If you can spare me a minute or two of your time, I’d like to talk about how many dollars that time is worth to you.
Deep within the hidden machinations that is transport decision making in New Zealand, there is a tiny little cog that turns a lot of very big levers. It’s a big part of the reason our cities look the way they do, why we have built lots of certain types infrastructure but very little of others, and why so many people will die on our roads in the coming years while we willingly choose not to do the things that we know would save them.
I’m talking about travel time values.
Many have written before about the headaches travel time values cause us. Ian Munro wrote a nice article here, NZTA have done several pieces of research into it including this one, Greater Auckland have written several pieces on it including this one. One I thought was especially on the money is Paul Buchanan’s paper on it here – in fact if you’ve only got a few minutes I’d recommend skipping my blog and just reading his paper instead.
Travel time values are important as, in our current economic analysis framework, they make up by far the biggest benefits of our transport projects.
I got thinking about this the other day while I was reading some recent research the Transport Agency has commissioned to work out the value of travel time savings.
The work itself is fine, it answers the question it was supposed to answer. But reading the report it struck me how overly-simplistic the question being asked is. It basically asked “how do we figure out what dollar value people put on their travel time?”. The current Transport Agency Economic Evaluation Manual has a bunch of values below. They are split out into 3 purposes, different vehicle types, and even whether you are sitting down or standing up. So for example, if building a new road was going to save one car driver one hour driving to work, then that would be considered to add $7.80 of benefits to society.
The piece of research developed a survey that asked a series of questions to basically determine updated values for these travel time values, as these ones are a few years old now.
These factors are then used to determine which projects get built, and how transport systems then get operated. If building a new road saves lots of people lots of travel time, then it has a good chance of getting built, if it doesn’t then it doesn’t. If phasing traffic signals a certain way saves lots of people lots of time then it has a good chance of getting done, if it doesn’t then it doesn’t.
On the surface this sort of makes sense. But think about it for more than about 10 seconds and it quickly starts to seem way too simplistic. The way people value their time is very complex. If you’re rushing to the hospital with your wife in labour, you are probably going to be valuing your travel time extremely highly (i.e. you’ll be willing to pay a few dollars to get there a little quicker). But when you’re driving back home afterwards, you’ll probably be taking your time and not be too bothered how long the trip takes. Even the same trip on different days can have a very different value – some days I’m late for work and desperate to get there as quick as I can, other days I’m a little early and am not bothered if I get delayed for a few minutes for some reason. Every single trip on our transport system has a different value to the person making it.
In a free market this variation sorts itself out through pricing. For example a plane ticket on a certain route at a certain time has a certain price. If you’re willing to pay that then you can make that trip, if you’re not then you can’t. There are a massive range in prices from extremely cheap if you’re willing to fly at unpopular times or unpopular routes, to extremely expensive if you want to fly at the highest-demand time. So people choose their time and route based on the prices and how important it is to them personally to make that trip.
But our roads are different. We have only extremely blunt pricing (basically just rates and fuel excise). As a result there are currently lots of trips being made on our roads that are not actually of high value to the people making them. And because people making these trips are clogging up the roads, the trips that are genuinely important can’t get made properly.
There’s only really one proper solution to this – road pricing. This will almost certainly happen in time and will change everything about the way we plan and build our cities.
But even road pricing doesn’t really fix our problem of thinking about travel time values in an overly-simplistic way. We’re still going to need models of transport sytems, and still going to need to be able to predict if an idea for a project is any good or not.
I think a good start would be to treat travel times as a distribution rather than a blanket value, and incorporating this into the economic evaluation process. This would mean assuming travel times are distributed in some sort of bell curve from very low values, to very high values, and everything in between. When demand for roadspace exceeds supply (i.e. when there is congestion) it can be assumed that the trips that don’t get made are the very lowest value ones. If a project is being built to increase the capacity of the road system, and hence induce some of this suppressed demand back onto the roads, these trips should only be assessed at a much lower travel time value, not the same value as all the other trips currently being made. This would provide a more realistic (lower) estimate of the true value of road capacity increase projects.
This wouldn’t solve everything wrong with our transport decision making process, but it would at least remove one of the biggest, most obvious biases inherent in the current methodology.