[Update: cited by  Megan McArdle in the Washington Post.]

Is it possible to buy goods and services that do not exist? This existential question was raised by the Fed’s actions March 3. Cutting their target interest rate by 50 basis points sends two messages. First, the Fed believes that current conditions are very contractionary. Second, the Fed is trying to promote confidence. This will not work.

But what about fiscal policy? The Wall Street Journal’s Chief Economics Commentator, Greg Ip, concurs that monetary policy will not work. But he then argues that fiscal policy can work because there’s plenty of headroom for the government to borrow.

But the real firepower lies with fiscal policy. Congress is working on a $7 billion to $8 billion package to combat the virus. The U.S. could easily borrow more: Congress authorized $65 billion in recovery spending after the 9/11 attacks, $100 billion after Hurricane Katrina in 2005, and $51 billion after Hurricane Sandy in 2012.

President Trump has suggested a payroll tax cut. On Monday, Sen. Elizabeth Warren proposed $400 billion of stimulus—roughly 2% of GDP.

This will not work. Both fiscal and monetary policy rely on the availability of goods and services for recipients of the funds to purchase. You cannot buy stuff that does not exist. The main result of expansionary policy, be it monetary or fiscal, will be inflation.

A Brief History of Macroeconomics

Not many people remember the phrase “fine tuning.”^[1] In the late 1960s some economists started using that phrase in the belief that fiscal and monetary policies could push the economy to any desired combination of inflation and unemployment. The Phillips Curve was the intellectual basis for this idea.^[2]

Figure 1 A Phillips Curve

The basis for the Phillips curve was empirical work done by A.W. Phillips relating the unemployment rate to the percentage change in wages. U.S. economists, led by Robert Gordon,^[3] noted that wages were a large part of the cost of production. Since prices reflect costs, they reasoned that they could replace the growth rate of wages with the inflation rate on the vertical axis. They were correct in a limited sense of the word.

Let’s look at the data. Here’s the standard Phillips Curve graph using data from the 1960s. The red line is the best-fit regression.^[4]

Figure 2 Phillips Curve, U.S., 1960s

Now let’s look at the same data for the 1970s. For obvious reasons I did not try to estimate this curve.

Figure 3 Phillips Curve, U.S., 1970s

Look closely at Figure 3. You can imagine two Phillips curves. The first is 1971, 1972, and 1973. The second is 1976 through 1979. It looks like the Phillips curve shifted up and to the right. Figure 4 shows this. What could have caused such a shift?

Figure 4 Two Phillips Curves in the 1970s

The 1970s were an era of economic upheaval. The OPEC cartel tripled the global price of oil not once, but twice. At one point there was a flat-out embargo on oil sales from OPEC member countries to the U.S. This was exacerbated by wage and price controls (President Nixon) and price controls on gasoline (President Carter). That decade led Paul Samuelson to coin the word “stagflation,” an era of high inflation and high unemployment. Arthur Okun codified this in his “misery index,” the sum of the inflation and unemployment rates.

The key factor that shifted the Phillips curve was high inflation. Inflation rates rose between 1971 and 1974. This inflation was unanticipated in that many people did not see it coming. When they adjusted their expectations in 1975, they overestimated future inflation. The new, higher level of inflation expectations define the second, higher, Phillips curve.

You had to live through that era to understand the impact on the economics profession. It happens that I was studying economics at Stanford at that time. I watched macroeconomists throw out their models. Some gave up. But others did the hard work that led to an understanding of the important role of expectations in market economies. An increase in expected inflation shifts the Phillips curve up and to the right. A decrease shifts the curve down and to the left.^[5]

A Simple Model

This material is straight out of baby macroeconomics. In one version or another it is included in almost all principles of macroeconomics textbooks. This model is the result of decades of research.

