because of uncertainty over the quantum of time that will be saved, and because of uncertainty over the value people place on their time. As has been said in objections, the Strategic Business Case overestimates both.

The traffic forecasts should give no comfort. Traffic forecasting is unreliable even when done well, as a legacy of failed tollway projects attests. But the figures in the Response to Submission are particularly worrying, because many of them are patently wrong. The actual carrying capacity of local roads such Fountain St do not seem to have entered into the AECOM scenarios. Not only does the model predicting that roads, not bordered by road reservations, which are currently at capacity, will still carry orders of magnitude more traffic than they do now, both in the build and no build scenarios.

The model also makes other, even less credible predictions, for example that Fountain St, which is a short street that begins and ends at t-intersections, will somehow carry more traffic than one of these roads (Mitchell Rd), and almost as much as the other (Euston Rd). Further, the model apparently predicts that some local roads will carry more traffic in the 9 hours between 10pm and 7am than they will in the 15 hours between 7am and 10pm (e.g. Maddox St, 2031).

These blatant errors lead to the question: how many of the other figures displayed in the Response are less obviously miscalculations? Can any of them be relied upon for what is a $16.8B decision? It shows that the EIS and Response have not sufficiently investigated the effects on the local road network. Yet it is these effects that mean the full project will not be able to achieve its stated outcomes.

There is also no explanation as to why the 2031 figures ‘with project’ have been omitted, despite their omission being pointed out in submissions. This is a worrying lack of disclosure.

If somehow, regardless of all of the above, the total benefit were as high as estimated, benefits are not revenue. Benefits cannot be spent if they cannot be captured, and the benefits identified in the business case, if they do manifest, will be highly resistant to ‘value capture’.

A minority of drivers value their own time far more than the majority ($53.60/hour, vs $21.32 according to the Strategic Business Case). A toll which is low enough to be affordable to the majority of drivers will only capture a fraction of the total benefit created. A toll which is higher than that will cause a higher than predicted displacement of traffic, creating disbenefit, and forgoing revenue.

Based on the proposed toll regime, the estimated time savings, and the estimated value of time saved by class of driver, the proposed cost of tolls will exceed the value of time saved for most drivers. That it is to say, setting the toll as high as proposed will leave the majority of drivers worse off than they now are. If they continue to use the M4/M5,the extra cost of doing so will exceed their value of time saved. If they change modes, they will be facing a slower and more expensive trip than the one they now enjoy. Only those with the highest value of time will be better off, and the majority of that benefit will not be captured in tolls.

This is part of the reason that so many recent toll roads have been financial disasters.

Calculations based on average income overestimate the value of time saved to drivers, and overestimate the percentage of value that can be captured. To be credible, the calculation of value of time saved should be based not on the average income earner, or even the median, but on some lower percentile – lest one exclude too great a percentage of potential drivers, and one must remember that incomes in Western Sydney are lower than the average for Sydney as a whole.

The business case predicts a fall in usage of the M4 and M5. This drop will be a measure of the number of drivers for whom the toll exceeds the benefit of using the M4/M5 below zero, compared to other alternatives. It does not measure the number of drivers for whom the toll exceeds the value of time travel saved. They may continue to use the road, because it is still better than the alternatives, even if it not as good for them as the current regime. Only cashrich- time-poor drivers will be better off overall.

The drop in traffic is probably underestimated, unless tolls are to be set at a level significantly lower than the proposed level. But if tolls are set at a lower price point, to maintain something closer to the current traffic volumes, then the total take from tolls will still be lower than required by the business case, and furthermore, the promised congestion savings will disappear.

Even ignoring value capture, the forecast benefit to cost ratio of 1.71:1 is highly questionable.  It excludes indirect costs. It includes indirect benefits. Both the costs and benefits are optimistic. The estimated Value of Travel Time Saved and the forecast quantum of Travel Time Saved in particular are dubious.

Although this is not addressed in the Response, we have received a communication from Tim Reardon on this matter.  The communication from Mr Reardon does not address the surprisingly high Value of Travel Time or that percentage of ‘business travellers’ (who have the highest attributed value of travel time) is, 24% of all travellers, fully double the 8% to 12% recommended by Principles and Guidelines for Economic Appraisal of Transport Investment and Initiatives Transport; Economic Appraisal Guidelines. The claimed benefits should not be relied on until these two values are adjusted downward.

The communication from Mr Reardon does state that the time savings largely arise from motorists in the project area but not on the M4, M5, or M4-M5 link. This seems improbable.  The effect of the WestConnex projects will be to offload traffic from the M4 and M5 onto other roads in the project area, as is stated in the EIS:

“Traffic modelling indicates that some motorists could choose to use alternate parallel routes, such as King Georges Road, Stoney Creek Road, Forest Road and Canterbury Road, instead of the M5 East Motorway corridor.”

If the WestConnex project has an effect on overall congestion, it will be to increase it.  The Response to Submission disputes this. It states that traffic diverted from the M4 and M5 will not cause congestion on parallel arterial roads:

“This diversion of traffic onto arterial roads to avoid the toll is most likely to occur outside peak traffic periods, when there is generally spare capacity across the network. During peak periods these arterial roads are already at capacity and there is less incentive to divert from the motorway to avoid the toll.”

This claim is almost certainly true, and it is a crucial acknowledgement. In a congested environment, localised changes in road capacity result in little or no change in congestion on other roads.

Local arterial roads will not be made much worse during peak hours, because they cannot.  They may become just a little more congested, and the peak will last a little longer. But most of the traffic displaced from the M4/M5 will not appear as traffic on other roads. Some commuters will change to public transport. Some will work from home a little bit more.  Others will move house, or change jobs, or change where they shop for goods and for services. People will pack more activities into fewer trips.

The same thing happens in reverse when extra capacity is added. Uncongested roads encourage people to travel more, and further. Demand for travel is not fixed. It varies with the cost of travelling, both measured in dollars and in hours. Congestion cannot be ‘cured’ by adding capacity.

What is now needed is for this insight to flow through to the rest of the Response to Submissions. Just as raising tolls on the M4 and M5 will reduce usage of the M4 and M5 with little or no impact on congestion on other roads, increasing capacity on the M4 and New M5 will have little or no impact on congestion because the momentary reduction of congestion induces additional traffic until congestion returns to more or less where it was.

If anything, time lost to congestion will increase because there will be more people experiencing congestion. But it would be a mistake to think that this would be a net disbenefit.

If there are benefits to be achieved from WestConnex, they will come not from reduced congestion or reduced time travelling but from an increase in the number of people travelling and in an increase in the distance travelled. That the forecast is for less people will use the M4 and M5 after the project than before it is evidence that the claimed benefits in reduced travel time will not, for a majority of people, justify the extra cost. This project has far more losers than winners.

The way NSW estimates the benefits of road projects needs to change. The claimed benefits from reduced congestion are mythical. Road projects do not reduce congestion. Benefits, if there are to be any, will come from allowing people to live further from where they work and shop. The value of this is not easy to quantify, because urban sprawl has a complicated mix of costs and benefits that defy easy enumeration. But developing (or importing) a methodology that addresses this is necessary if we are to properly value the net benefits, or otherwise, of our future road projects. There is no value in a simple but wrong answer.