One thing that came across most clearly in the excellent Radio NZ piece about the City Rail Link project that played on Sunday morning was that the difference between the government’s measurement of the project’s cost-effectiveness and the assessments undertaken for Auckland Council doesn’t just come down to whether one is right and one is wrong. This is a subjective business, based on what you include or exclude, whether you are optimistic or pessimistic about the impact of the project on land-use change and, ultimately, whether you think there’s more to deciding whether a project makes sense or not than simply a single number – its cost-benefit ratio. I must say I have always wondered what the cost-benefit ratio of Britomart station would have been, as barely 1000 people a day were using the trains when it was opened – but in hindsight you could hardly say that it was a mistake or a project that “didn’t make a difference”.
Of course that’s not to say that cost-benefit analyses are pointless exercises. They do provide a very detailed analysis of the impact a project will have, and when you’re spending large sums of public money you need to know that the money is not going to be wasted. In my mind, the real issue relates to what the process measures and (perhaps more importantly) what it doesn’t measure. Radio NZ’s preview of their Sunday piece about the CRL highlighted this debate:
The Transport Agency is taking a new look at how it measures the benefits of major projects.
NZTA believes Auckland’s proposed downtown rail tunnel has only a quarter of the $1.3 billion of benefits claimed by Auckland Council – prompting criticism of the methods it uses. The tunnel would cost $2.4 billion.
The Government questions the claimed benefits, both the growth in rail patronage, and its economic impact in transforming Auckland.
Transport Minister Steven Joyce says Auckland has been tweaking the model used to measure the benefits.
Mayor Len Brown says a lot of the debate over benefits is crystal-ball gazing and government agencies have always underestimated Auckland’s growth.
Part of the reason the Government attributed fewer benefits than Auckland to the project was the method it uses.
NZTA takes a narrower view than used in some other countries for measuring wider economic benefits.
But PricewaterhouseCooper director Chris Money says work done in Britain found that traditional ways of measuring, undervalued urban public transport.
Business commentator Rod Oram says that unlike the Transport Agency, the Auckland Council case measures economic growth from a more dynamic downtown area. He thinks the agency’s focus is wrong.
NZTA has now put six research projects looking at new ways of measuring major transport projects, out for tender.
Auckland Council will next year prepare a more detailed case for funding the project.
More on the debate over Auckland’s downtown rail tunnel can be heard on Insight on Sunday.
The really interesting thing is that NZTA are looking at this issue of whether they measure the cost-effectiveness of transport projects in a way that accounts for full benefits – particularly in terms of how public transport project benefits are measured. I found the part of the show which discussed how the UK has changed their approach over the past few years, as well as debates about different discount ratios, quite interesting. It seems pretty crazy to me that we completely ignore project benefits more than 20 years in the future – the New York subway had many of its major lines built over 100 years ago yet they still provide massive benefits right to this day.
AKT reported on a couple of NZTA’s proposed studies a while back, and one has to assume that this is the same issue that NZTA is looking to get a better understanding of. In fact, if you look at a board paper about what NZTA thought of the City Rail Link review process, it seems that they were pretty sceptical of the Ministry of Transport’s extremely pessimistic position on the project (I have highlighted what I think is the key sentence):
I’m hopeful that some of the research projects that NZTA seems to have underway into how they measure the benefits of transport projects may shed some light on this issue, bring our system in line with best practice overseas and hopefully mean that the divergent numbers of the Council and Government can be brought together.
Processing...
“It seems pretty crazy to me that we completely ignore project benefits more than 20 years in the future – the New York subway had many of its major lines built over 100 years ago yet they still provide massive benefits right to this day.”
The trouble is that when you discount a $1m projected benefit 100 years in the future for the benefit-cost ratio, it’s only worth $335 in today’s dollars (at 8%). This doesn’t help the business case, even though as you’ve said there are massive ongoing benefits.
Absolutely true Chris. Of course if you don’t build it now because the future benefits don’t stack up it gets more expensive in nominal terms….the original cost of the Morningside Deviation was two million pounds, which would be a laughable amount now. Unfortunately the Reserve Bank’s calculator doesn’t go back that far to give us a modern value on this amount, but it wouldn’t likely be $2.4 billion.
If you’d invested 2 million pounds in 1930 and managed 10% rate of return (which is a bit higher than you’d get now, but inflation was running around 15% for bits of the 1970s and 1980s), then your investment would be worth 4 billion pounds now. Or $8billion after decimalisation. Enough for the CBD tunnel, rail under the harbour to Albany, and rail to the airport.
