Inflation: Man vs. Machine
Wednesday, October 26, 2005
Before Mr. Bernanke came to Washington -- and long before President Bush nominated him to run the Federal Reserve -- he proposed creating a machine to crank out the central bank's interest-rate decisions.
Ben S. Bernanke, then chairman of Princeton University's Economics Department, worked with a colleague in 2001 to design a mathematical model that could absorb economic information and recommend how to adjust short-term interest rates to keep both inflation and unemployment low.
"We don't mean to suggest seriously that machine will replace human in [Fed] policy-making," Bernanke and his co-author wrote in an academic paper. "But having such a system would have several advantages," similar to having an automatic pilot on an airplane, they said.
That was Bernanke the professor, sharing his musings with fellow theoreticians. But some critics and observers worry that Bernanke, if confirmed as Fed chairman, may seek to put the central bank on a type of autopilot. Before joining the Fed board in 2002 and moving to the White House in June, Bernanke spent much of his academic career researching and advocating inflation targeting -- the practice in many countries of establishing a public, numerical inflation goal.
Bernanke's many papers and public remarks provide a window onto his thinking on the subject and reveal that he may have better political instincts than critics imagine, according to colleagues and analysts familiar with his work.
Skeptics of inflation targeting, including retiring Fed Chairman Alan Greenspan and some other Fed officials, question whether such a system would unduly limit the Fed's flexibility to lower interest rates in a financial crisis or an economic slump. They question whether the Fed needs such a system when it has managed well without it.
If Bernanke does try to move the Fed toward targeting, as many analysts expect, he will need to convince both his Fed colleagues and Congress that the benefits trump the potential disadvantages.
Soon after joining the Fed, Bernanke emphasized the importance of human judgment and flexibility in any inflation-targeting approach. He suggested moving toward a targeting system in small steps. He also discussed possible tactics for winning congressional support for such a change.
"There has been an evolution in his thinking. . . . He is much more aware of the subtleties of making this work in the American political process," said Frederic Mishkin, a Columbia Business School professor and longtime collaborator with Bernanke on inflation-targeting research. "He is not an egghead academic who doesn't get it. He knows how to learn and operate in the real world."
Bernanke, chairman of the president's Council of Economic Advisers, will need to demonstrate those skills first to members of the Senate Banking Committee, who plan to hold hearings and vote on his nomination before the end of the year.
In his first remarks as Bush's nominee Monday, Bernanke said his priority would be to "maintain continuity" with Greenspan's policies.
But an effort to move the Fed toward adopting an inflation target would be "a pretty significant departure" from Greenspan's approach, said Sen. Paul S. Sarbanes (Md.), the ranking Democrat on the committee.
Sarbanes said he will want to explore Bernanke's commitment to the Fed's "dual mandate": Congress's instruction that the central bank use monetary policy -- primarily the adjustment of interest rates -- to accomplish two goals, keeping inflation low and promoting a healthy level of employment.
When the Fed lowers rates, lower borrowing costs prompt consumers to spend more and businesses to expand, creating more jobs. That is good during a slump, but it can push inflation higher if the economy is growing at a healthy rate. Conversely, higher rates cool growth, which dampens inflation pressures but can lead to layoffs.
Fed officials unanimously agree that low, stable price inflation provides the best condition for economic growth, which generates job creation. "I subscribe unreservedly to the . . . dual mandate," Bernanke said in a March 2003 speech.
Some observers have doubts. "Bernanke is a creature of academic orthodoxy," said Thomas Schlesinger, executive director of the Financial Markets Center, a nonprofit organization that follows the Fed. "I'd be concerned about Bernanke's advocacy of inflation targeting. It poses real questions about the Fed's already incomplete commitment to full employment."
Bernanke sought to allay such concerns in a speech on the subject two years ago. Speaking as a Fed board member, he proposed "an incremental move toward inflation targeting, in the form of the announcement of a long-run inflation objective."
The Fed could say, for example, that it seeks to guide the inflation rate to about 2 percent over the long term. But it would include no specific time frame and no requirement that it hit that goal, according to the text of his speech. That "might help the Fed communicate better and perhaps improve policy decisions as well, without the costs feared by those concerned about potential loss of flexibility."
Bernanke concluded by discussing how to address potential congressional concerns that an inflation goal might undercut the dual mandate.
"The entire rationale for the [inflation goal] can be expressed in terms of jobs and growth," he said. "We post the objective in terms of inflation only because that is what the Fed can control in the long run."
Even before Bush's announcement Monday, some White House officials and analysts questioned whether Bernanke might lack the business and financial-markets experience needed to guide the U.S. economy through booms, busts and potential crises as skillfully as his predecessor.
His paper on the interest-rate machine might bolster or ease such concerns.
Bernanke and his co-author ran a test comparing the decisions their "expert system" recommended with those made in the real world by Alan Greenspan since he became Fed chairman in 1987. The machine lost. Its suggestions would have produced about the same average inflation but higher average unemployment than the Fed's policies.
"We are moderately encouraged about the potential of an expert system to help policymakers," they wrote, concluding, "However, there clearly remains considerable scope for human judgment about special factors or conditions in the economy in the making of [Fed] policy."