A Q&A with Behavioral Scientist, Richard Mathera, on using behavioral economics interventions to transform transportation in an urban environment.
No one likes being stuck in daily traffic on the way to work. And yet, we endure. The average commute is just shy of 30 minutes. Commuters not only suffer with bumper to bumper traffic, but worse, they tend to be less physically active, have higher rates of obesity and lower overall well-being!
In this article we talk to Richard Mathera about what has worked (and what hasn’t) to improve our roadways and commutes. He covers why nudges aren’t always the answer, examples of systemic changes that could work and the role of incentives.
A couple of notes:
- While COVID-19 has drastically changed our commutes along with our mental models around them (we’re more concerned than ever before about getting on a crowded bus), we also see a potential opportunity in the disruption. Humans are more likely to change their habits and routines after a major disruption—like, for example, a global pandemic. Note that the following recommendations are for sustainable travel in post-pandemic conditions.
- While there are a lot of issues with the current system, we’ll be focusing our discussion on reducing single occupancy rides and increasing uptake of carpooling and bus lines, or multi-occupancy rides.
Q: What are examples of past or present transportation programs and interventions that use a behavioral economics approach?
Several travel programs have shown success at scale in reducing fatalities and congestion:
- Sweden’s Vision Zero project prioritized the reduction of fatalities and thus slower speed limits. This ended up saving lives.
- London’s congestion charge in the early 2000s significantly reduced traffic.
- Cities like Seattle, San Francisco, and Everett, MA have created bus-only lanes, reducing bus and car commute times and increasing public transportation reliability.
However, research has also found that achieving travel-related behavior change is difficult when only using small nudges.
These smaller nudges were unsuccessful at driving an increase in multi-occupancy travel:
- Letters to encourage carpooling, including those with peer testimonials (the testimonials appeal to our desire to follow social norms).
- Emails to registered carpoolers to increase carpooling, including: a) A matching process that assigned people to carpool groups and b) Loss framing: highlighting the opportunity cost of driving to work every day.
- Giving people free bus passes, including specific routes to take.
- Emailing personalized travel plans to people based on where they lived, including an on-call counselor that would provide advice to employees.
- Offering lottery tickets to bus riders for a weekly cash prize.
Rather, the most successful travel-related behavior change initiatives employ systematic interventions.
We’re talking about top-down, comprehensive changes to the existing infrastructure. Given what we know about current behavioral interventions, we believe we need to significantly change the decision-making environment for commuters.
Given what we know about current behavioral interventions, we believe we need to significantly change the decision-making environment for commuters.
Q: How do you systematically change the decision-making environment for commuters?
We need to start by entirely reframing our commutes. To discourage single occupancy commutes, we need to make them harder, costly and unattractive. On the other side, we need to make shared transit easier, cheaper and more attractive.
This is a bold, big idea but it’s not new. System-level initiatives are the foundation of our current transit landscape, which is centered around cars. In the 20th century, American cities were built for cars on the assumption of 1 person = 1 car. And this has become the unchallenged default.
If we’re going to get commuters to adopt multi-occupancy transit, we need to build our cities’ infrastructures so that the easiest and most attractive option is shared rides and public transit.
Q: What kinds of interventions would make single occupancy commutes more difficult and less attractive?
Barriers increase friction which ultimately decreases the likelihood of a specific behavior.
Behavioral barriers to deter single-occupancy commutes include:
- Closing specific roads to commuters (either always or at specific times).
- Restricting roads to High-Occupancy Vehicles (HOV)-only.
- Increasing the cost of driving a specific route (tolls).
- Limiting which cars can enter a city on a specific day.
Examples of other possible more nuanced, behaviorally-informed approaches include:
- Limiting some forms of transportation for short time periods and asking people to try new modes of transport. London found that people took new, more efficient routes to work after a travel disruption, suggesting that the status quo bias has a high impact on our travel decisions.