It is always worth putting off capital spending for the future if you can. In 100 years time average incomes will be several multiples of what they are now, and future people will have access to all sorts of cool technology.
Given the government can generally borrow at the lowest rate 10% is way too high. And it’s not like the government of the day had 2 million pounds looking for a home that they could have invested or build a rail tunnel with so any actual or imagined return on funds invested is not relevant. The rate of return of the project and the cost of borrowing is relevant.
Actually you can do a reverse look up- 2.4 billion is 30 million pounds converted 2011-1920. Rate of inflation for the period is 4% (while there was high inflation for the 70s and 80s there was also very low and negative inflation in other periods).
http://www.reservebank.govt.nz/statistics/0135595.html
Are you basically saying we should never build any capital item because it will be cheaper in the future?
Commercial investment should always aim to earn significantly more than the rate of inflation. If it didn’t there would be no point in investing or creating new businesses.
You should buy the capital item when you need it and not before. I can’t find figures for NZ, but real (not nominal) incomes doubled in the States between 1950 and 2000. Assuming NZ is about the same, then real incomes in 1930 are about a third what they are now. Much easier for me to pay for a project than it was for my grandparents. That’s why discount rates are important in business cases.
“You should buy the capital item when you need it and not before”
You mean like last Friday? For something that will take years to build you have to plan ahead just a little.
Obi I certainly agree with you regarding the net present value issue. Of course we should have a discount ratio – the real question is whether 8% is too high and what the effect of having such a comparatively high discount ratio is on promoting some transport projects (ones with short term benefits) over other types of projects (ones with much longer term benefits).
It sounds like we’re a bit out of step with how the rest of the world looks at the assessment of transport projects, and we should wonder why this is the case and whether we should change.
Given the long run rate of 4% for inflation (though that is lower more recently) and a historical real rate of return (risk free) of 3.5%, 8% is about right.
Admin… I agree with your point and also the point the quoted article is making. My gut feeling is that all the transport projects that are seriously on the table at the moment are worthwhile both economically and for other reasons (such as resilience of infrastructure). But many of them show a BCR less than 1. I’m not sure that matters as long as they evaluated on the same basis… which was the problem with the City Council tunnel business case. But, full quantification of economic benefits and a longer term view probably makes it easier to sell the benefits of infrastructure development IF the same methodology is applied to all proposals.
Obi I think it’s accepted by all parties (except perhaps APB&B) that the original business case for the City Rail Link had some flaws. The real question is whether you believe council’s review of the business case or the MoT’s. NZTA thinks the “real” answer is somewhere in the middle.
I wonder if there is a ‘real’ answer, or whether too many of the inputs are subjective and subject to politics (in a quite legitimate way). So there’s an argument that it might actually be nigh on impossible to apply the same methodology EXACTLY to all proposals.
But of course no problem to build the whole motorway network twice in a short space of time…. the NY subway system could do with a few upgrades but the expensive bits, the actual tunnels, function for a very long time; much of London’s network is Victorian. Of course the CRL will repay it’s investment for decades and decades.
I’d be interested to understand more of the rationale behind the costs that get included too. The City Rail Link business case has costs for building the tunnel, but a substantial amount is allowed for new rolling stock and expanding other lines. It also allows for all of the costs, like on-going maintenance, staffing, etc. This all seems fairly reasonable as you need all these things to operate.
When it comes to roading projects, do they also include equivalent costs? Does a project like Waterview or Pu-ford include all of the additional vehicle costs that will be incurred? More cars means more maintenance, WOF checks, Police, ACC levies, etc. You need all of these things for a motorway to be useful, so do they get included? We’re going to end up paying them to be able to get any benefits, so I hope so!
Arthur Grimes of MOTU (and a director of the Treasury National Infrastructure Unit) has some interesting perspectives on these issues:
http://ips.ac.nz/events/downloads/2010/Motu%20infrastructure%20workshop/Arthur%20Grimes%20IPS%20&%20Motu%20Presentation%20July%202010.pdf
He offers a alternative view on the value of large infrastructure investments, in that such projects such as Broadband, large transport schemes etc can result in impacts that cannot be envisaged or fully quantified at the time that a project business case is made. Therefore there can be an option value for projects which may counter the traditional view that it is better to defer expenditure if costs or benefits are uncertain.