The Value of Time
by Peter Nunns
July 3rd, 2013
Readers of this blog will be familiar with the notion of the benefit cost ratio (BCR), a figure that compares the
forecasted benefits of a project with the financial cost of building it. It’s often used as a shorthand for the quality
of a project: If the BCR is high (i.e. substantially above 1) it is seen as a good use of public money; if not, it can
be criticised as a boondoggle.
Everyone plays this game. Opposition politicians often criticise motorway projects such as Puhoi-Wellsford and the
Kapiti Expressway on the basis of BCRs that fall below 1, while the Minister of Transport has in the past expressed
scepticism about the City Rail Link on the same grounds.
However, there is relatively little public discussion of the hows and whys of these seemingly consequential numbers.
How, exactly, does one calculate a BCR?
The procedures for conducting an economic evaluation of a transport project are set out in excruciating detail in the
Economic Evaluation Manual (EEM) published by the New Zealand Transport Agency. This manual defines the exact procedures that need to be followed when
evaluating any transport project and specifies the values that should be used in the evaluation.
Read it at your leisure (or peril). Here’s the summary version.
The traditional method of estimating the economic benefits of a transport project involves three main steps.
First, you need to define the project carefully – that is, you need to figure out what you are planning to build, when you will build it, and the new service patterns
that you’ll introduce as a result. Take, for example, the case of the Avondale-Southdown rail line, which is in the
Auckland Plan but hasn’t been defined carefully enough to enable us to figure out what it will do. We can’t evaluate
that until we know exactly what we’re building and when.
Second, you need to forecast the effects that the project will have on travel behaviour. Let’s stick with the example of Avondale-Southdown. In the short term, we might expect adding a new rail line to
encourage some people to switch from buses or cars, thereby reducing congestion, and to encourage some people to take
trips when they wouldn’t otherwise have travelled. In the longer term, it might change the patterns of population and
employment location, by making it easier for people to live and work in certain places.
This forecasting is typically done by regional transport models, which estimate (on the basis of existing travel
patterns and forecast changes to land use in a region) how people will get around in the future and how much time it
will take them.
Third, you need to quantify the benefits of changes to travel behaviour. This is where the EEM comes in. It summarises the types of benefits that you should expect from transport projects,
and defines values that allow you to monetise those benefits. Broadly speaking, the benefits considered in the EEM fall
into four main categories:
• Reduced travel time
• Reduced vehicle operating costs
• Reductions in accidents and health costs
• Reduced vehicle emissions.
(There are obviously a whole bunch of important things that are not assigned any value under this framework. We might,
for example, want a transport system that provides us with a choice of multiple modes, or increases our resilience to
shocks such as tectonic plate activity or sudden oil price increases. These benefits are often considered in other
stages of the evaluation but not quantified.)
NZTA considers travel time savings to be the most consequential economic benefit. Forecast travel time savings make up
the largest share of quantified benefits from most large transport projects in and around Auckland. This is based on the
idea that the time we spent in transit could be spent more productively on other activities. If we weren’t stuck on
congested road and PT networks, we would be working more, or doing things that we found at least as satisfying or
remunerative as work. For many projects, the time savings on an individual trip might be small – but they can add up
quite rapidly over large numbers of trips.
Take the case of the CRL. Removing the Britomart bottleneck is expected to increase capacity and hence frequency across
the whole network. This will make it quicker for PT riders to get into the city centre. The effects are larger on the
Western Line but still significant from the South and East, as this table released by Auckland Transport indicates:
Click for big version.
CRL time savings
In short, reductions in travel time are an important topic! So how, exactly, do we place a monetary value on them?
The answer is buried in the appendices to the EEM – Tables A4.1 and A4.2 to be exact. I’ve summarised some of the key
features in two handy charts. Unfortunately, the figures themselves present some logical conundrums.
The first chart compares the value of time assigned to trips on urban arterial routes and in rural areas. Urban travel
during the weekend is considered to be less valuable than weekday travel – which is fair enough, as most people work
during the week. But – perplexingly – urban commuter travel is considered much less valuable than rural travel of any
type.
In plain English, the EEM places a much higher value on the average JAFA’s time when s/he is on holiday in the Bay of
Islands than commuting across the Harbour Bridge to get to work.
Is this logical? It’s hard to say, because the EEM contains no attribution or explanation for these figures. They are
apparently derived from willingness-to-pay surveys , in which people are asked what value they place on their own time. But they don’t seem to bear any relation to
differences in productivity, which is another important measure of the value of time.
The econometric research suggests that urban areas in New Zealand, and in particular the Auckland city centre, have a
large productivity premium over other areas. Take, for example, the findings of a 2008 paper by David Mare (“Labour productivity in Auckland firms” ). Based on this, one would expect Aucklander’s time to be counted as relatively more valuable, not less:
So far, so strange. Now look at the second chart, which compares the additional value of time assigned to commuter trips
on different modes. The EEM places a much lower value on trips that aren’t taken in single-occupant cars. If you’re
walking or cycling rather than driving, your time is worth $1 an hour less; if you’re taking the bus or train, it’s
worth $3 an hour less.
Once again, it’s impossible to say whether or not these values make any sense, because no source or attribution is
provided. In theory, there might be some logic to these differences. For example, time spent in a bus might be less
onerous as it enables one to multitask while travelling. And people might not object to time spent walking or cycling
due to the health benefits. But counter-arguments could also be raised – many people enjoy driving cars and don’t mind
spending a bit more time on the road.
What can be said is that the values of time defined in the EEM have consequences. They allow the transport outcomes from
different projects to be quantified and compared with each other. Decisions about what to build and not to build are
then based on those comparisons. So it’s tremendously important to know that we are evaluating projects using a method
that undervalues the time of urban travellers relative to rural travellers, and undervalues the time of PT users,
cyclists, and car-poolers relative to drivers.
*************
Peter Nunns is an economist working in Auckland with an interest in transport past, present, and future.