A few months back when Auckland Council and the government released vastly different results from their (supposedly) joint review of the City Rail Link’s business case, we learned that the process of determining whether a project is “worth it” is not the objective exercise that is so often portrayed. The government’s review said the project’s cost-benefit ratio was 0.4 while Auckland Council’s review put that number at around 1.1. As both parties were fairly agreed on the costs, it was debate over the magnitude of the benefits that really led to the disagreement. As an interesting sequel to this debate, a recent paper discussed at the Council’s Economic Forum today, put together by Auckland Council’s chief economist and based on work done by the NZ Institute for Economic Research that looks at a key component of this debate: the ‘discount ratio’ used. The concept is discussed below:
A country’s social discount rate policy should attempt to solve two distinct problems (with ideally two distinct instruments). The first is to take account of an infrastructure project’s ‘wider economic investment’ effects. This means ensuring that public projects do not crowd out/replace even more profitable private sector projects, as measured by their social opportunity cost – as well as taking account of how productivity improvements stimulate more private sector investment. These wider investment benefits and opportunity costs can differ significantly between projects, depending on how the original investment was financed and the kind of benefits they provide. Taking account of these features is critical to ensure value for money.
The second problem is determining the social rate of time preference. It is well accepted by economists that individuals and society in general, place a higher value on benefits and costs that occur in the near future vis-a-vis those that occur many years into the future. Consider for example, asking a two-year-old if they would like a lollipop today, or two lollipops in a years’ time. The social discount rate makes a judgement on the value of benefits and costs received in the future. This “time value” judgement is critical for the aspirations of sustainable investment in New Zealand’s transport network and our corresponding level of wealth over time. The social discount rate currently used in the assessment of transport projects is 8% (real, meaning net of general price inflation; all discount rates here are expressed in real terms).
This is a somewhat overly complex explanation in my opinion. My understanding of discount rates is to ensure that we know it’s better of putting money into a project than just chucking it in the bank. Having $1000 of benefit now is more valuable than getting that $1000 of benefit next year, because if you put your money in the bank you’re going to earn something off it.
In any case, the main debate in the paper is not over whether we should both having a discount ratio, but rather what that ratio should be and whether the benefits should be “chopped off” at some point in the future. Different projects have different types of benefits, some which happen quicker and some which last longer. A useful example of how two projects (one with short-term benefits and one with long-term benefits) compare is outlined below:
This 8% real social discount rate means that policy-makers equate a unit of benefit2 received in 30 years’ time, with 6 cents received today. A reduction in the social discount rate would alter this equation significantly. For instance, at a discount rate of 4%, policy-makers would equate a unit of benefit received in 30 years’ time with 31 cents received today (so benefits in the future are given a higher value). This has important implications for the type of projects approved for funding. A lower discount rate changes the prioritisation of transport projects from those that have benefits in the short term to projects that have benefits further into the future.
For example, Figure 2 illustrates this for two hypothetical projects that are mutually exclusive (that is, doing one means you would not do the other). Project A (e.g. widening a road) realises benefits immediately and sustains them indefinitely, while project B (e.g. implementing a rapid transit corridor) has smaller benefits in the short to medium term but substantial benefits in the longer term.
The decision for which project to invest in lies, in large part, in the value society places on future benefits – a judgment made implicitly by the social discount rate. Using New Zealand’s current discount rate of 8%, project A is preferred over project B, because the benefits from project B are realised in the distant future, which is heavily discounted. In fact, transport project evaluation is currently capped at 30 years because any benefits realised after this period are often discounted to zero. This means that the majority of the benefits of project B are excluded in the benefit cost analysis. In contrast, using a discount rate less than approximately 5% means project B would be preferred over project A.
Here’s the image:What gets really interesting is when this theoretical debate gets applied to real projects, like the City Rail Link – which has very long-lasting benefits, but benefits that accrue somewhat slower than smaller projects like widening a road.
Figures 3-6 below illustrate this point using the City Rail Link (CRL). Figure 3 shows the undiscounted benefit stream of the CRL. Notably, the project’s benefit stream is very similar to the long-lived project in Figure 2 above (project B), where benefits are shown to increase into the future. Figure 4 shows the benefit stream of the economic appraisal using a standard New Zealand appraisal following the Treasury guidelines of an 8% discount rate and a time period of 30 years. This gives a discount benefit stream of 1.8 billion and a Benefit Cost Ratio (BCR) in the order of 1.1. However, if the appraisal period was lengthened to reflect the long-lived nature of the CRL, then the benefits attributable to the project would increase to 2.7 billion and the BCR would increase to a figure in the order of 1.5. This is shown in Figure 5 using a 60 year appraisal period. This example shows how current methodology is biased against long-lived projects, because many of the future benefits are simply not included in the standard evaluation.
The graph showing the gradual but consistently growing benefits of the City Rail Link is a very good match with the theoretical “project B” outlined in the graph above:60 years into the future the benefits of the project are quite truly massive, and still growing. This is not really too surprising, the New York Subway continues to provide that city with enormous benefits that grow each time its ridership grows – even though the subway opened over 100 years ago. That was an investment with benefits that have truly lasted the test of time.
Looking at the standard way of measuring benefits, you can see what a tiny fraction of the undiscounted benefits are actually caught:Of course it is appropriate to discount benefits and (possibly) to cut them off at some point. But the quantity of benefit measured using the standard approach seems to be completely tiny compared to the actual amount generated. Extend the appraisal period out (but keep the 8% discount ratio, and you start to capture some of the longer-term benefits of a project like the CRL: But what’s really interesting is looking at what happens when you combine both a lower discount ratio and a longer appraisal period to the process. Interestingly, this is exactly what’s done in the UK and a comparison of the results is shown in the diagram below – comparing the quantity of benefits of the CRL under the NZ and the UK system of assessment: The report goes on to note that the only differences are the discount ratio and the length of time the benefits are measured over:
Figure 6 then uses the CRL benefit stream (from Figure 3) and shows the difference in perceived benefits using the New Zealand standard appraisal and the United Kingdom (UK) appraisal methodology. The only difference between the two is the discount rate and the appraisal period. The CRL shows a present value benefit stream of 1.8 billion under the New Zealand methodology and a BCR in the order of 1.1. Under the United Kingdom methodology, the same project would be shown to have a present value benefit stream of approximately 11 billion and a BCR in the order of 5.2. Using the same benefit information, the UK methodology would place the benefits of the CRL in the order of 6 times as great as under the standard New Zealand methodology.
In other words, if we were undertaking a cost-benefit analysis of the City Rail Link in the UK, the project would have a cost-benefit ratio of 5.2 rather than the 1.1 the Council has measured it as having. That is an absolutely massive difference.
While I imagine transport projects generally score higher in UK assessments, meaning that the threshold for funding is probably a higher cost-benefit ratio than here, the key issue is that the City Rail Link is exactly the kind of project the NZ system is weighted against. If you compare how a project with much quicker (but shorter-lasting) benefits performs under the two systems, the difference is much smaller:So is the UK unusual in the way it measures projects, or are we? Looking an international comparisons it seems that we are the ones who apply an unusually high discount ratio:The report to Council’s Economic Forum proposes that the Council lobby central government to relook at this matter, so that better consideration of long-term projects can be made. I certainly hope the issue is looked at closely so that we’re brought in line with international best practice and to ensure that the full benefits of projects with long lifespans can be given adequate consideration.