“Rational choice has come quite a way,” a Stanford University professor, Barry Weingast, said. “Thirty years ago, it was far outside the mainstream in political science in almost every respect.” Mr. Weingast led a roundtable discussion at the American Political Science Association’s annual meeting in Philadelphia recently, focusing on the influential growth of rational choice theory, which assumes that people are self-interested agents out to maximize power or wealth. The theory uses mathematical models and posits that people exercise instrumental reason in making utilitarian choices.

A University of Washington professor, Margaret Levi, spoke first, saying rational choice was now deeply embedded as a theoretical approach in comparative politics. She said she was originally among a marginalized group within that field. But over time, she said, rational choice began to influence the way political scientists thought about questions such as ethnic identity or bureaucratic behavior.

More recently, she said, comparative politics has come to use a combination of approaches and theories, and rational choice theory has been absorbed: “There is almost no one who doesn’t recognize that there are free-rider and collective-action problems, incentives, interests, cost and benefits that have to be calculated.”

Mr. Weingast next introduced a Stanford colleague, Terry Moe, who he said was “one of rational choice’s greatest critics and most creative and unique appliers.”

Mr. Moe opened by noting that it was fair to say that rational choice had taken the discipline of political science “by storm.” He gave two reasons: first, it had huge analytical strengths. More than a methodology, he said, rational choice was “a family of theories” or “a technology for developing theories.”

At its heart, it involves simplification, focusing on what appears to be absolutely essential, and building models on that basis. The assumptions that rational choice theorists make, he said, are clear: “This is unusual in traditional political science.” A Harvard professor, Kenneth Shepsle, likewise said, “We find a much healthier respect in the discipline for science.”

Mr. Moe also said rational choice had helped integrate the discipline of political science.”This is a really well-developed, coherent, analytic, powerful body of literature. It sounds corny to say so, but I never thought we’d see this sort of thing in my lifetime.”

Although Mr. Moe said he was a big proponent of rational choice theory, he also spoke of a “downside.” Mr. Moe said simplification could drive out what is important and interesting. “A lot of models are trivial and worthless. They don’t help at all to understand politics, and some are flat out misleading.”

He cautioned that because rational choice involves math and formal models, practitioners can sometimes become too absorbed in technique and equations, which could lead others to ask profitably, “Who cares?” Mr. Moe said rational choice theorists should be able to demonstrate that their models matter, by pointing to an empirical phenomenon that they help to explain.

He said it’s not true that anybody who could do math well could necessarily become a good modeler and great political scientist.

Mr. Moe also conceded that rational choice, rooted in voluntarism, did not always adequately address certain topics such as power. There was another side to politics: coercion and force. A Yale University professor, Ian Shapiro, a co-author of “Pathologies of Rational Choice Theory: A Critique of Applications in Political Science” (Yale University Press), concurred, saying rational choice could be “rather innocent in its assumptions about power.”

Mr. Weingast pointed out the work of a University of Rochester professor, William Riker (1920-1993), as one of the “main progenitors” of rational choice theory in political science. From that “frigid and isolated location in upstate New York,” Mr. Shepsle said, his teacher developed other outposts at the California Institute of Technology, Michigan State University, and elsewhere. Mr. Weingast said Riker likened those early years “planting flags,” one at each institution.

Mr. Shepsle said a deep respect for political history as a laboratory for ideas and evidence took root at Rochester. Likewise Ms. Levi said some of the best work nowadays combines statistical and rational choice approaches with qualitative research (“in-depth understanding of a particular place and time and period and people”).