Wellington Scoop

Analysis of inner-city bypass shows cost escalations and lower benefits

A paper presented to this week’s transport conference in Wellington states that Wellington’s inner city bypass had “fewer than expected beneficiaries and lower than expected benefits.” And while safety benefits were expected from the bypass, the result was the reverse – “disbenefits, involving crashes between pedestrians and vehicles.”

The paper was presented to the conference by Dr Nadine Dodge, a Senior Advisor on transport strategy at the Wellington City Council.

This is her conclusion:

A number of national and international studies have indicated that ex-ante cost-benefit analysis tends to under-estimate project cost and over-estimate project benefits, resulting in inflated cost-benefit ratio estimates that negatively influence decision-making in the transport sector. Failure to account for induced demand has been identified as a key source of underestimation of travel demand and overestimation of project benefits

The present study has conducted an ex-post appraisal of Wellington’s Inner City Bypass. Consistent with other studies, final outturn costs for the Inner City Bypass were substantially higher than initially estimated.

While international studies have shown that induced demand caused by roading investments frequently results in an under-estimation of travel demand, in the present case study travel demand was lower than forecast.

Because increased demand is proportionate to the percentage increase in transport capacity, the Inner City Bypass may not have resulted in measurable levels of induced demand due to its relatively limited scope. Lower than expected travel demand allowed for sustained travel time savings, but also resulted in fewer than expected beneficiaries and lower than expected benefits.

Analysis indicates that in the highlighted case study, the ex-ante BCR was between 3.3 and 8.5 times higher than the actual result, calling into question the reliability of ex-ante forecasts as a decision support tool. This demonstrates that both underestimation and overestimation of travel demand can reduce transport benefits, highlighting the need for accurate forecasting of travel demand if CBA is to be an accurate decision support tool in the transport sector.

The most inaccurately forecast benefit category was safety, where ex-ante appraisal indicated substantial benefits but there were actually disbenefits. This may have been due to changes in road layouts and traffic directions, resulting in crashes between pedestrians and vehicles.

This outcome indicates that future roading improvements should take care to avoid safety disbenefits and impacts to vulnerable road users, especially in urban environments.

In the absence of certainty surrounding travel demand forecasts, sensitivity testing of proposed investments under a wide range of future travel demand scenarios can provide greater certainty investments will deliver value for money under a range of possible futures.

The present study also suggests that the routine usage of a +-20% margin of error is inappropriate, as the actual margin of error appears to be much larger.

While the NZTA post implementation review process assesses investments relative to their forecast performance at the time of funding approval, the current review has indicated that using time of funding approval at the baseline year may be problematic.

CBA is frequently used to decide between alternative investment options well before final approval is given, and the accuracy of CBA at this stage of a business case is important to ensure that CBA assists with identifying the best performing option.

The case of the Inner City Bypass has demonstrated that costs escalated considerably between the identification of the preferred option and ultimate funding approval, resulting in a substantially reduced BCR. While a post implementation review methodology may have indicated that the project performed well relative to forecast performance at the time of funding approval, an ex-post analysis compared to initial CBA forecasts tells another story.

Conducting post implementation reviews relative to CBA estimates at the short list or programme business case stage may provide better insight into the overall reliability of CBA and its usefulness as a tool to decide between alternative investment options.

Ex-post analysis is relatively infrequent in New Zealand, and NZTA reviews are confined to a small percentage of small and medium sized projects. There is therefore substantial scope to expand the usage of ex-post analysis to ensure that ex-ante appraisal techniques accurately predict investment performance.

The case of the Inner City Bypass suggests that more frequent use of ex-post CBA can inform the accuracy of transport investment decision making in New Zealand


  1. Patrick Morgan, Cycling Action Network, 7. March 2019, 17:01

    This raises questions:
    How can we trust NZTA BCRs?
    What are the implications for other projects?
    Why does NZTA not routinely check BCRs after construction?

  2. Graham C Atkinson, 7. March 2019, 17:54

    Remember the Inner City Bypass was part of a broader concept including the dual Mt Vic tunnels (4 lanes to the airport); obviously when only part of a concept is implemented the benefits will be less.

