The scene for the presentation at the evenf of 5 May was set Tom’s paper in TransportNetwork (see https://www.transport-network.co.uk/Analysis-Learning-from-a-crisis/16611) which noted that the need for and reliance on epidemiological modelling during the COVID-19 outbreak had been exceptional and very visible to a public seeking answers rather than wanting to understand how such models were derived.
Decision makers in responding to the crisis had been influenced by the models and modellers and the lifesaving measures put in place would not have been possible or thought of as acceptable to politicians and the public if the model evidence had not been available.
It was therefore suggested that the COVID – 19 outbreak can teach transport modellers the following six lessons:
1. Modellers need to follow the lead of COVID modelers and be more assertive but humble in how the results are presented. The public and politicians seek answers, rather than wanting to understand how these were derived. Models need to support the public’s need to know what kinds of interventions need to be in place.
2. What-if scenarios can help show how we emerge from lockdown and whether the future transport system can be cleaner and greener. The assumptions and caveats must be made clear and focus the debate on those. Need to move from a “predict and provide” paradigm to one of “decide and provide”? Modelling needs to start providing support for the measures, interventions and policies necessary to support a manageable, deliverable and affordable transport system in the short, medium and longer term. Existing models still have value in testing alternative scenarios. For example:
a. What is the actual proportion of people who work-from-home (WFH), where do they work, what will be the impact of X% of them continuing to WFH in the future?
b. Impact of future declines and increases in oil prices and costs of using cars?
c. What is the percentage of people using active modes of transport? Will they continue to do it for exercise and health, and will they transfer that behaviour to other types of travel (e.g. for education, work and leisure) and how will that affect demand for road space? Impact of reallocating road space to facilitate demand for walking and cycling and what level of transfer is required to warrant the reallocation of road space?
d. How does the need for social distancing affect the capacity of public transport?
e. What are the alternative population growth rates and economic growth forecasts that are needed for business cases for infrastructure projects and are those business cases developed pre-COVID still valid?
3. Also, model, model well, and present the Do-Nothing situation. Discussions about models need to be rational about different model approaches, forms and the assumptions made in the models and how the assumptions affect the outcomes.
4. Modellers should welcome scrutiny and challenge from scientists outside of their field. Models cannot make decisions!
5. Any model forecast pre-COVID will be wrong. Good quality research is necessary to underpin new model assumptions.
6. Transport is a derived and elastic demand, but not everyone has the same opportunity to respond. Model (and monitor) the wider health, environmental and economic impacts of transport interventions post-lockdown with a social lens to look at how this impacts differentially across society.
The Question to be answered was: What can we as transport modellers do to support post-lockdown and post-COVID project and policy decision-making?
The summary of the discussion was:
The fact that long term forecasts cannot take “black-swan” events into account makes the case for more effort devoted to scenario planning and “what-if” sensitivity analysis and the communication of multiple plausible futures to decision makers and the community in a form that is helpful.
The problem that the likelihood that long term forecasts used for infrastructure planning will be wrong (but don’t know how wrong) highlights the need and value of scenario planning and “what-if” sensitivity analysis.
Although opinion surveys are not a reliable source of input for models they are all that is available and can act as a starting point.
It is not expecting that there will be any 2020 and 2021 base year models developed but that Big Data passively collected in 2018 and 2019 may fill the gap in the interim with more confidence that passively collected data may have value over traditional surveys in the future.
Under the “decide and provide” paradigm, models can be used to test for resilience (e.g. what-if shared mobility does not occur, what-if there is no post COVID economic bounce back, what-if CCAM does not occur, etc).
New project metrics are needed beyond the simple CBA measure. Preference is to adopt a dashboard type approach to present results that are meaningful and relevant to decision makers and the public.
Overall the presentation achieved what it set out to achieve, namely, to outline what transport modellers can/should do to support post-lockdown and post-COVID project and policy decision-making under a new “decide and provide” paradigm.
A recording of the Webinar is available for AITPM members here.