Before becoming chairman of the Federal Reserve, Ben Bernanke said it was a "top priority" to "maintain continuity" with Alan Greenspan's leadership. However, in a new study, Haas Professor John Morgan questions the power structure of the Federal Reserve and determines that a single, powerful leader may not be the wisest way to create sound economic policy.
In a paper titled "Do Monetary Policy Committees Need Leaders? A report on an Experiment," Morgan and coauthor Alan Blinder of Princeton University examine whether designated leaders of monetary policy committees make a difference. Their collaboration resulted in a surprising finding: "Groups without such leaders do as well as or better than groups with well-defined leaders" when making monetary policy decisions, Morgan and Blinder conclude in their article, which was recently published in the American Economic Review.
The topic first arose during a Princeton cocktail party ten years ago. Morgan, who holds the Gary and Sherron Kalbach Chair in Business Administration and is a member of the Haas Economic Analysis and Policy Group, and Blinder were talking and wanted to know if the Federal Reserve's "will of the maestro" approach is the right way to determine monetary policy. "Our results indicate a loosely led group may be the right strategy," Morgan says.
Morgan used the Xlab, an experimental social science laboratory at Haas, to conduct a four-month experiment to reach those results. Morgan, founding director of the Xlab, recruited 200 undergraduate students who had taken at least one macroeconomics course. First he studied committee size. Students were assigned to groups of four or eight, then given macroeconomic equations for inflation and unemployment, both resembling the US economy. Their job: pretend to be real-world central bankers and control the nominal interest rate. "The situation becomes very real when subjects are put in charge of monetary policy decisions," says Morgan.
The experiment then introduced a realistic "shock" to aggregate demand. Morgan studied how quickly the groups hit inflation and unemployment targets by raising or lowering the nominal interest rate. Raising the rate lowers inflation and raises unemployment; lowering the rate does the reverse. The larger groups barely outperformed the smaller groups. The lab scored the quality of each student's performance and the person with the highest score became the designated leader of the group.
In subsequent experiments, Morgan and Blinder found individual leaders worked just as fast as a group and more accurately. However, the group emerged stronger in terms of overall performance and skill. The paper explains, "While the leader may be the best player, he or she seems incapable of improving the performance of the group." Morgan adds, "There's a sort of magic to the group and it made better decisions systematically than the lone wolf decision maker model many Feds used for years."
The paper raises questions about the "strong leader" model of the Federal Reserve and its formal chairman. Morgan contemplates whether organic, more collegial leadership may be more effective. "Maybe the Senate's line of questioning for Bernanke and future chairmen shouldn't be so much about the intellective criteria but other criteria that facilitate group interaction," Morgan says. "Our results suggest that the European bank model of seeking compromise and consensus can, in fact, be highly performing."