Over the course of the pandemic it has become clear how wedded the transport sectors are to the idea of “forecasting”, in particular predicting when or if traffic levels by air and rail will return to pre-pandemic levels.
We know the pandemic has changed many things, but the message does not seem to have got across that the pandemic has changed the relationship between our past and our future in ways that have profound implications for decision-making in the present. This is true in most aspects of economic life. In transport it has particular importance, both because the pandemic has affected these sectors so severely and because forecasting plays such a central role in transport planning and provision.
“Forecasting” is a problem for economists. The activity is fraught with difficulty. In the common understanding economic forecasting and “prophesy” seemed to be confused, i.e. telling the future. The use of the word “forecasting” makes it hard for economists to deny this implication. But most economic modelling is not intended to tell the future. It is about understanding the causal relationships between variables in the past to better understand the present. Insofar as these models are used to “forecast” the point of the exercise is to better understand the consequences of policy actions, on the assumption that historic causal relationships continue to hold. When the future does not work out as we expected, we ask: did we miss something in the past, or have the causal relationships changed? Each iteration is part of the process of better understanding the present.
But in transport sectors projecting future traffic levels, “forecasting”, takes on a special importance because of its role in key instruments for service provision and performance management, including concessions, franchises, PPP contracts and cost/incentive-based regulatory systems across road, rail and aviation.
At the heart of all these processes bidders (and public authorities) have to assess the level of passenger traffic they “expect” to use the service over the life of the contract. These estimates inform an assessment of the capacity needed, its cost, the price to be charged to users and the amount the private operator is prepared to pay to win this contract. But these forecasts are primarily an extrapolation of past relationships: price and income elasticities measured on historic data applied to other economic forecasts of macro-economic variables (per-capita disposable incomes, house prices, etc.).
The truth is we know these past relationships are broken. In time, new ones will be established, reflecting the long-run impact of the pandemic on working patterns and leisure travel. We can speculate about how altered these relationships will be and when they will settle down, but we cannot “know”; not least because the continued progress of the pandemic and public health responses to it remain unclear.
Part of the answer to this is, of course, scenario planning. Where possible, adopt flexible strategies that can adapt to different possible states of the world, as they reveal themselves.
But recognising the new uncertainty is vital to thinking about transport provision today. Concessions and franchises presenting private operators with volume risk will be very difficult to sell. Operating contracts, as now envisaged for the UK railways and already popular in continental railways, are a solution, but present major challenges in creating performance incentives not game-playing. At the same time, rolling stock and aircraft lessors will be thinking creatively about contractual arrangements to encourage the use of their fleets.
Price-controls for regulated infrastructure providers, such as airports or track providers, also suffer under the new uncertainty. Old approaches cannot be expected to prevail. More thought will need to go into making explicit the risk sharing agreement between infrastructure providers, customers and the state.