Finding New Applications for Economics
To call Charlie Plott an economist may be accurate, but it does not come close to doing justice to the depth and breadth of his five-decade career.
Since Plott, the William D. Hacker Professor of Economics and Political Science, joined the Caltech faculty in 1971, he has studied some topics one might expect of an economist, like stock market behavior, price controls, and monopolies. And more fundamentally, he contributed dramatically to the development of experimental laboratory methods in economics. But he has also focused on things that do not at first blush seem within the realm of economics, such as opening-weekend movie revenues and political decision-making.
Most recently, his economics research has moved in another unexpected direction: transportation for special-needs students in Australia.
The problem
Australia is a big country—almost as large as the United States in size—but it has less than a tenth of the population of the U.S. This means that the population is, on average, much more spread out, especially outside of big cities.
Because people in rural areas are so sparsely distributed, it can be a challenge to transport children to and from school. For parents of students with disabilities, the problem can be even more daunting. Although public school systems do provide transportation, creating effective bus routes can be a logistical nightmare.
"One kid might need a supervisor, while someone else's kid needs a wheelchair, and another kid needs a restroom on the bus. Some of these kids need the equivalent of an ambulance to take them to school," says Plott. "These are very difficult problems. How does one decide which bus picks up which child? Which route does the bus take? How are the buses equipped?" Because these are special-needs children, equipping all buses to attend to the needs of all children would be a very expensive undertaking.
"Currently the transportation provided by the government is one size fits all, and services are not good," Plott adds. "If you're in one place picking up some kids, other kids have to wait. Some of these kids ride the bus up to two hours going to school and another two hours going home. The question posed by the research is how to engage the services of the local transportation industry and channel competition in ways that match the preferences of the families."
Plott says this school transportation issue constitutes what economists call a "lumpy" economic system. A bus and its route are, in essence, a big lump of resources that cannot be divided up among consumers with different needs. The task can be done by many buses and many routes, but then the economics of shared services is lost. Things get very expensive.
Lumpy economic problems do not respond well to normal decentralized market forces like supply and demand, Plott says, so it has typically fallen on a government administrator to make decisions about how resources get allocated within such problems.
"In principle, these problems have features that prevent any kind of market solution, so you have a government agency taking over. Some administrator has decided who gets what, who's going to do what, how it's going to be done," Plott says. "As economists, we know that that's typically very inefficient. If you could make it more efficient, you would pay less for it. You could make people better off, including taxpayers, if you could find different ways of doing it. The question is, can we find market-based processes for solving these problems that work better?"
The solution
If a school district only has a small number of disabled students to transport and all have similar needs, it is relatively simple to create a properly equipped bus and a bus route that accommodate their needs. However, each additional student makes the problem exponentially more complex. Just a few dozen students can result in millions of possible routes, and that is before the special needs of each student are accounted for.
In developing their solution, Plott and his research team started with a school in northern Melbourne with only 80 students. The relatively small number of students to pick up plus the fact that the area is urbanized made the problem more manageable, he says. The team gathered data on factors such as the locations of each student, the needs of the students, and how long it takes students to get on and off a bus—information that helped them build a mathematical model of the most efficient routes to use. Their model helped them settle on seven routes and seven vehicles that would cover all 80 students. Next, they designed a special auction process that allowed the seven routes to be serviced by seven or fewer private bus companies. But here the researchers ran into a new problem. The bus companies were not familiar with the new competitive system and preferred the traditional method of granting all routes to a single company that would organize its own routes and services. Because of this, there were so few competitors participating in the auction that it would be easy for them to collude, causing the routes to be auctioned off at higher prices, Plott says.
Fortunately, Plott has more than a little bit of experience with auctions. He's been studying the economics of auction design for nearly 40 years.
"Charlie Plott and his collaborators confronted similar problems in the 1980s, when working on railroad car allocations, and in the early 1990s, when developing the auction protocol that the Federal Communications Commission used to sell parts of the radio wave spectrum," says Jean-Laurent Rosenthal, the Rea A. and Lela G. Axline Professor of Business Economics and Ronald and Maxine Linde Leadership Chair of the Division of the Humanities and Social Sciences (HSS). "In both cases, the mechanisms Charlie designed made a big difference, and his ongoing research continues to have real-world impacts."
In the type of auction people are most familiar with, bidding continues as long as bidders keep raising the price, and only one item gets auctioned at a time. What Plott proposed was a simultaneous auction of all seven routes. If a company was outbid on one route, they could hop over to another route and bid on that one instead. To help prevent collusion between bidders, the entire auction was computerized. Companies that participated were unable to see how many other bidders there were nor the identities of the other bidders. All they could see was the current price and how much time was left in the auction. Last-second bidding was avoided by new methods of ending auctions. The new method worked, and the school ended up paying less for the bus company contracts than it previously had.
"We're saving money here. As a matter of fact, we were about 5 to 10 percent under what the contracts had cost in previous auctions," Plott says. "And since we're talking about roughly $6 million here in sales, everyone was very happy with that. Even the bus companies were happy because the auction was so fast."
However, the greatest beneficiaries under the new system were the students. Plott says his team received a letter from the mother of one of the students letting them know her child's commute time on the bus dropped from two hours to just 30 minutes.
Plott says he hopes that this research will ultimately be applied to more than just school transportation issues.
"The transportation issue reflects problems that are found with the procurement of government services in general," he says. "Government service procurements involve a wide range of issues and are provided throughout the economy. How they are procured impacts both the cost and the quality of service. It could be housing for the elderly, it could be Meals on Wheels, it could be flood control. The government provides thousands of services. If I can solve this one, then we have a model that could be applied in many places."
"Charlie has had major impacts on several fields within economics and politics. The key to his success is that he is both scientist and engineer: he attacks the fundamentals and develops new theories and paradigms while at the same time creating applications that solve real problems," Rosenthal says. "Unlike most economists, he does not just dream up policies. He subjects them to rigorous experimentation in his laboratory. It's a hard furrow to plow, but a half century in, it has been an extremely bountiful one."
Caltech's Division of the Humanities and Social Sciences is holding a symposium March 2 to honor Plott's long career and his many contributions to his field. The symposium will be held in the Beckman Institute Auditorium from 8 a.m. to 5 p.m. The program for the symposium can be found here. To register for the symposium, click here.