I remember someone once telling me that a significant problem with the “transport field” is that many of the experts start out their academic studies as civil engineers, and the first thing they study is stormwater. Now I have a great appreciation for stormwater engineers, as they make sure our cities don’t flood when it rains, but ultimately it’s fairly logical stuff: x amount of rain falls and you need to get rid of it without flooding the place and without polluting the environment. There are some reasonably simple calculations to make: what will the demand on the stormwater system be (ie. how much rain will fall) and what will the supply of the system be (how wide do the pipes need to be to get rid of all the water).
Inevitably, when engineers move on from studying stormwater to studying traffic, they often seem to end up thinking of vehicles passing along roads in exactly the same way that they think of stormwater rushing through pipes. There’s a certain level of demand (the number of vehicles passing through the system) and a level of supply (the number of lanes, whether it’s grade separated, intersection capacities and so forth). If you’re a road engineer then life is all about trying to ensure the level of road supply is greater than the level of road demand, as otherwise you end up with congestion.
Reading through the documentation behind a proposed upgrade to the intersection of East Tamaki Road, Preston Road and Ormiston Road in Manukau City, I found a perfect example of how road engineers continue to think of people like we are mindless stormwater. The project to simplify the currently complex intersection probably makes quite a lot of sense, as the current intersection is a damn mess and the proposal would simplify that significantly.
However, in analysing the scheme assessment report I found myself getting a little bit suspicious of the level of anticipated benefit when I saw that the cost-benefit ratio of the project was supposedly 8.0!
Delving into the way that this cost-benefit analysis was undertaken provides some quite useful information that perhaps goes a long way towards explaining why roading projects always seem to promise huge time savings benefits that never seem to materialise (or only materialise for a very short period of time before induced demand eats them up).
What time savings benefit analyses for roading projects tend to do is not actually compare the situation as it is now with the situation that will exist in 20 or so years, but instead compare an expected do nothing outcome with an expected do the project outcome in 20 or so years. Now obviously that makes sense in a general sense, as you may well be justifying your project in terms of how it answers the question “what difference will this project make in 20 years time?”, but what it does mean – to put it a tad crudely – is that you’re actually comparing one guess with another guess. And it is actually the “expected do nothing outcome” that I think we’re probably measuring very poorly at the moment, leading to an over-estimation of the time savings benefits that roading projects are expected to bring.
This issue is highlighted in the do nothing analyses for this East Tamaki/Ormiston/Preston project, which I have included a table of the “expected situations” in future years if the project doesn’t happen:
Looking at this data, one’s first response is likely to be “well crikey if we don’t do the project look how terribly congested and slow this intersection will be by 2031!” And, taking the modelled results without question, yes that certainly appears to be the case. Passing through the East Tamaki Road/Preston intersection in 2031 will apparently take you nearly six minutes during the morning peak, compared to just over one minute now. Similarly for the Ormiston Road/Preston intersection, it will supposedly take you just over two and a half minutes to pass through this intersection in 2031 during the AM peak, compared to 70 seconds now. Average speeds will supposedly decline from around 30 kph to around 6 kph.
The next step in working out the benefits of the project is to model what will happen in 2031 with the upgrade, and then compare the two. Here’s similar data for the new intersection: While congestion still goes up, it is not nearly as bad as in the “do nothing” situation. The difference between the two situations is aggregated, a cost allocated to every saved minute, and voila, we have a significant amount of “time savings benefits” that are used to justify the project. Here’s the table showing the outcome of the assessment:
As the project only costs around $6 million, its cost-effectiveness is pretty whopping according to this assessment, and therefore it is seen as a priority to fund.
However, getting back to where I started this post, does anyone else see a problem here? The question I am interested in answering is whether, once the existing situation started getting worse than it is now, people would continue to pile into using that particular intersection, or whether they would start seeking alternatives – like going another way, catching the bus (if there was one that wasn’t affected by this congestion) cycling, walking or simply not taking that trip. Now obviously there would be some level of economic disbenefit from people not being able to make these trips, or having to go a more non-direct way in order to avoid the congestion of this intersection – but shouldn’t that actually be what we’re measuring to determine the cost-effectiveness of the project, rather than some mythical “congestion armageddon” that will clearly not happen.
This is where we get back to road engineers thinking that people are like stormwater. If this intersection were just a low capacity pipe, and it started to rain, then of course the water would continue to keep piling up. But people are not stormwater. People will be put off taking trips, people will choose alternative routes, people will choose alternative modes (if they exist and have a speed/convenience advantage) and so forth.
I actually think that the “20 minute savings” promise that NZTA’s CEO went on about at yesterday’s opening of the duplicated Mangere Bridge makes this same mistake. It expects that people will continue to keep piling onto the bridge no matter how badly it is congested. It is this misunderstanding, or ignorance, of the fact that congestion will put people off using that stretch of road, that I think significantly contributes to the over-estimation of time-savings benefits for our roading projects. This inevitably ends up being made worse by politicians who pick up on the mentioning of a “time saving” and assume it’s the difference between the road as it operates now and how it will operate when the project finishes. In actual fact, it’s a comparison between a ‘congestion armageddon” in 20 years time that will clearly not happen, and how the road will (supposedly) operate once completed – a calculation which inevitably ignores induced demand.
In other words, it’s usually total rubbish – because people are not stormwater. We’re people.