It is a common refrain, “why didn’t the planners see that coming?”

A 2021 report from the International Transport Forum - Travel Transitions. How Transport Planners and Policy Makers Can Respond to Shifting Mobility Trends - considers this. It looks at why transport planners didn’t foresee changes in mobility patterns over recent decades, their generally poor performance in predicting transport demands, and how planning can be improved.

The report shows that common analytical and decision-making methods for transport planning are no longer fit-for-purpose. The way planners think about the future also needs to change.

Its messages are relevant to other fields of planning, and to futures thinking more generally.

The main point is that traditional forecasting models provide a false sense of certainty in changing times. Uncertainty is usually hidden (unintentionally or not), rather than being explored.

The foresight field too is strewn with over-confident predictions of what are the important factors shaping the future.

The report doesn’t reject the use of models, but argues that better models (and mindsets) are needed, and that they shouldn’t be the sole basis for planning decisions.

As any good foresight practitioner will tell you, taking the time to formulate the futures questions is critical to success. You shouldn’t just grab a model to generate an output value. Asking “how may demands for transport be influenced over the next two decades?” leads to a very different approach to planning than asking “what will demand be in two decades?”

Six important foresight messages from the report are described below.

Novelty is often the Achilles heel of models

Models usually don’t cope well with significant unanticipated changes, such as new travel options and behaviours.

It can also be difficult to determine whether newly emergent behaviours and activities are short-lived fads or will become an important influence. If they are viewed as fads, they are ignored or discounted (as the first automobiles were).

On the other hand, extrapolating out from a fad can distort expectations too. The augmented reality game Pokémon Go was very big in 2016, but this type of mobile game hasn’t yet become the norm.

If they do become an important trend, there remains considerable uncertainty over what time period, and how, the impacts could be felt. This is the case for autonomous vehicles, when for the last decade fully autonomous driving was expected by many to be just around the technological corner. Assumptions about use of autonomous vehicles also haven’t been well tested.

This requires resisting the temptation to identify the next big thing, or prematurely calling the demise of the last big thing. A range of potential trajectories needs to be actively considered, and monitored over time, to improve understanding.

The need to take a broader system view

Planning and other models may not take account of broader system factors that influence future demand (or supply), or they assume that the influences won’t change. For example, the ITF report notes that long term travel trends are influenced more by wider societal changes (such as the nature of work) than by changes in transport systems.

Similarly, whether working away from the office will become common or not depends on a range of factors, which may not be accounted for in models, or in “hot takes.”

So, when times are changing you shouldn’t just rely on the usual sources and suspects for information.

Focus on hypotheses not simplistic explanations

A discussion paper that supported the report noted that studies of earlier transport transitions usually developed ad hoc qualitative explanations rather than hypotheses of change. There has often been a belief in the importance of a particular factor that influences future transport demand, which subsequent research shows played only a modest (if that) role.

And even if one or a few factors accounted for much of a previous transition, it doesn’t mean they will for future ones.

Hypotheses encourage exploration. For transport planning, the report notes that development of good hypotheses requires interdisciplinary collaboration. Testing them requires good data sources, long-term perspectives, and a focus on processes not just outcomes.

Being data informed not data driven

Quantitative approaches are common in most areas of planning, and becoming so in other foresight projects. As more and larger data sets, and methods to analyse them, become increasingly available it is tempting to add them into models. These can provide benefits, but the type and quality of data, and the algorithms used, need to be carefully considered.

Theory-free approaches, where patterns in the data are identified without hypotheses for how they come about, can be useful initial steps. But patterns don’t always illuminate processes, so hypotheses will still need to be developed to help plan better.

As Nobel Laureate Paul Nurse commented, “data should be a means to knowledge, not an end in themselves.”

It is often assumed that the critical indicators to measure activities have been identified, and the appropriate data is available. That’s not always the case, especially in periods of change. Available data, rather than the best data may be used. If they can’t be easily quantified critical factors may be omitted from models. Clarity on why particular data are used and how, and why other data sets are not, is becoming more important to understand the strengths and limitations of models and their outputs.

Models may not take account of considerable variability between regions and between different population groups. For example, transport patterns can vary in different parts of the country and between different age groups.

The report also recommends combining rather than replacing traditional data sources (like surveys) with new sources (such as mobile phone records), since this can help identify and understand longer-term trends as well as more rapid changes.

Focus on plausibility and adaptability, not precision

 The report advocates for a shift from “predict and provide” to a “decide and provide” approach. The focus shifts from precision to plausibility. “Decide and provide” is a vision-led strategy, where a desired future state is defined and actions set out to help this come into being. It notes some planning initiatives are already taking this approach, which was developed in part from Waka Kotahi’s (NZ Ministry of Transport) futures work in 2014 (see Lyons & Davidson 2016).

The “decide and provide” approach is intended to provide adaptability in response to unanticipated changes. It moves away from reacting to trends and locking in policy decisions and infrastructure, toward what it calls “regime-testing.” This involves more explicit questioning of how the world or system is operating, leading to the development of a vision that informs rather than reacts to policy decisions.

Developing a range of scenarios is advocated. These shouldn’t be used as just descriptions of how the future may look, and then picking one. Instead they help to highlight uncertainties, and shape the vision and the actions to achieve it.

Recognising governance challenges

The report doesn’t just examine the planning process. It also highlights governance challenges associated with changing planning approaches. Overcoming these challenges, the report recognises, is much harder than simply identifying them.

One is siloed organisational structures, which often inhibit taking a broader view of the system the planning activity operates within. Such as having transport planning run by the transport agency, rather than collaboratively by a broader collective of agencies that cover the key influences on transport.

Another is the multi-layered nature of transport governance. This involves local and regional councils or governments, and the national government. Power and incentives usually differ between them, creating conflicts. The power usually rests with those who collect the taxes, but, the report notes, the land use decisions are usually set at the local and regional levels. Such governance complexity will often be common for other foresight projects. This highlights that power structures and differences in values and practices shouldn’t be overlooked when thinking and acting for the longer-term.

A third important challenge is having the organisational capacities to change. The report identifies “professional impotence” as a barrier to change. This is where individuals can see a need for change but the norms and expectations associated with traditional practice make it difficult to actually change.

The report identifies four types of capacities (based on a framework from Hölscher et al.)  needed to transform approaches to governance:

The report closes by noting that transformation doesn’t, and can’t, depend on perfect institutional arrangements. But it is dependent on “… the willingness and openness to reflect, reconsider, explore and challenge established processes and standard ways of thinking and doing.”

Featured image by Nick Fewings on Unsplash