Scenario Planning in an Age of Uncertainty
For more than 50 years, transportation planners have been tasked with developing forecasts of future transportation conditions. These forecasts are then used as input into regional plans, infrastructure design, and entitlement efforts. Expecting that we could have engineering certainty about human behavior related to travel choice, we’ve continued to provide these forecasts. Unfortunately, most of those forecasts were wrong…
For the first 40 of the last 50 years, vehicle miles traveled (VMT) in the US grew at 3-times the rate of population growth, and transit ridership declined. Few transportation forecasts got that right. A 2005 study by Flyvbjerg of 25 international rail projects found that passenger forecasts were overestimated by an average of more than 100%, and that 84% of the rail projects had actual ridership more than 20% below forecasts.
Forecasting travel has become even more complicated over the last ten years. VMT per capita has steadily declined since several years before the Great Recession, and transit ridership has risen. Reasons include lasting changes in employment and the economy, demographic trends, and the growing role of technology as a substitute for travel and mobility.
Travel forecasters are developing ways to learn from the first 40 years’ shifting travel dynamics, and to understand and project the significance of trends that began to emerge over the last ten. We’re using newly available Big Data from GPS and cellular devices along with more sophisticated and interactive models to improve the accuracy and responsiveness of our forecasts. An equally important realization is that we can better inform decisions on transportation projects and programs if we’re clear about the level of confidence and variability in our forecasts, the reasons for those uncertainties, and to the extent that those reasons can be influenced, the means through which decision makers can more narrowly focus the range of possible outcomes.
We’re dealing with a world that has become dramatically more dynamic and unpredictable and determined to make quick decisions. Technology and mobility services have become the primary focus of new transportation investment and system performance, led by the private sector business models, venture capital, and speed to market with prototype testing. Due to infrastructure funding and appropriation shortfalls, the more deliberate public process to plan, review, approve, and build major 20-50 year transportation projects will also require even more reliable and stable forecasts of costs, benefits and returns-on-investment. For both private and public investments, it’s become essential to reliably estimate consumer preferences and ability to monetize services and optimize use of public infrastructure. Empirical evidence on user preferences and the effectiveness of transportation services arrives daily into the data mines of Uber, Lyft, Google, Amazon and others. It’s our responsibility to obtain the best data available and use it to improve our travel models and forecasts.
reducing "path" uncertainties...
One way to address these significant uncertainties is to employ scenario-planning approaches.
Scenario planning first came to planning in the 1990’s with Oregon’s LUTRAQ and Salt Lake’s Envision Utahbeing early examples. Envision Utah, for example, evaluated the consequences of four alternative growth strategies on land consumption, public infrastructure, transportation choices and walkable communities. The performance-based scenario planning process also establishes metrics and evaluation criteria for monitoring the effectiveness and unintended consequences of plan implementation.
This comprehensive scenario planning process is well established and has been used throughout communities in the US and even internationally. However, there is one significant drawback to the current application. Scenario planning, as currently practiced, always focuses on predicting the future based on an extension of historical trends. This approach may work well when there is limited shift in current trends.
However, the last ten years have witnessed dramatic shifts in the US economy, demographics and the influence of technology on travel and mobility. VMT per capita, which had risen by 5-10% a year for the preceding forty years, dropped by about 8% in total during the last ten years. During those ten years, US per capita transit ridership increased, and rail ridership grew by about 24%. Expert forecasts range from a return to the growth trajectory of the first 40 years to a continuation of the more recent trend. The “extended-new-normal” forecasts for 2040 is about 8000 VMT per capita, while the “return-to-business-as-usual” forecasts for 2040 is about 18,000 VMT per capita, 125% higher. Click on the links below for white papers:
- “Demographic Trends and the Future of Mobility” – Lindsey Hilde, Alex Rixey, Eric Womeldorff and Jerry Walters
- “Effects of Next-Generation Vehicles on Travel Demand and Highway Capacity” – Jane Bierstedt, Aaron Gooze, Chris Gray, Josh Peterman, Leon Raykin and Jerry Walters
In response to this evolution, Fehr & Peers developed the TrendLab+ interactive scenario visualization tool to address this unprecedented change. TrendLab+ allows planners, policy makers and the public to express their opinions on each of 16 economic, demographic and technology trends that have begun and will continue to influence travel behavior, and the need for and consequences of different transportation policies and strategy choices. Click here to find out more about our TrendLab+ tool.
The use of TrendLab+ can be demonstrated by its use to analyze two vastly different scenarios as presented below.
Scenario #1- This scenario assumes economic growth, new opportunities for labor force participation in next-generation jobs, a rise in per capita incomes, and continued suppression of fuel prices.
Scenario #2- This scenario assumes that millennials continue their preference for urban areas as they enter their family-raising years and that technology will eliminate the need to travel for many day-to-day needs:
The awareness that these two different perspectives would mean a difference of about 24% in VMT per capita will inform scenario planning and policy and investment decisions by State, regional and local governments. A difference of 24% in per capita VMT will make a substantial difference across a region or even a City. It could mean that roadway infrastructure projected to be above capacity would have its useful life extended by 10, 20 or even 30 years. Alternatively, it could mean that a freeway project is over capacity immediately after its construction.
Scenario planning is an important element of many statewide, regional, and even local studies. It provides an important framework to look at alternative concepts and facilitate broader discussions with key stakeholders. However,, most scenario planning projects rely on looking at the past to project the future. When the future isn’t that different from today, using the past as a basis to forecast would be an appropriate approach. In an era of unprecedented change in our society and travel behavior, the use of tools likeTrendLab+ that look at changes in these trends becomes a critical element of our work and an approach that should be implemented as broadly as possible.