Exhibit 1: This shows average air fare yields for non-business travellers (bargain hunters), which is to say the average of all fares paid by this category of air passenger. It shows the average fare between Sydney-Brisbane for example to be approximately $55, and Melbourne-Brisbane approximately $95.

Guest writer Gerard Drew is Research Director with Beyond Zero Emissions (1):

After 30 years of discussion on high speed rail in Australia, and not a kilometre of track built for it, a cynic might assume that it’s never going to happen. But that would ignore what progress is being made overseas and recent analysis that changes the outlook for high speed rail in Australia.

The Zero Carbon Australia High Speed Rail report was published last week. It is a collaboration between think-tank Beyond Zero Emissions (BZE), the University of Melbourne Energy Research Institute, and the German Aerospace Research Center.

The most important finding in the report relates to fare pricing assumptions associated with the estimation of high speed rail (HSR) commercial revenue. The point of difference is the air fare benchmark used as the competitive price level for HSR. This is a key point I believe was overlooked and misunderstood in the article on The Urbanist on April 15, High Speed Rail: too good to be true?

In the Phase 2 report produced by AECOM et. al. for the previous Federal government (completed in 2013), these air fare assumptions are not explicitly declared in the whole of the 534 page report, nor its 2,500+ pages of appendices, so some confusion is understandable.

We investigated this issue in some depth, as we understand it is critical to understand HSR revenues. We also analysed it at length as part of our submission responding to the Phase 2 report.

For brevity I will discuss only the non-business fare category, however business fares are also assessed in our submission. What we identified is that in the Phase 2 report, average air fare yields were significantly underestimated in relation to actual market prices. Average fare yields presented in Appendix 1F of the Phase 2 report are shown in the first exhibit.

This shows average air fare yields for non-business travellers (bargain hunters), which is to say the average of all fares paid by this category of air passenger. It shows the average fare between Sydney-Brisbane for example to be approximately $55, and Melbourne-Brisbane approximately $95.

Concern may be immediately apparent to regular flyers that this is a somewhat below average estimate. To verify this, we surveyed the air fares for the five air routes indicated over six months, June-December 2013. Incidentally, this is a period in which the three major airlines – Qantas/Jetstar, Virgin Australia and Tiger Airways – all reported significant losses.

The results of our survey confirmed that the Phase 2 report was inaccurate, as shown by the sample of a single day’s flights shown in the second exhibit.

Exhibit 2

The dots indicate the cheapest airfare available on each flight from Sydney to Melbourne over the course of one day – disregarding any additional baggage, credit card or other fees which may be incurred.

A few lines are also shown. The load factor average ($145) accounts for the fact that not all flights are full (only 80% capacity for this route); being the mean of all but the most expensive 20% of flights.

The floor average ($118) accounts for the split between non-business and business passengers (a 60-40 split according to Tourism Research Australia data); this being the mean of the cheapest remaining 60% of flights. This is the lowest possible average fare yield for bargain hunters on the flights available on this day, hence floor average.

The last line, in red, is the average fare yield indicated for non-business passengers in the Phase 2 report ($70). This is 40% lower than the lowest possible that we identified on this particular day of flights. This discrepancy is consistent across all routes over the six months of air fares surveyed (as detailed in our submission).

Passengers make their decisions on relative price, but the absolute value is what is important when calculating the resulting revenues. The underestimate is inconsequential for airline operators, who will continue to earn what they actually earn regardless. But when HSR fares are constructed to be comparable to such unrealistically low air fares the resulting HSR revenue is correspondingly underestimated to a large degree: passenger numbers are estimated to be in the tens of millions in both studies.

The additional revenue resulting from aligning HSR fares with realistic air fares has been shown to be sufficient to repay the estimated capital cost. This “astounding” finding may be unexpected and go against intuition, but this does not make it incorrect. It is the result of thorough analysis and should be judged on the assumptions and methodology used to determine it (detailed in our Phase 2 submission).

Flow-on considerations

As this substantially alters the commercial case for HSR (allowing capital repayment), other assumptions affected by it are also worth testing such as advancing the implementation schedule.

Although not stated explicitly in the Phase 2 report, the main drag on implementation is financing. When there are insufficient financial returns, government financing is required (as is assumed in the Phase 2 report), which leads to lengthening the implementation period so as not to overload annual budgets. This situation changes considerably when there is a prospect for commercial returns and the private sector can contribute to financing.

A ten year implementation was noted in our report as an ambitious target in-line with what has been accomplished internationally. While the evidence of what has been accomplished in Taiwan, Spain and China is undeniable it was also noted in the report that a staged implementation may be more practical.

Our passenger estimates differ from the Phase 2 report, referencing different points in time, utilising two different models constructed independently. What is surprising perhaps is that the two results are so similar. In any case, it is not incumbent on BZE to issue a thorough comparison of results to the Phase 2 report. Nevertheless, a comparison of estimated passenger kilometres is available should anyone request it. It indicates a very similar passenger volume in the early stages, with the BZE estimate actually growing more conservatively over time.

Exhibit 3

Costs associated with construction are many and varied as well as highly sensitive to terrain. The resulting cost estimate is the result of many months of topographical optimisation and savings in comparison to the Phase 2 report are likely to result from an accumulation of reductions in complexity. It has been incorrectly stated that cost savings have resulted from sharing tracks with commuter services. The HSR system proposed is completely independent of the commuter rail system, with no track sharing. A number of locations were identified were the required land space was available in the easement of existing rail infrastructure corridors to accommodate HSR tracks. The synergies associated with utilising existing rail corridors are manifold, but shared track is not part of the plan.

Lastly, muddying the distinction between facts and estimates is a subtle bias introduced into this commentary. BZE is very clear to identify where figures are estimates yet these have been misrepresented in the presence of what are referred to as “facts”. Simply because one groups estimates fit more neatly with a pre-determined view than another’s does not make it a fact – no matter how blunt your axe is.


  1. AECOM has been offered the opportunity to respond to BZE’s criticism of their methodology; so has the Department of Infrastructure and Regional Development.