The Phelps Factor

Joseph E. Stiglitz

(NOBLE PRICE OF ECONOMICS)

On December 10, Edmund Phelps, my colleague at Columbia University, will receive the Nobel Prize in economics for 2006. The award was long overdue. While the Nobel Prize committee cited his contributions to macroeconomics, Phelps has made contributions in many areas, including the theory of growth and technological change, optimal taxation, and social justice.

Phelps’ key observation in macroeconomics was that the relationship between inflation and unemployment is affected by expectations, and since expectations themselves are endogenous – they change over time – so, too, will the relationship between unemployment and inflation. If a government attempts to push the unemployment rate too low, inflation will increase, and so, too, will inflationary expectations.

This insight holds two possible policy implications. Some policymakers have concluded from Phelps’ analysis that the unemployment rate cannot be lowered permanently without ever-increasing levels of inflation. Thus, monetary authorities should simply focus on price stability by targeting the rate of unemployment at which inflation does not increase, referred to as the “non-accelerating inflation rate of unemployment” (NAIRU).

But the NAIRU is not immutable. The correct implication, which Phelps repeatedly emphasized, is that governments can implement a variety of policies, particularly structural policies, to allow the economy to operate at a lower level of unemployment.

Policies that focus exclusively on inflation are misguided for several other reasons. As a practical matter, even controlling for expectations (as Phelps’ work insists that we do), the relationship between unemployment and inflation is highly unstable. It is virtually impossible to discern the relationship from the data except in a few isolated periods.

Changes in education levels, unionization, and productivity are part of the explanation for this instability. But, whatever the reason, policymakers face considerable uncertainty about the level of NAIRU. Thus, they still face a trade-off between pushing unemployment too low, and setting off an episode of inflation, and not pushing hard enough, resulting in an unnecessary waste of economic resources.

How one views these risks depends on the costs of undoing mistakes, which in turn depends on other properties of the inflation-unemployment relationship that Phelps’ analysis did not address. The weight of evidence indicates that the cost of undoing the mistake of pushing unemployment down too far is itself very low, at least for countries like the US, where the relationship has been carefully studied. In this view, the Federal Reserve should aggressively pursue low unemployment, until it is shown that inflation is rising.

By contrast, inflation “hawks” argue that inflation must be attacked preemptively. While most central banks are inflation hawks, this stance is a matter of religion, not economic science. There is simply little or no empirical evidence that inflation, at the low to moderate rates that have prevailed in recent decades, has any significant harmful real effects on output, employment, growth, or the distribution of income. Nor is there evidence that inflation, should it increase slightly, cannot be reversed at a relatively minor cost – comparable to the benefits of additional employment and growth enjoyed in the excessive expansion of the economy that led to the increase in inflation.

In the early 1990’s, the Fed, and many others, thought that the NAIRU was around 6%-6.2%. Based on changes in the economy, I and the staff that worked with me on President Bill Clinton’s Council of Economic Advisers argued that the NAIRU was considerably lower. We were right. Unemployment fell to 3.8% without any surge in inflation.

This matters because, as the great economist Arthur Okun argued, reducing unemployment by two percentage points would increase output by 2%-6%, or $0.5-1.5 trillion dollars in the case of America. Even for a rich country, that is a lot of money. It could be used to put America’s social security system on a stable footing for the next 75-100 years. It could even pay for a substantial share of the cost for a war like that in Iraq!

Phelps’ work helped us to understand the complexity of the relationship between inflation and unemployment, and the important role that expectations can play in that relationship. But it is a misuse of that analysis to conclude that nothing can be done about unemployment, or that monetary authorities should focus exclusively on inflation.

That view belongs to a school of modern macroeconomics that assumes rational expectations and perfectly functioning markets. In other words, individuals – usually assumed to be identical – fully use all available information to forecast the future in an environment of perfect competition, no capital market shortcomings, and full insurance of all risks. Not only are these assumptions absurd, but so are the conclusions: there is no involuntary unemployment, markets are fully efficient, and redistribution has no real consequence. But, while government policies, according to this school, are ineffective, that matters little. Because markets are always efficient, there is no need for government intervention. More perniciously, many supporters of this view, when confronted with the reality of unemployment, argue that it arises only because of government-imposed rigidities and trade unions. In their “ideal” world without either, there would, they claim, be no unemployment.

For more than three decades, Phelps has shown that there is an alternative approach. He has tried to understand what we can do to lower unemployment and increase the well-being of those at the bottom. But he has also striven to understand what makes capitalist economies dynamic, what lies behind the entrepreneurial spirit, and what we can do to promote it further. Phelps’ economics remains one of action, not resignation.

Joseph Stiglitz is a Nobel laureate in economics. His latest book is Making Globalization Work.