Don’t believe the traffic projections

We have frequently raised concerns about projected future traffic growth, given that in recent years there has been an extended flat-lining of traffic growth. What’s perhaps most concerning about these projections is how they ignore what has actually happened in the recent past and how those producing the projections don’t seem to learn from past mistakes.

This isn’t just an Auckland problem. An article I came across recently looks at projections for a bridge across Lake Washington in Seattle highlights how stubborn the projections of growth are despite evidence to the contrary:

traffic-projections-678What’s crazy about this graph is how persistently wrong the projections have been – yet without any change to reflect the reality of declining traffic volumes over a 15 year period between 1996 and 2011.

Yet it’s not just a specific example of a bridge in one American city where we see these persistently wrong projections coming through. Let’s look at a comparison of official traffic projections across the UK over the past 20 years and compare those with what actually happened:

uk-traffic-projectionsIt’s hard to know whether these repeated mistakes are just accidental, ignorant or wilfully neglectful of reality.

Our own local example of this ignorance is in the traffic projections being used for the stupid Additional Harbour Crossing Project, where modelled traffic growth rates completely ignored recent trends and therefore were calculated from a base that was significantly too high:

This graph was from a year ago and in the past when I’ve posed it, there have been some that say “look it’s starting to rise again” but the reality is it isn’t. The most recent monthly data shows traffic have flat-lined and volumes are still less than it was a decade ago (monthly figures only started in late 2007).

AHB Monthly Traffic Sep 13

Similarly another frequent comment we see when this is discussed is to the effect hat the downturn is only due to the current state of the economy. However many economic indicators are pointing to the economy being much healthier today than it was a few years ago. Other indicators highlight that while on a per capita or percentage basis we might not be doing as well as in the past, on a total basis we are doing well. For example despite the percentage of people who are unemployed being higher than it was in 2007/08, in total there are actually significantly more people employed at the moment.

I suspect traffic projections keep making these mistakes because they are calculated using models with fundamental problems in them. They are generally designed to predict the future based on extrapolating our behaviour from some point in the past. That may have worked in the 90’s (and earlier) but it doesn’t work now and one of the key reasons is that we are seeing generational changes occurring with young people choosing not to drive as much as older generations. Yet while road models might be well over estimating vehicle trips, PT models have been doing the opposite. One of the best examples is Britomart where we exceeded the 2021 projected daily patronage in 2011.

Britomart Projection Numbers Graph

And even the Ministry of Transport in their response to the City Centre Future Access Study said that private vehicle trips were probably being overestimated.

MoT OIA docs - Modelling 1

When there are tens of billions of dollars of public money is riding on these faulty projections, it suggests we need a new approach starting with not believing the current projections.

Shared spaces for busy intersections

All our shared spaces so far have been on relatively quiet local roads: Elliott Street, Fort Street and Lorne Street in the city centre and the newly bypassed Totara Ave in New Lynn. In the UK things are a little more ambitious, as highlighted in a recent Better Cities & Towns article:

It’s been nearly a year since a major traffic light was removed at an intersection with 26,000 vehicles per day, heavily used by truck traffic in Poynton, Cheshire, England. A section of the town’s High Street was also renovated so that pedestrians, bicyclists, and vehicles all mix. The volume of vehicles is nearly double the upper limit for “shared space” intersections according to industry standards.

“This was the most ambitious shared space project — certainly in the UK — any anywhere else that I am aware of,” says Ben Hamilton-Baillie, a British urban designer who led the project.

Accidents have gone down so far, although Hamilton-Baillie cautions that it is too early to draw conclusions on safety. Traffic queues have been drastically reduced, despite an increase in pedestrian space of more than 100 percent.

Here’s what the intersection looked like previously (H/T Gordon Price):poynton-3Compared to now:poynton_selection2Perhaps what is most interesting is that traffic flow has actually improved, as well as safety improvements and a vastly nicer pedestrian environment.

The corner of Quay Street and Hobson Street, outside the entrance to the viaduct and a place where huge volumes of pedestrians are treated like scum, would be a great place for something like this.


View Larger Map

The possibilities are endless…

The elephant in the room: a familiar tale of traffic decline

While the UK trends have been around a bit longer than ours, this could very easily be a good description of the current disjoint between reality and the government’s transport policies in New Zealand:

Build on falling traffic trend – not on the countryside

10 November: Quarterly traffic figures are out, showing falling vehicle numbers and undermining the case for new roads.

The statistical releases from the DfT cover motorways and main roads in England, Wales and Scotland, and all show a falling trend in vehicle traffic volumes, alongside rising speeds. Across all types of vehicles and roads, motor traffic volumes were 0.5 per cent lower in the third quarter of 2011 than in the third quarter of 2010, continuing a long-term trend.

The releases also shed light on traffic speeds and congestion. For local authority A roads in England in particular, the data is interesting. It shows absolutely no increase in congestion on these roads since 2006 – in fact the overall change observed is that traffic on A roads is now moving 1.8 per cent faster than it was during 2006/7 (25.1 vs 24.6 mph).

