One year ago, Professor of Government Kenneth A. Shepsle was in the middle of the country, choosing between coasts. Then a professor of political science at Washington University in St. Louis, he had turned down several offers from schools around the country and was still weighing attractive offers from Stanford and Harvard.

"I had great doubts about the west coast. I think it's a place one goes for a year to revitalize oneself, but then one has to go back to the real world. The east coast is the real world--it's fast, it's competitive," Shepsle says. And as he puffs on his pipe, Shepsle certainly looks more Eastern Establishment than California Bohemian.

And so, despite the fact that Shepsle, a shining star in the branch of political science known as mathematical modeling--had taken two sabbaticals at Stanford, Harvard got him.

Mathematical modeling of political behavior takes a more scientific approach than traditional political science, and it is still somewhat controversial. It is distinct from quantitative political science--modeling is concerned mostly with the logical flow of ideas, whereas quantitative political science is concerned mostly with statistical relationships.

If one engages in mathematical modeling, "one does look for simplified mechanisms, but let's not fool ourselves--that's what all science does," says the Washington, D.C. native. "I don't think that any science prospers without good observation. Much of what we want to understand in political science are the regularities we observe."

Shepsle was a math major as an undergraduate, at the University of North Carolina, Chapel Hill. When he was closed out of a mathematical physics class as a junior, his advisor suggested that he take a political science class instead. He did--and he loved it, and went on to graduate studies in the field at the University of Rochester.

Mathematics mixed with political science was "an extremely novel combination at the time," Shepsle says. "Twenty years later--well, I don't want to say it's mainstream, but it's certainly a respected tradition." And mixing math with government is certainly what Shepsle is good at. "Ken combines an interest in theoretical government with expert empirical hoking and poking around in real institutions. There are not very many people who do his sort of inductive analysis who have also spent time hanging around Congress interviewing people," says Robert H. Salisbury, chairman of Washington University's political science department.

Shepsle sees his work, which combines science-like theory with the study of concrete institutions, as partly a reaction to the behavioral revolution in political science of the 1950's and '60's. He says that before this 'revolution,' students of government often wrote detailed, voluminous tomes which were all descriptions of institutions, such as Congress, without including any theory. "But a little too much of the baby was thrown out with the bathwater," he says. "The revolution was correct in that detailed description was not enough, but the baby that was thrown out was the institutions themselves."

The result, Shepsle says, was an extremely individualistic view of government, which lacked the structure provided by a study of the institutions.

Shepsle's research often connects theory to practice, through in-depth study of institutions like the U.S. Congress--he has studied legislative politics, congressional committees, budget processes, and voting strategies, among other things. In one of his works, Shepsle claims that the nature of congressional subcommittees leads to uncoordinated federal growth and unnecessary, uncontrollable expenditures.

Since 1970, he has written two books, edited two others, and penned more than three dozen articles. Much of his work was done in collaboration with other scholars--"I love collaborating," he says.

Currently, the 41-year-old professor is beginning a long-term study of congressional conference committees--such as the recent tax committee--which are established to reconcile the House's version of a bill with the Senate's. "I'm primarily interested in this as a theorist--in a sequence of processes, having the ability to affect things late in the game is a great advantage."

Shepsle says that while he enjoys teaching, research is what he likes best. "It's the party-line--I've been taught to say it. You teach to put the bread and butter on the table, but your true love is your research," he says. "And it's absolutely true, especially at first. Also, to be any good at teaching, you have to be an active researcher, and vice versa."

He said that students should come to schools like Harvard realizing that research is the priority, but rather than resenting it, should be thankful because it guarantees that they will have an active, vibrant faculty.

Smart Choices