“The Enigma of Alan Greenspan”

Remarks by
Greg Ip, senior special writer, The Wall Street Journal

Donald W. Reynolds Visiting Fellow
Washington & Lee University
Delivered Feb. 7, 2006
 

In a speech early last year in Scotland, Alan Greenspan said, “In the broad sweep of history, it is ideas that matter. Indeed, the world is ruled by little else… Emperors and armies come and go; but unless they leave new ideas in their wake, they are of passing historic consequence.”

Ever since Greenspan gave this speech, I’ve been asking myself: What ideas has Greenspan left us? Many ordinary people are familiar with Keynesianism, Reaganomics, and Milton Friedman’s doctrine of monetarism. Many economics students, by the time they’ve graduated, will have studied the Phillips curve and the Taylor rule. But 20 years from now, will students study “Greenspanism” or “the Greenspan rule?” I suspect the answer is, “No,” because the essence of Greenspan’s thinking is his distrust of any “ism” or rule.

Anyone who knows anything about Greenspan’s early years would snort in disbelief to hear him described as non ideological. He was of course a hard-core libertarian, opposed to government interference in almost any aspect of society. As an associate of the libertarian philosopher Ayn Rand in the 1960s, he inveighed against the Food and Drug Administration, the Securities and Exchange Commission, and antitrust laws, which he called a “jumble of economic irrationality and ignorance.”

Yet in his later years, he was more likely to describe himself as a follower not of Ayn Rand but of Thomas Bayes. Thomas Bayes was an 18th-century British Presbyterian minister who had early insights into decision making under uncertainty. Some things are relatively straightforward to predict: for example, the probability that a flipped coin will come up heads or the percentage of boys born last year who will grow up to be at least six feet tall. But many complex things like the U.S. economy are inherently uncertain because so many factors affect its performance and they are always changing. A Bayesian accepts that rules based on history can easily break down when applied to the future. He makes a decision based not on the most probable outcome but on a range of possible outcomes.

Greenspan declared in August 2003, “Uncertainty is not just an important feature of the monetary policy landscape; it is the defining characteristic of that landscape.” At the Fed, he and his closest colleagues would describe their work as “epistemology,” the study of knowledge and its limits.

This committed agnosticism may be the most important factor in Greenspan’s success. He does use models of the economy, but avoids getting invested in any. That makes it easier to shift gears when the model stops working.

In 1996, he concluded that productivity growth in the United States had accelerated, and thus the economy could grow faster and unemployment fall further without generating inflation than some of his colleagues maintained. The Fed delayed raising interest rates, which probably helped extend the 1990s expansion – though critics say it also contributed to a stock bubble.

Listen to this passage from a Fed meeting in 1999 where Greenspan explains how he’s concluded that productivity growth must be higher than generally realized:

As we all know, when econometricians get regression results that appear out of line with the real world … they have to look for the missing variable. I submit that there is a missing variable… I think it may be about time to try to substitute this variable for NAIRU (the non-accelerating inflation rate of unemployment). Let me put it this way: Neither one is an observable phenomenon, but neither was the planet Pluto before 1930. Scientists figured out that there had to be something there, given the extent to which Uranus and Saturn were deviating from their forecast orbits. Well, I submit that at some point we are going to come to the conclusion as statisticians that the simultaneity of a falling inflation rate and an ever tightening labor market is trying to tell us something.

This was just one case of his “missing variable” methodology. Another was in the early 1990s, trying to understand why the economy was growing so slowly. He concluded it was the credit crunch. Then, in the mid-1990s, when he was puzzled by the failure of wage growth to pick up as unemployment dropped, he struck upon the insecure worker hypothesis. And more recently, in trying to understand the low level of long-term interest rates and inflation world wide, he has concluded it must be globalization.

In addition to these insights Greenspan also had a lot of duds. In the year 2000, he argued that the economy would slow down of its own accord after consumers had bought all the cars and houses they could possibly want. In fact, housing sales went on to break annual records, automobile sales remained well above their 1990s average, and we are still waiting for consumers to reach their saturation point. Greenspan, as far as I know, hasn’t repeated this theory lately.

At least since the 1960s he has argued that the ability of homeowners to “extract” the equity in their homes through the mortgage market represents a unique channel of stimulus to the economy separate from the “wealth effect” of higher home prices. Though he has won many converts to this view, most of the Fed’s professional staff are not among them. As Greenspan himself has admitted, the theory is as yet unproven.

Even Greenspan’s productivity insight in 1996 had come to him in an earlier, less successful form. Back in February 1987, at Townsend Greenspan, the consulting firm he ran before joining the Fed, he wrote, “Indirect evidence suggests that productivity growth in many key nonmanufacturing areas of the economy, where high tech predominates, may have been significantly underestimated in recent years.” The problem was that the economy wasn’t behaving like this was true. As growth approached its old speed limit, inflation pressures emerged.