This shows up a major flaw in the analysis on which new road plans are based, with predictions of small vehicle time savings – added up over decades – making up the majority of measured ‘benefits’ of building a bypass. These time savings are not calculated in comparison to the current situation, but against a projected level of congestion that is modelled on an assumption of continued traffic growth. Growth that may never actually happen.

And it would be a mistake to assume that reduced traffic is simply a product of the recession, and that economic recovery will automatically bring cars flooding back onto our A roads.

Falling traffic is a positive trend that started well before the financial crisis in 2008. At least partly, it represents a growing desire – particularly among younger people in towns and cities – to live without car ownership and make use of high quality alternatives.

The choices that national and local governments make over the next few years will have a lot of influence on what happens to traffic, offering people more choice or locking them into car dependency.

As Phil Goodwin and other academics have pointed out, London’s reliance on the car has been declining from a peak way back in 1993, and the city saw car ownership and mode share fall sharply throughout the long financial boom of the 2000s, while public transport services improved. So prosperity and traffic growth are not as intertwined as some local authorities bidding for DfT funding for their bypasses would like to believe, and these councils are making a grave mistake by assuming new roads will bring real financial benefits to their local areas.

If councils like Norfolk, Devon, East Sussex and Bristol are keen to spend money improving transport, they should be looking instead at ways to make the most of people’s desire for alternatives to the car in their towns and cities with smarter travel and land-use planning; building on the trend for less traffic, rather than building new roads and developments out in the countryside.

I’ve put the key paragraph in bold. Pretty much all our planned roading projects are justified on the basis of solving huge future congestion problems caused by assumed massive future traffic growth. If those growth predictions are wildly inaccurate and if there’s not a huge problem now (like is the case for many of the RoNS projects) then we’re basically spending a huge amount of money to solve a problem which doesn’t, and won’t, exist.

Interestingly it seems as though UK traffic growth predictors suffer from a bizarre problem of simply ignoring reality where it disagrees with their modelling outputs – check out this comparison of predicted traffic volumes versus real traffic volumes:

Seven wildly inaccurate predictions later and they still haven’t learned their lesson?

Meanwhile, in the USA traffic volumes are back to 1995 levels:

And in New Zealand state highway volumes have basically gone nowhere in four years (the pink line is all traffic):

I wonder how long it’ll take for the transport profession to come to the realisation that traffic just isn’t growing anymore. I guess as agencies like NZTA have a massive vested interest in completing ignoring these trends, it might be quite some time yet.  Falling traffic volumes are the absolutely giant “elephant in the room” when it comes to transport policymaking because as soon as we accept what’s finally happening, everything changes.

People travelling less in the UK too

The slowdown in traffic growth seen both in New Zealand and the USA is also occurring in the UK, as detailed in a recent Economist article.

Car and van mileage has fallen over the past four years, mainly because of the economic slump. Yet this comes atop a longer-run trend: for around 15 years, Britons have been making fewer journeys. According to the Department for Transport, the average person now goes on only slightly more trips than he did in the early 1970s, mostly by car. Between the mid-1990s and 2010 individuals made 19% fewer shopping outings. Jaunts to see friends dropped by fully 22%, thanks to a fall in visits to private homes.

The article is accompanied by a couple of really useful charts which start to break down the difference between trip numbers and distance travelled, and then looks at which types of trips are increasing and decreasing the most:  The decline in trip numbers is discussed at length in the article, which looks at long term cultural changes such as people grouping together shopping trips much more, a general reduction in “Couch and kitchen socialising”, the rise of the internet and higher petrol prices making think twice about ‘discretionary’ trips. Yet the article also argues this does not mean Britain is turning into a country of hermits:

A rise in hours spent staring at computer screens and televisions—and a concurrent decline in journeys to see friends at home—does not necessarily mean that Britain is becoming a nation of hermits. Mr Gershuny argues that those who engage with friends online also tend to see them more in person, even controlling for age. The internet may make socialising more “efficient” and diverse—people can research and plan where they are going, or what they want to buy, eat or do when they are out. In fact, many trips to visit friends at home are being replaced by jaunts out with friends, reckons Oriel Sullivan of Oxford University.

The ONS’s national well-being survey suggests that socialising with friends is still one of the most popular pastimes. People spend more time chatting on the phone too. Yet the travel data may obscure such engagement because socialising is increasingly combined with another activity. The number of trips to meet friends outside their homes has held steady. Other types of outing have become more popular, such as what the ONS describes as “entertainment or public activity” and “day trips”, all of which are likely to include friends or family. For many people, work is also a social encounter.

Retail travel also follows this trend. Despite the decline of the high street, there are more grocery shops at transport hubs such as railway stations, which makes buying fresh food possible without a special outing. “Multitasking” has become a popular shorthand for the predicament of modern workers. It may increasingly apply to their leisure time too.

There is an interesting debate to be had, I think, around the extent to which we might expect trends like these to continue into the future. I do wonder whether there’s a risk of us over-building transport infrastructure – particularly new roads where volumes are very much static (unlike booming PT patronage) – because our future projects are based on past assumptions which simply no longer hold true.

I feel the real test will be what happens to travel patterns and traffic volumes once economic growth properly returns (assuming it will do so).