Let’s begin with some definitions. Aggregate demand (AD) is total planned spending in the economy. (There can be a difference between planned and actual spending that results in unexpected changes in business inventories.) Aggregate supply (AS) is total actual production in the economy holding unit production costs constant. Significant, economy-wide increases in cost shift the short-run aggregate supply curve (SRAC) up and to the left. In the 1970s the cost increase was the large runup in oil prices.^[6]

The position of the short-run aggregate supply curve is determined by per-unit production costs. But in the long run the economy tends to move toward full employment equilibrium. We model this by drawing the long-run aggregate supply curve (LRAC) as a vertical line. Figure 5 shows an economy in long-run equilibrium at full-employment GDP (Y_e).^[7]

Figure 5 The AD-AS Model

Alert observers will notice that the vertical axis is the average price level, not the inflation rate. Converting the model to use inflation instead of the price level is a pretty straightforward, but tedious, exercise. I’ll ask fearless readers who have made it this far to trust me on this.^[8]

Now suppose there is a negative (adverse) supply shock. Some factor may cause wage rates to rise across the economy. Or a major virus outbreak in China can disrupt supply chains. The decrease in quantities of intermediate goods causes prices to rise. Figure 6 shows the short-run equilibrium.

Figure 6 A Negative Supply Shock

Note something interesting about Figure 6.  The SRAS curve has shifted up because production costs have increaed.  Firms will take two actions to offset this cost increase.  They will lower output to Y₂. And they will raise prices to P₂. This the consequence of any adverse supply shock.

Now economic policymakers face a dilemma. The source of this dilemma today is the same as it was 50 years ago. Both fiscal and monetary policy have their most immediate impacts on aggregate demand. There may be some influences on aggregate supply but those take longer to appear.

Consider Figure 6 again. Suppose policymakers decide to fight unemployment. They do this by engaging in expansionary policies: lower interest rates, increase government spending, and/or cut taxes. That will shift the AD curve up and to the right as shown in Figure 7.

Figure 7 Policy to Fight Unemployment

You can see the problem. Increasing employment also causes the price level (inflation rate) to rise.

OK, suppose the folks working on policy decide to fight inflation. Figure 8 shows what happens then.

Figure 8 Policiy to Fight Inflation

Oops! GDP falls. Which means employment falls. By definition, unemployment rises.

Stagflation presents policymakers with real dilemmas. They can only solve one problem by making the other worse.

There is a good argument to be made that economic policy should do nothing when faced with an aggregate supply shock like this. The economy will eventually return to full employment on its own. The mechanism is that the real wage – the dollar wage adjusted for changes in the price level – must fall. This can be accomplished in two ways. First, nominal wages (the wage rate in dollars per hour) can fall. Remember, at Y₂ there is high unemployment. Eventually the excess labor supply will cause wages to fall.

The second way the real wage can fall is with a higher price level. This is the reason many central banks around the world set 2 percent inflation as a target. A bit of inflation erodes real wages just a little bit. But the economy will return to full employment faster with a touch of inflation.

Figure 9 shows the effect of a fall in the real wage rate.

Figure 9 Adjustjment to Long-Run Equilibrium

Conclusion

A highly infectious virus is causing a negative supply shock. This leaves policymakers in a quandary. They can fight inflation. But that makes unemployment worse. Or they can fight unemployment, making inflation worse. But one thing they cannot do is persuade people and businesses to buy goods and services that simply do not exist.

1. The phrase “fine tuning the economy” was first used in the 1960s. Tracking down the original source has been challenging. As far as I can tell, Gardner Ackley coined the phrase during his tenure as Chairman of the President’s Council of Economic Advisers. This is implied in his Wikipedia entry. See https://en.wikipedia.org/wiki/Gardner_Ackley. Accessed March 4, 2020. ↑
2. From EconomicsHelp.org https://www.economicshelp.org/blog/1364/economics/phillips-curve-explained/. Accessed March 4, 2020. ↑
3. There are a disturbingly large number of economists named Robert Gordon. This one was at Brookings in the 1960s. ↑
4. For those interested the regression used the logarithms of the actual inflation and unemployment rates. ↑
5. There is an enormous body of research on this topic. As an economist, it’s tempting for me to start slinging math around. Instead I’ll just ask anyone who’s interested to contact me via e-mail. ↑
6. Oil is used in more places than you can imagine. Most plastics start out as petroleum. Asphalt to pave streets, many solvents, some garden fertilizers and pesticides, and a host of other items use at least some oil. ↑
7. Employment changes roughly with GDP. In fact, in the short run using a model that says GDP equals employment times labor productivity works pretty well. ↑
8. Again, interested readers are more than welcome to make e-mail inquiries. ↑