  3. Andy Mellon, 7. March 2019, 19:09

    @Graham – that wholly depends on what basis the original BCR was calculated. The article heavily implies that the original BCR was calculated just for the inner city bypass and not including any of the benefits in the broader concept. As such, the post-implementation review conducted would be like-for-like and thus appropriate.

    More surprising to me is that NZTA are not conducting post-implementation reviews for all major investments as a necessity – a 1 year PIR, 3 year PIR, 5 year PIR and 10 year PIR would seem to be vital in both assessing NZTA performance but also to ensure under or over costings can then be taken into account when constructing subsequent business cases for investment.

  4. Isabella, 7. March 2019, 19:22

    Try reading the paper Graham. That section had its very own BCR which was able to be analysed on its own terms.

  5. Mike Mellor, 7. March 2019, 19:30

    Graham: the point is that NZTA significantly overestimated the BCR used to justify this particular project (and included safety benefits, when the project actually made the road network less safe), and it has taken an outside observer to identify this flaw in the process.

    Any benefits of a broader project are a different matter altogether.

  6. glenn, 8. March 2019, 10:02

    That’s if you assume her conclusion is correct. Ask everyone travelling through the city if there is any benefit.

  7. Dave B, 9. March 2019, 15:34

    One major area of benefit which the Inner City Bypass fails to deliver is to Vivian Street and Kent Terrace. Why does the southbound State Highway route continue to dog-leg via these roads? The Terrace Tunnel is bi-directional, but the ICB trench is not, even though they are the same width (I can assure any inclined to argue this point that the trench is 12.7m wide as measured at the Buller Street overbridge – i.e. wide enough for the same 2+1 bi-directional lane-configuration as the Terrace Tunnel).

    Why do we not extract full benefit from this expensive but under-performing asset by accommodating both directions of SH1 along the Trench > Karo Drive > Arras Tunnel route?

    The cynic in me reckons that because NZTA still quietly hankers after its grand scheme of duplicate tunnels and Basin Flyover, it does not actually want to improve things now as this would weaken its case for that. If there is a more practical and honourable reason why the present shambles has persisted for so long I would really love to hear it.

  8. steve doole, 10. March 2019, 10:16

    BCR is based on outdated thinking.
    Benefits include savings on “travel time” that a project will enable, but the value of everyone’s time is not treated equally. If you drive a vehicle the time reduction for you will be valued much higher than for me, as I don’t have a car. And so no surprise – projects that have high BCR are usually roads that accellerate car use (mentioned as ‘induced demand’ in the article, but there is increased car use due to other project outcomes as well.)
    Theorectical savings in travel time on roads without intersections and pedestrians are going to be higher than suburban and city roads. And proposed roads with high speed limits are forecast as though all vehicles travel at the speed limit, which is different from reality, but increases the calculated benefits.
    For 50years NZTA and predessors have used this skewed version of BCR to justify road expansion. Now NZ has highest car use per person in the world, bar USA and some tiny places.
    Government should direct NZTA to use a BCR that value’s everyone’s projected time savings separately from the way they are forecast to travel.

  9. Patrick Morgan, 12. March 2019, 8:36

    She used the same evaluation methods Transit used, but with accurate data. BCR was much lower. The paper is well worth reading. [via twitter]

  10. Iona Pannett, 12. March 2019, 12:57

    Cold comfort indeed that we were right on the Bypass. Time for a new vision on transport. Motorways are very 1960s. [via twitter]

  11. Andrew, 12. March 2019, 17:03

    Iona, the bypass would have been so much better as a cut and cover, which I believe was proposed. Local north/south movements would have been unhindered by the bypass.

    As far as I know A. cost and B. protest put that idea in the bin and left us with the half arsed result we live with now.

  12. Filosofos, 16. March 2019, 0:10

    If you exclude the atypical examples of Edinburgh and Sydney, light rail projects around the world almost always end up with much higher BCRs than predicted. Something to think about…

    It’s a win-win: Beneficial for pedestrians and drivers too (by decreasing demand on roads). Oh, yes, and the climate, too. Is this considered in BCRs?