I recite these examples not to play down the significance of Greenspan’s successful insights but to try to identify something important about the way he thinks. He had a lot of ideas. He tested them. And when an idea didn’t hold up, he discarded it. When his productivity thesis failed to explain the economy’s behavior in 1987, he applied the old rules and tightened monetary policy. Indeed, in September, 1987, at one of his first FOMC meetings, another governor posited that “inflation may behave in an untraditional way” because productivity growth might be under-measured. Greenspan responded: “There is always something different (about an economic cycle). But there is nothing very unusual about this one.”
On the other hand, when the economy’s behavior suggested in 1996 that productivity had accelerated, he acted accordingly by not tightening monetary policy.

In general, it’s difficult for anyone, most of all policy makers, to constantly question their views of how the world works. Most of us have presumptions and preconceptions that we use to make sense of the world and to make decisions. We will cross the street because we assume a car will not come at us from the wrong way. We will not buy a used car from that man because we assume used car salesmen can’t be trusted.

Moreover, we often gravitate to people with strong, confident visions, even as others find those very same visions completely wrong.

“People for the most part dislike ambiguity,” Philip Tetlock, a University of California at Berkeley psychologist, has written. “Human performance suffers because we are, deep down, deterministic thinkers…. We insist on looking for order in random sequences.”
Tetlock has undertaken some fascinating empirical research on what makes someone a good decision maker. His book Expert Political Judgment chronicles a 16-year experiment testing the predictions of 284 experts in geopolitics, on questions as diverse as whether Japan would recover from its early 1990s stock market collapse, whether Quebec would secede from Canada, or whether Nigeria would collapse into interethnic violence.
Tetlock concludes: “What experts think matters far less than how they think.” Experts on either the right or left who have a single, unified view of the world are more likely to be wrong, and badly wrong. Such “hedgehogs,” as Tetlock calls them, “know one big thing.” They are less prone to self-doubt, more likely to dismiss evidence that contradicts their vision, and less likely to admit to mistakes. “Foxes,” on the other hand, “know many little things.” They “draw from an eclectic array of traditions, and accept ambiguity and contradictions as inevitable.” (He attributes the hedgehog-fox labels to philosopher Isaiah Berlin, who in turn traced them to ancient Greece.)

Some hedgehogs turn out to be great leaders, businessmen or scientists precisely because they doggedly adhered to a single, simple belief that turns out to be right. Foxes can be maddeningly hard to pin down.

Greenspan is a fox. He spoke in opaque prose and avoided precision because he thought the constantly shifting structure of the economy made certainty and precision impossible. In 2004 he said one of the few things an economic forecaster can count on is that a company’s inventories can’t go below zero. That, he said, is “probably the full state of my knowledge about how to make a forecast.”

Indeed, as Fed chairman Greenspan was always circumspect about forecasts, such as those developed by his very capable staff at the Fed. Perhaps that’s because of his own unimpressive forecasting record. Back in 1984, Greenspan and a partner got into the money management business. According to an article in Forbes Magazine, the fund’s selling point “consisted mainly of Greenspan's macroeconomic analysis of secular and cyclical trends.” But the fund’s performance, according to the magazine, was “barely passable … In 1985 … (it) turned in one of the least impressive records of all pension fund advisers.”
At his 1987 confirmation hearing, one of the Senators quoted from this article. Greenspan’s response to this critique was, “All I can suggest to you, senator, is that the rest of my career has been somewhat more successful.”

Milton Friedman once told me he thought Greenspan was very good at reading economic trends. When I pointed out to him that some had found his forecasting record to be unimpressive, Friedman replied, “I was only judging by the fact he was able to make a living at it.”

Indeed, Tetlock finds that foxes are not especially good forecasters. But, he says, they make fewer big mistakes than hedgehogs. For a central banker, that means having the occasional recession but avoiding catastrophes like the Great Depression of the 1930s or the Great Inflation of the 1970s.

Greenspan calls his own flavor of Bayesian decision-making “risk management.” It amounts to deliberately risking small mistakes to avoid much bigger ones. An example of this came quite recently. In early 2003, the recovery seemed stalled and inflation, already low, was going lower. Though the Fed discussed deflation, the risk of it seemed remote: it could probably have raised interest rates later that year without much harm. But if in fact deflation had occurred, such a strategy could have been disastrous: once prices and wages start falling, companies and individuals are crushed by their debts, and even zero interest rates may not stop the downward spiral. To be sure, keeping interest rates low to prevent this remote possibility risked inflation. But as Greenspan told Congress: “We know how to deal with inflation.” By contrast, on deflation, he said, “our knowledge base was virtually nonexistent.”

We do know how the economy performed under Greenspan: extremely well. He left office with unemployment and inflation both lower than when he took office. There were just two mild recessions and the longest expansion on record. Growth in the last few years has been the strongest of the major industrial countries.

We are not, however, certain what Greenspan did to bring this about. It could be partly luck. In the 1970s, we had a massive oil price shock and a decline in productivity growth. In the 1990s, we had low oil prices, an acceleration in productivity growth, and the integration of China into the world economy, all of which made it easier to keep inflation down. Moreover, the U.S. was hardly alone in enjoying strong economic performance. In the last 15 years, inflation has fallen in most countries, from Italy to Congo, in some dramatically. Australia, Canada, the United Kingdom and Spain have done as well or better than the U.S. in reducing inflation and unemployment since 1987.