- Expanding our mental model of transportation options. Mental models are our preconceived understanding of how things work and often bias our actions or social behaviors. People likely have not considered some forms of transportation as possibilities because of certain mental models. For example, some Parisians may have a mental model that the metro is only used by tourists, while buses are used by locals. Expanding the mental model for what transportation options fit into a set of choices will be important for getting people to consider new transportation arrangements.
- Paying people up front to take a different form of transportation. After an initial period of payment to create the habit, these payments could be ratcheted down over time.
- Making the costs of driving more salient. As our Auto Loan Calculator illustrates, the costs of driving and car ownership are largely invisible. When you take the train, you see the cost every trip you take. But when you drive, you don’t observe the price of the ride. The fact that you fill up your gas tank or pay for insurance infrequently means that when you drive, it seems like you’ve already paid for the ride. Behavioral approaches that make each trip more salient at the point of driving could lower the likelihood of people driving.
- Charging for parking. Parking is an ideal opportunity to making the costs of driving salient. If you remove free parking and charge someone to park every time they want to park, you are increasing the salience of the cost of driving and inserting a shock to their driving habit system that may ultimately impact their transportation decisions.
Q: What’s the process for developing a behaviorally-informed transportation intervention?
First, it’s important to conduct a behavioral diagnosis for the existing modes of travel. This means understanding behavior at the most granular level possible. We use quantitative and qualitative methodologies to ensure a deep understanding of the challenges and opportunities that exist for high-impact, cost-efficient interventions.
- Collect Data About the Problem. We review and analyze existing data sets about the problem (commute patterns, congestion, accidents, current incentive programs and how they have worked).
- Synthesize scientific literature in the relevant domains, highlighting tactical insights that can influence the hypotheses we make.
- Map what actually happens. What are people doing? We do this mapping with a high level of attention to detail. We would follow commute lines and measure ability to park at public transit stations, time to wait, noise on trains, duration of delays, etc. This is where we want to get specific–every decision point should be mapped, regardless of how trivial that decision may seem. Every decision point is a barrier that must be tracked (e.g. the decision to drive my car to work vs. taking the bus is a barrier).
Q: What role do incentives play in a behaviorally-informed transportation intervention?
Where barriers deter people from doing specific behaviors, incentives play a massive role in motivating people to do specific behaviors. A variety of incentive types could be leveraged to shift travel behavior.
If monetary incentives are used, they should be used to create a habit, with relatively high amounts paid at first to encourage adoption. Over time, the amount and salience of these payments should decrease, allowing more sustainable incentives, such as social or experiential reinforcements, to maintain the habit.
To the extent that monetary disincentives are used, the pain of paying should be salient and maximized each time someone drives. At the same time, the pain of paying for public transportation should be eliminated through prepayment or removed altogether (if possible) to leverage the particular psychological appeal of free items.
The biggest non-monetary influences associated with transportation are hassles. Intuitively, hassles should be increased to make behaviors we want to discourage more difficult and reduced to make behaviors we want to encourage easier.
Other non-monetary reinforcements include: social norms, accountability, lottery psychology, salience, mental models, identity, and others. Our team has achieved significant results without the use of monetary incentives. These can be explored further during behavioral diagnosis.
Q: But wait, what about COVID?
The pandemic represents a unique moment for human behavior change but also a huge amount of uncertainty vis-a-vis people’s future behavior. We are at an inflection point, and it is unclear what will happen in the post-pandemic world. Considerations include:
- Many people have stopped commuting. Intervening as they return to a regular commute offers a very strong opportunity to create new habits. We recommend pursuing this immediate opportunity to change transport habits. People’s transportation decisions during the pandemic will not necessarily be reflective of their future transportation decisions, so any experimental result during the pandemic may or may not hold.
- If COVID-19 does not ever subside, people may shy away from public transportation in perpetuity, increasing single-car usage and congestion. The public transit or shared ride system may have to support new standards (cleanliness, distancing, etc.) to draw ridership.
- In the long-term, will a new remote work culture and the high cost of living encourage people to move out of cities and/or drive less permanently, reducing congestion? Furthermore, the continuation of some remote work norms may be a way to decrease single occupancy commute rides.