Daniel McFadden, "The Behavioral Science of Transportation"
Daniel McFadden delivering the keynote address at the World Conference on Transport Research
In his keynote address to the World Conference on Transport Research in Berkeley on June 25, 2007, UC Berkeley Nobel Laureate Daniel McFadden urged transportation researchers and policy makers to employ insights from economic and behavioral science to make sure that human behavior doesn't get left out of the calculus as they design transportation systems. He emphasized that they must not overlook the way people use and perceive transportation as individuals and the way they shape it collectively through their political, economic, civic and cultural institutions.
McFadden has direct personal experience in such cross-discipline thinking about transportation. An economist by training, he put his theories about consumer behavior and econometric forces to an early test in a pioneering study of mode choice intended to predict how many commuters would switch to the Bay Area Rapid Transit (BART) system, two years before its 1975 opening. His study predicted that BART would entice about six percent of Bay Area commuters to change modes, less than half the official estimates when the rail system was proposed. When BART opened for business, McFadden's prediction proved to be correct.
McFadden was awarded the Nobel Prize in Economics in 2000 for his discrete choice theory, which evolved with his work on BART, which he did in association with the Institute of Transportation Studies.
Embracing behavioral science runs counter to traditional transportation thinking, he acknowledged.
"I asked him if he would be working on the Southern Crescent, the premier passenger train. 'Oh no,' he said, 'if I did that, I would have to deal with people. Railroad men would rather work with freight.'"
McFadden told a story to illustrate the old way of thinking. "When I was eight years old, a neighbor was promoted to conductor on the Southern Railroad. I asked him if he would be working on the Southern Crescent, the premier passenger train. 'Oh no,' he said, 'if I did that, I would have to deal with people. Railroad men would rather work with freight.' Today, it is important for transportation workers, and transportation researchers, to recognize that there is no escape from humans and the impact of their behavior on transportation systems. One has to work with people. "
McFadden outlined the three levels at which human (or "consumer") behavior has been modeled over time:
- physical analogies, such as gravitational attraction to explain movement of people and goods,
- economic theory, in which it is assumed that transportation users make choices that maximize their preferences, and
- more recently, popular behavior that can't be expressed by strictly rational constructs, but which requires analysis with tools from sociology, anthropology, cognitive psychology and brain science.
He illustrated the use of economic models with three examples:
- the linkage of demand and utility by 19th-Century French engineer Jules Dupuit, also the namesake of the WCTR Society's highest prize,
- the use of Newton's law of gravitation by Ravenstein in 1885 and Zipf in 1946 to explain the movement of people within a transport system, and
- disaggregate travel demand forecasting models developed by Domencich and McFadden in 1972, which used random utility maximization, and which they used to make their BART projections.
"Dupuit is one of the founding fathers of both transportation science and economic theory," McFadden noted. "He recognized that a product will be demanded up to the point where the dollar value of the marginal utility of an additional unit purchased falls to the opportunity cost of that dollar amount." This gave transportation planners the tool to quantify the benefits and costs of policies such as setting tolls and to evaluate investments in new infrastructure.
The gravity model postulated in 1946 by Zipf was widely embraced as a way to make long-range travel forecasts, McFadden explained, though it does not work as well under heterogeneous conditions and does not reconcile easily with individual choice or decision models.
The final type of model, which he and Domencich developed, is based on the premise that people have different tastes and needs. Coupled with observed individual (or "discrete") choices, it can be used to estimate utility and how it changes when the transportation system is altered, such as whether BART is a commuting option or not. While these models have been used regularly, he pointed out that there might be less awareness that they can be used for broader questions over larger geographic regions and longer time periods than many researchers currently do.
This sort of thinking about models is especially important given the interest in using market forces, such as congestion-pricing, to wring more efficiencies out of the transportation system. Although congestion pricing is relatively easy to implement, and has been demonstrated to work well under the right conditions, there is still strong resistance, McFadden noted.
From the traditional economics point of view, this irrational resistance to congestion pricing is a puzzle. To get the answer, McFadden said, it is necessary to turn to cognitive psychology and brain science. People feel their losses more deeply than their gains, so that, for example, the majority of drivers in a traffic jam are convinced that the other lanes are moving more rapidly than the one they are in: a logical impossibility. This loss aversion results in an innate fear of change in the status quo and, by extension, market-based schemes that require consumers to choose among a menu of options, all of which require them to act differently.
The cause for this aversion to change? The way our brains are wired, according to neuroscience research. Our evolutionary path results in our processing potential threats at a much more primitive level of the brain than where we process potential rewards. Trade, or participation in the marketplace, occurs at this very primitive level, with a large emotional component, which explains the cliche about sex and shopping activating the same neural regions.
Pelotons, the wedge-shaped groupings that bicycle racers form to conserve energy and limit wasteful sprinting, are an example of a human adaptation to this fear of being singled out for punishment. In exchange for giving up some freedom of choice, the peloton riders are rewarded with a more certain and more efficient outcome.
As a result, people are natively agoraphobic, afraid of the marketplace, because it can punish them if they make the wrong choices and because they don't trust themselves to be fully in command of their marketplace decisions. Pelotons, the wedge-shaped groupings that bicycle racers form to conserve energy and limit wasteful sprinting, are an example of a human adaptation to this fear of being singled out for punishment. In exchange for giving up some freedom of choice, the peloton riders are rewarded with a more certain and more efficient outcome—until they see their chance to break away.
McFadden posited the likelihood that a small group of congestion-pricing opponents have formed their own pelotons, and the majority of consumers have joined up out of a defensive reaction. He suggested work be done on better understanding pelotons as they apply to collective decision-making for achieving social welfare, or "how consumers pack together when they drive, ride transit, or vote on transportation projects.
"Perhaps by understanding the formation and stability of anti-tax, anti-road-pricing pelotons, one might see how to encourage pelotons that support efficient market solutions to congestion in transportation." Without careful framing of market solutions, the public will continue to resist them, he said.
Finally, he repeated his advice to transportation researchers to work closely with behavioral scientists and take advantage of new findings as they continue to be made, so that researchers can better understand and estimate the ways that human choices will shape and respond to transportation systems.
Other links:
- "The Behavioral Science of Transportation," Daniel McFadden address to WCTR, June 2007, PowerPoint file.
- Daniel McFadden's Nobel Prize Acceptance Speech, with an introduction by UCTC founding director, the late Mel Webber, excerpted in ACCESS Magazine, Spring 2002, the publication of the University of California Transportation Center.