Nor is Greenspan’s paradigm of risk management entirely new. Allan Meltzer, a Fed historian and economist at Carnegie Mellon University, has noted that in the mid 19th century, British economist Walter Bagehot said that during financial crises the Bank of England should suspend the gold standard and lend freely. “Call it risk management,” Meltzer says.

All that said, it seems unlikely the U.S.’ good economic performance can all be attributable to luck. There were many dark economic moments during Greenspan’s tenure and the fact that the U.S. came out of them as well as it did seems likely to have something to do with how Greenspan approached each challenge. It may be true that much of what Greenspan preaches now is standard operating practice for many central banks, but I think Greenspan has done a good job articulating it, and other central bankers have often said how much they’ve learned from watching Greenspan.

This brings me to the final question. Is Greenspan’s way of thinking learnable? Can it be passed on to future generations? And can Fed chairman Ben Bernanke emulate Greenspan and continue his track record?

One of the advantages of being a hedgehog is that your ideas tend to be memorable, even if they aren’t right. One of the disadvantages of being a fox is that it’s harder to sum up your way of thinking in a succinct, memorable way. A few years ago Harvard University economist Greg Mankiw dug up a scholarly paper Greenspan published in 1964 called “Liquidity as a determinant of industrial prices and interest rates.” Mankiw said he liked this paper because it foreshadowed many of the hallmarks of Greenspan’s Fed chairmanship: an intense look at the data, and a desire to “integrate various points of view” that showed “a lack of dogma and a nimbleness of mind.” But, Mankiw wondered if it presaged Greenspan’s career in another, less favorable way: it “left no legacy:” it had not been cited even once in subsequent academic literature.

Ben Bernanke, on the other hand, has produced lots of memorable, often-cited papers. For example, in 1983 he came up with what later became known as the “financial accelerator” theory of the Great Depression. According to this theory, deflation, debt and the banking system interacted in such a way as to dry up the supply of new loans to households and businesses. He argued this could explain, in a way that other theories could not, why the Depression was so long and deep. This was a rather bold statement from such a young academic. John Gunn, professor emeritus of economics here at Washington & Lee, just today reminded me how the Great Depression is a phenomenon not readily explained by a single, unitary hypothesis.

One is tempted to conclude from this that Bernanke is more hedgehog like, or at least less fox-like, than Greenspan. But is he? A little over a decade after these initial insights, Bernanke was crediting other scholars for advancing our understanding of the Great Depression by looking at how different countries performed depending on their adherence to the gold standard.

As you know, Bernanke advocates a public, numerical inflation target, which Greenspan did not. This is often cited as evidence that he is more attracted to rigid rules than Greenspan. But as ex-Fed Vice Chairman Alan Blinder, who was a colleague of Bernanke at Princeton University, noted a few weeks ago, “Like Judaism, inflation targeting comes in reform, conservative and orthodox variants.” Blinder said Bernanke once was somewhere between conservative and orthodox but “it's very clear while on the FOMC (from 2002 to 2005) he migrated to reform inflation targeting.” Bernanke’s time as governor on the policy-making Federal Open Market Committee exposed him to how the Fed’s dual mandate – to maintain both full employment and stable prices – and its committee-centered decision making would constrain the establishment and pursuit of an inflation target.

Finally, it’s worth noting that Bernanke says one of the most important lessons of the Great Depression is to be flexible and creative in the face of big challenges. When he delivered the H. Parker Willis Lecture in Economic Policy here at Washington and Lee University here almost two years ago, he concluded by saying:
“One lesson (of the Great Depression) is that ideas are critical. The gold standard orthodoxy, the adherence of some Federal Reserve policymakers to the liquidationist thesis, and the incorrect view that low nominal interest rates necessarily signaled monetary ease, all led policymakers astray, with disastrous consequences. We should not underestimate the need for careful research and analysis in guiding policy.”

On another occasion, he said of FDR’s response to the Depression:

“Roosevelt's policy actions were, I think, less important than his willingness to be aggressive and to experiment -- in short, to do whatever it took to get the country moving again. Many of his policies did not work as intended, but in the end FDR deserves great credit for having the courage to abandon failed paradigms and to do what needed to be done.”

I think these statements suggest that Bernanke is aware of the danger of hedgehog-like thinking.

I’ll conclude by suggesting we not be too quick to use someone’s early views as a template for what kind of central banker he will be. Consider the following quotes, from a much longer essay defending the gold standard published in 1966: “Gold and economic freedom are inseparable… Deficit spending is simply a scheme for the ‘hidden’ confiscation of wealth. Gold stands in the way of this insidious process. It stands as a protector of property rights.” Who would have guessed then that Alan Greenspan, the author of those worlds, would eventually become a central banker whose hallmark was his avoidance of rigid rules such as the gold standard?