Viva predict and provide! Or the battle of the truisms

Tom van Vuren

Inevitably, for decisions as important as the long-term consequences of a transport policy or project for urban structure and society, we need some kind of forecasting or modelling. Irrespective of the complexity of those models, from simple trend extrapolation to state-of-the art activity- or agent-based models, we cannot provide confidence without predicting the outcomes of alternative decisions.

Predict and provide is a much maligned term, but prediction is necessary even for a decide and provide approach, if the decision is not whether to build or not, but aimed at understanding the outcome of each possible intervention before deciding which one to pursue. You can decide what to build, but not what the outcome will be. For that we need to model. In other words, we need to predict to decide what to provide, and explore how to do that well.

Every model is full of assumptions and simplifications. Any debate on travel demand forecasts should, first and foremost, focus on the assumptions rather than the model itself. It’s not just the task of the modeller, but also the transport planner to explain the likely outcomes in the light of these assumptions and explore if different assumptions in the model about the future will lead to different decisions, policies and plans. And let’s not ignore the value of modelling in the evaluation of the success of an intervention after implementation – modelling the counterfactual (what would have happened if we had done nothing or something different) helps explain to a critical public and helps decision-makers learn for the future.

Uncertainty does not reduce the value of modelling; it actually increases that need to explore and explain. And concerns about future uncertainty will most likely lead to the implementation of more adaptable policies and plans. Any intervention that can be tuned up or down, such as pricing, will be easier to adjust than the building of fixed physical infrastructure. Monitoring the effectiveness of an intervention in a volatile environment, measuring against the predicted outcomes, and managing how to adapt the policy intervention against the objectives, will also require modelling, adjustment of the parameters, changed inputs and amended future forecasts, development of new functional structures.

From predict and provide to predict, decide and provide, to explore and explain, to monitor and manage. It’s a mouthful!

Find out the interesting modelling-related sessions at the 2021 Transport Practitioners Meeting (held online between 6 and 7 July) here:

Tom van Vuren is Regional Director UK & Europe for Veitch Lister Consulting, and Visiting Professor at the Institute for Transport Studies, University of Leeds.

Leave a Reply

Your email address will not be published. Required fields are marked *