From the Commons Transport Committee oral evidence on the new (MOAR ROADS) National Policy Statement on National Networks, March 31st 2014:

Witnesses: Robert Goodwill MP, Parliamentary Under-Secretary of State, Department for Transport, and John Dowie, Director of Strategic Roads and Smart Ticketing, Department for Transport, gave evidence.

Q91   Chair: Specifically in relation to forecasts, we have had a lot of representations saying that the forecasts are inadequate. Are you going to revisit that?

Mr Goodwill: Forecasting road traffic and rail passenger demand is very difficult given the challenges in predicting growth in the population, travel behaviour, the economy and oil prices. Our view is that we should not focus on a single prediction but plan and invest for the future based on a range of scenarios. Our lowest forecast for road and rail growth is based on low population and on economic growth, and suggests a substantial increase in traffic and rail over the coming years. Forecasts from such strategic models are not used in justifying individual schemes; these are assessed more by using mode or location-specific models designed specifically for that purpose.

In terms of the reliability of forecasts, we believe that the Department’s national transport model has not systematically overestimated traffic; it has over and under-forecasted traffic, and this reflects past over and under-forecasting of the key underlying drivers of traffic, for example, GDP, population growth and fuel prices, which are themselves uncertain. While forecasting road traffic demand over decades is difficult, our forecasts perform well when the trends in those underlying drivers are accurately captured. There is a chart on page 11 of the national policy statement for national networks which illustrates this point. The NTM has successfully forecast the trends in traffic since 2003, and the forecast for 2010 was within 1.3% of observed traffic data. I understand that traffic in London is not representative, but even our lowest forecast for traffic growth suggests a substantial increase in traffic levels over the next 25 years, and we have made similar forecasts for rail demand and rail freight.

Q92   Chair: I think that means you are not going to change how you do it.

Mr Goodwill: We are not going to write these forecasts in stone and not respond to events that may change. If the price of oil were to double or population growth did not follow the trend we expected, or if for some reason older people no longer used their cars into their ripe old age because of insurance issues, we could certainly revisit that. Mr Dowie might want to come in to explain how we do that. This is a 20-year forecast, but we are not going to leave it on the table for 20 years without revisiting it.

John Dowie: As an ongoing process, we keep looking at the traffic forecast and the various assumptions we use and how the model works. It is an ongoing learning process. Some groups have rightly flagged up that the model’s accuracy in terms of what is going on in London looks weaker than at a national aggregate level. Clearly, that is something we will be looking at. We are not freezing the traffic forecast and the numbers once and for all in this document.

Q93   Chair: It is an ongoing process.

John Dowie: Yes, but we are just trying to avoid spending weeks at a public inquiry debating the inner workings of the model.

Just to prove where the BS lies in this, this from the Campaign for Better Transport’s is Professor Phil Goodwin’s famous graph of what the DfT said would happen against what actually happened:

I’ll leave it to the reader to judge if Mr. Dowie should be explaining the inner workings of a model that obviously doesn’t work very well.

The consequences for Boris and TfL ought to be profound here – if the DfT, which can’t see anything wrong with models that consistently over-estimate national trends say there’s a problem specifically with forecasting London traffic, how reliable is anything that produces in terms of a basis for hugely expensive, destructive new road schemes?



Set your Twitter account name in your settings to use the TwitterBar Section.