Stephen Rees's blog

Thanks are due to my friend Eric Doherty who has been using the list serves and the Livable Blog to draw attention to a new report from the Campaign for Better Transport in the UK. They have been digging through the “post-opening project evaluation” reports that the Highways Agency there uses to examine the “schemes” that they have built. They go back one year and five years after a new road is built, or widened, and look at the impact compared to what was forecast before the project started. In other words, they look at what was forecast to happen and what actually happened. These reports are remarkably detailed, clearly objective and look at a range of data – not just the vehicle flow. We, of course, do not do that here.

Here is the opening of the Executive Summary of the Campaign’s report

Each year the Highways Agency spends around one billion pounds building and widening trunk roads. Ministers decide whether to invest in motorways and A-roads based on the Agency’s detailed economic analysis. Their evidence is supposed to spell out what the impact of road building would be: the impact on the economy, on the local environment, on traffic and on greenhouse gas emissions. These forecasts need to be reliable, because of the amount of money being spent and the potential for severe impact on congestion, traffic and the environment.

Unfortunately a number of reports commissioned by the Highways Agency show that these forecasts cannot be relied upon. Instead of accurately predicting what will happen, the Agency’s forecasts underestimate the effect on traffic, air quality, noise and greenhouse gas emissions. They also fail to predict the economic impact and whether schemes will be good value for money.

If you would like to download the whole report as a pdf that link will do it. Or if you are really into this kind of stuff you could find all the HA POP reports on major road projects . It is quite possible that the campaign was a bit selective in its choice (they took the five most recent ones) but it seems to me that they do notice some of the inherent problems that persist in modelling transportation improvements in general and roads in particular. As a former economist at the predecessor Department of Transport, I can say that we had similar concerns back in the 1980s, and a lot of us thought then that as computers got smaller, cheaper and more powerful, the models would get better. And, of course, they have done. But even so, there are persistent problems with the way these road expansions turn out

post-opening project evaluation reports (POPE) found that overall traffic levels rose significantly as a direct result of each scheme. Two of the three bypasses studied simply moved the congestion elsewhere.

Forecasting was generally inaccurate, especially of future traffic levels, and the economic forecasts did not reflect the actual impact on local businesses. CO2, air quality and noise impacts are generally worse than expected, and walking, cycling and public transport did not improve, even on local roads where traffic has decreased.

That last bit is significant for us too. As in the UK, major road building projects here – such as the Highway 1 widening – are claimed by the government and Translink to improve cycling and transit (I don’t recall them even mentioning walking!).

Some of the problems reside in the limitations of the available transportation models. Perhaps the greatest weakness has always been their inability to understand the way that projects induce traffic – by encouraging more and longer trips – but also change the way that land in the region around the project is used. The model used for all the Gateway projects, and for the Sea to Sky Highway widening, simply used the same land use and population distribution forecasts for all future scenarios. In other words, the number and distance of trips in the future was fixed as a factor of land use (population in residentially zoned areas, employment in others and so on) – and while that land use was different from the “before” case it was the same in both the “with project” and “without project” future cases. This, of course, both understates traffic growth, but also overstates benefits in terms of travel time savings ascribed to the project. Underestimating future traffic in our case means that environmental impacts such as noise, air pollution or greenhouses gas emissions are also less than reality, but show up as lower “costs” on the evaluation. More importantly, it also creates the illusion that congested will be lower in future. As Gordon Price has been saying about the Gateway for years, just tell us one place where road building has worked to reduce congestion.

The five year after (5YA) reports found that overall traffic increased substantially once each new road was opened. This was substantially higher than the background level of traffic growth and each report attributed this to new trips generated by road building.

In the UK they talk about “trips generated” but here we use the term “induced” – because the model itself uses the term “trip generation” for the function by which it forecasts travel based on population distribution.

The reports show that traffic growth was not limited to new road space; while in each case some drivers using local roads switched to the new road (providing limited congestion relief on minor roads), nearby roads tended to see large increases of traffic.

That is going to be one of the most significant impacts of the Highway #1 widening, and one which the government has deliberately under estimated, as a way to calm the real fears from neighbourhoods that will be swamped with new traffic. It is being explicitly dealt with by engineers in both Burnaby and Vancouver  – to the extent that they can. But I suspect that there will be howls of rage from the very people who currently expect better travel conditions south of the Fraser once the new bridge opens. Of course, we have seen this effect before – just ask those who were here when the Alex Fraser Bridge opened. Once again, the model simply assumes that the same number and length of trips are better distributed on an enlarged network that has greater capacity. But since induced travel is not factored into the equation, the forecasts are wildly misleading.

It is not as if any of this is new either. As earlier blog posts attest.

Eric thinks we need a moratorium on building new roads here, just like the one the UK Campaign wants. I would agree, except that I don’t think it needs to be a temporary halt to projects until we can do a better job of project evaluation. We need to stop increasing the capacity of the highway system because we know that not only does it not cure congestion, but if we are to reduce our impact on the planet significantly we have to take drastic steps to reduce the use of cars. Indeed, if all we wanted to do was reduce congestion, simply spending the same amount of dollars on transit projects would be a good start, but we would still need to do something about land use AND we would still need to find a way to finance transit operations and maintenance. A road building moratorium might be a good gesture, but it would not deal with the fundamental problem. In North America we drive single occupant vehicles too often and too far – and we claim that this is “necessary” because we are locked into a car dependent land use pattern we are reluctant to see changed. Until there are viable alternatives – and better designed communities – I cannot see that changing.