Travel Demand Modeling in the Boston Region: TDM23 as a Tool for Transportation Planning

by Tegin Teich

The Boston region’s transportation system is undergoing a generational transformation following the opening of the Green Line Extension in 2022 and the substantial investments made possible by the Bipartisan Infrastructure Law. Over the next twenty years, the Boston metropolitan area will see significant changes to the roadway and public transportation networks on which people walk, bike, roll, ride, and drive throughout their communities. The reimagination of highways and boulevards in Allston, Somerville, and Charlestown are but a few examples of how cities and towns in the region are engaged in planning transportation networks that allow for easy access to multiple forms of travel. While the infusion of billions of dollars in federal funding creates these transformational opportunities, the region will be contending with the effects of increased remote work on travel patterns and transit ridership, as well as technological developments such as electric bikes and trends such as ridesharing.

The choices made today about building public transit and roadway infrastructure will have lasting effects for decades. With that responsibility in mind, planners and leaders must make the best use of available data and methodologies when making investments to create the equitable, safe, modern, and multimodal future desired. Travel demand modeling is an important method at our disposal. As a public agency committed to creating tools that are useful companions in transportation planning, design, and decision-making, the Boston Region Metropolitan Planning Organization has been working over the last two years to develop a new travel demand model, TDM23.

In transportation planning, we use travel demand models to explore how transportation infrastructure decisions made now might influence the future. Like most models, travel demand models start with representations of current conditions. In the case of travel demand models, this includes the demographics of people who live in the region, where they live, where jobs are and what types of jobs they are, existing transportation infrastructure, and how people choose to travel to work, to run errands, and to socialize. 

From that representation of today, models can then be modified with inputs representing how we believe any of those factors will change in order to enable the exploration of potential outcomes of those changes. Outputs of travel demand models can be used to compare different infrastructure plans based on estimations of traffic and transit ridership that we think projects will generate, greenhouse gas emissions that could be produced, and potential effects on populations that have historically experienced negative impacts from the transportation system—low-income communities and communities of color.

A model is not a crystal ball but a compass. Navigating a ship with a compass serves to illuminate the direction, but the currents, winds, and weather influence the actual journey. Similarly, a travel demand model can chart potential futures, but actual outcomes depend on variable factors such as economic shifts, policy changes, technological advancements, and leadership decisions.

So, while we strive to build models of the future based on the best data available, they are still projections, aggregated and summarized, and not inevitabilities. We want models to be as accurate as possible when we compare the outputs with future data, but we never expect them to be “right.” In fact, we know that many events will happen that we did not know about when we carried out the modeling—from a significant new technological development to a global pandemic—that could detract from the accuracy of predicted outcomes. In the face of such uncertainty, the best way to critique whether a model has been useful is to assess whether it informed decision-making in a way that helped reach the goals established at the beginning of the planning process.

Despite its complexity, TDM23 must be an understandable and reliable resource if it is going to serve as a common framework for making transportation decisions. That is why we released it as a publicly accessible resource with dynamic documentation that makes its structure transparent. Following the release, we published the TDM23 Users Guide, which includes visualizations designed to enable those with less software development experience to explore examples of outputs. This is a new way to approach and define the practice of travel demand modeling. Together with partners in public agencies, academic institutions, and beyond, we can improve this resource and navigate to a better transportation system for the future.

Tegin Teich is the Executive Director of the staff of the Boston Region Metropolitan Planning Organization, the public agency responsible for carrying out the federal transportation planning process in the region. To learn more about the Boston Region Metropolitan Planning Organization’s TDM23 travel demand model, visit bostonmpo.org/travel-demand-model.

Read this article on the MPO's blog.