The Natural Rate of Unemployment

The Natural Rate of Unemployment

On December 29, 1967, Milton Friedman stood before the American Economic Association to deliver his presidential address. The title was dry --- “The Role of Monetary Policy” --- but the argument was explosive. Friedman told the assembled economists that the central pillar of postwar macroeconomic policy was built on an illusion. Governments could not buy permanently lower unemployment by tolerating higher inflation. The economy had a “natural rate” of unemployment, and attempts to push below it would only produce accelerating prices.

It was one of the most consequential speeches in the history of economics.

The Argument

The Keynesian consensus of the 1960s held that policymakers faced a stable tradeoff between inflation and unemployment, embodied in the Phillips Curve. By running the economy a little hot --- a bit more government spending, a bit more money creation --- they could keep unemployment permanently below what it would otherwise be, at the cost of somewhat higher but stable inflation.

Friedman rejected this entirely. His argument ran as follows: suppose the government increases the money supply to stimulate demand. Firms see their revenues rising and hire more workers. Workers see higher nominal wages and are willing to work more. Unemployment falls. But here is the crucial step: the increase in money eventually pushes prices up. Workers gradually realize that their higher nominal wages have not actually increased their purchasing power --- real wages have not risen. They revise their expectations of inflation upward and demand higher nominal wages to compensate. Firms, facing higher costs with no real increase in demand, lay off the extra workers. Unemployment returns to its original level, but now with a higher rate of inflation baked into everyone’s expectations.

The “natural rate” is the unemployment rate at which inflation is stable --- neither accelerating nor decelerating. It is determined not by monetary policy but by the real structural features of the labor market: the efficiency of job matching, the generosity of unemployment benefits, the degree of unionization, regulations, demographics, skills mismatches, and the pace of economic change.

Phelps’s Parallel Discovery

Edmund Phelps, working independently at the University of Pennsylvania, reached essentially the same conclusion through a different analytical route. Phelps focused on the wage-setting behavior of individual firms and workers, showing that each firm sets wages based on what it expects other firms to pay. When expectations of inflation are incorporated into wage bargains, the short-run Phillips Curve shifts upward. Only by continuously surprising workers with more inflation than they expect can the government keep unemployment below its natural level --- an obviously unsustainable strategy.

Phelps’s work was more technical and less polemical than Friedman’s, but the conclusion was the same: the long-run Phillips Curve is vertical at the natural rate. Phelps was awarded the Nobel Prize in 2006, in significant part for this contribution.

From Natural Rate to NAIRU

The concept migrated into the policy mainstream as NAIRU --- the Non-Accelerating Inflation Rate of Unemployment. The name change was partly cosmetic, reflecting some economists’ discomfort with the word “natural” (which seemed to imply the rate was desirable or immutable). But NAIRU became the operational concept that central banks used to guide monetary policy.

The logic was straightforward: if unemployment is below NAIRU, inflation will tend to rise, and the central bank should tighten monetary policy. If unemployment is above NAIRU, inflation will tend to fall, and the bank can afford to ease. The art of central banking became, in large part, the art of estimating where NAIRU stood and calibrating policy accordingly.

How It Changed Policy

The natural rate hypothesis fundamentally altered the ambitions of macroeconomic policy. Before Friedman, policymakers believed they could choose a permanent point on the Phillips Curve --- perhaps 4 percent unemployment and 3 percent inflation. After Friedman, the consensus shifted: monetary policy could smooth short-run fluctuations, but it could not deliver a permanently lower unemployment rate. If governments wanted to reduce the natural rate itself, they needed to focus on structural reforms --- improving education and training, reforming labor market institutions, reducing barriers to job creation --- rather than turning the monetary dial.

This intellectual shift had enormous practical consequences. It contributed to the prioritization of inflation control over full employment in central bank mandates during the 1980s and 1990s. The independence of central banks, the adoption of inflation targeting, and the general retreat from activist demand management all owe something to the natural rate hypothesis.

The Empirical Problem: A Moving Target

The concept’s elegance is matched by its empirical slipperiness. The natural rate cannot be directly observed. It must be estimated from noisy data, and those estimates are subject to enormous uncertainty. Worse, the natural rate appears to move over time. In the early 1960s, it was plausibly around 5 percent in the United States. By the late 1990s, estimates had drifted down to around 4.5 percent. In the late 2010s, unemployment fell below 4 percent --- and then below 3.5 percent --- without triggering the inflation that NAIRU models predicted.

This raised uncomfortable questions. If the natural rate keeps changing, and we cannot pin it down in real time, how useful is it as a guide for policy? Central banks repeatedly found themselves estimating NAIRU after the fact, essentially fitting a curve to whatever unemployment rate happened to coincide with stable inflation. Critics charged that the concept was unfalsifiable: any unemployment rate could be declared “natural” in retrospect.

The Hysteresis Critique

A deeper challenge came from the concept of hysteresis --- the idea that prolonged periods of high unemployment can raise the natural rate itself. Workers who remain jobless for years lose skills, connections, and motivation. They drift out of the effective labor force, and the economy adjusts to their absence. Conversely, a period of low unemployment can pull people back in, lowering the natural rate.

If hysteresis is real, the Friedman framework breaks down in a fundamental way. Demand-side policy --- the very thing the natural rate hypothesis says cannot permanently reduce unemployment --- might actually do so by preventing the scarring effects of prolonged joblessness. Recessions are not just temporary dips below a fixed natural rate; they can permanently damage the economy’s capacity.

Why “Natural” Is Not Natural

Perhaps the most important critique is the simplest. The word “natural” implies something determined by nature --- permanent, optimal, beyond human influence. In reality, the natural rate is shaped by institutions, policies, and social choices that are anything but natural. Scandinavian countries with strong labor market institutions and active retraining programs have very different “natural” rates than countries with weaker safety nets. The natural rate is not a fact of nature but a product of the society that produces it.

Friedman’s insight that monetary policy alone cannot permanently reduce unemployment remains broadly accepted. But the broader implication --- that governments should focus narrowly on inflation and leave unemployment to structural forces --- is more contested than ever. The boundary between demand-side and supply-side, between cyclical and structural, turns out to be far blurrier than the clean vertical line of the long-run Phillips Curve suggests.