Opening
HOST
Welcome back. Today we are on the short run: the booms and slumps of an economy, and what they do to prices and jobs. Let's start with one month.
ECONOMIST
April 2020. Covid-19 had just closed shops, restaurants and offices. In that single month, the United States lost 13.6 percent of all its jobs. That's more than one job in eight.
HOST
And the growth models we have studied say nothing about that?
ECONOMIST
Not much. They tell us where an economy is heading over decades. Today we look at the ups and downs along the way.
HOST
So here is today's question. Why does output swing around its trend, and what do booms and slumps do to inflation and jobs?
ECONOMIST
By the end of this episode, you'll be able to do three things. First, describe how economists measure booms and slumps, and what U.S. recessions did to output, jobs and inflation, the rate at which prices rise. Second, explain why a boom that lasts makes inflation rise year after year, and why bringing it down takes a slump. Third, use a rule called Okun's law to work out how much unemployment a slump brings.
HOST
And we'll follow the chapter's mind map part by part?
ECONOMIST
Yes. Part one measures the cycle, meaning the repeated ups and downs of booms and slumps. Part two is the Phillips curve, and part three is Okun's law and policy. Each part has five leaves, small topics that we take in order.
Measuring the cycle
HOST
Part one of the map: Measuring the cycle. The first leaf is short-run output, the percent gap from potential. So, what is potential?
ECONOMIST
Let's visit Islandia again, our small imaginary island country. Over many years, its output climbs along a smooth path, set by its tools, its workers and its ideas. That smooth path is the trend in today's question. Economists call it potential output. It is what Islandia produces when everything is used at a pace it can keep up year after year. Think of a car's cruising speed, the speed it can hold on a long trip, not its top speed.
HOST
And actual output doesn't follow that path?
ECONOMIST
Not year to year. Shocks, meaning sudden events, push output away from potential. On Islandia, that could be a jump in the price of imported fuel, a storm, or a slump in Mainland, its trading partner. In a boom, output is above potential. Harbor Boatworks, the island's boat builder, runs overtime shifts and delays repairs. Its boats get hard to buy. In a slump, output is below potential, and work at its boatyards slows down.
HOST
So how do we measure how far Islandia is from potential?
ECONOMIST
Take the distance between actual and potential output, and divide it by potential output. That gives short-run output, also called the output gap. Suppose potential output is 100 units, and Islandia makes only 98. It falls short by 2 percent of potential, so short-run output is minus 2 percent. It is positive in a boom, negative in a slump, and zero at potential. We use a percent because the same shortfall of boats is a disaster for a tiny island, but almost nothing for a giant economy.
HOST
Next leaf: 1933 and 1944. What happened in those years?
ECONOMIST
These are the biggest swings in the U.S. data. In 1933, during the Great Depression, a deep slump that hit the whole world, U.S. output was 23 percent below potential.
HOST
Almost a quarter below. And 1944 was World War Two. I suppose output was right at potential then, since potential is the most an economy can make.
ECONOMIST
That's a common misreading of the word. In 1944, output was 35 percent above potential, as factories ran day and night to make weapons. Remember, potential is a cruising speed, not a top speed. For a while, firms can run extra shifts and skip repairs, but they can't keep that up. One caution: no agency estimated potential output for those years, so these early gaps are based on a smooth trend drawn through the data. Since about 1950, the swings have been much smaller, mostly within 4 percent of potential, above or below.
HOST
Next leaf: 2020, in today's data and in the 2023 data. Before the numbers, what exactly is a recession?
ECONOMIST
A recession begins when output falls below potential, so when short-run output turns negative. In recent decades, the deepest dip came in the spring of 2020, when Covid-19 closed much of the economy.
HOST
How deep was it?
ECONOMIST
In today's data, output fell 8.8 percent below potential. But in the data published in 2023, the same dip was 10.4 percent.
HOST
Wait, the past changed?
ECONOMIST
The past didn't change, but what we know about it did. Statistical offices revise GDP, their measure of total output. That means they correct and update it as better information arrives. And potential output is never observed. It is estimated with a model, and every new estimate moves the gap. So when you quote a gap, say in which year the data were published.
HOST
Next leaf: the Covid-19 recession and the jobs it cost. How many?
ECONOMIST
21.9 million jobs, in about two months. That's the total for the whole United States, from the peak in February to the low point in April.
HOST
Is that a lot, compared with other recessions?
ECONOMIST
Far more. In most U.S. recessions since 1980, the country lost between 1 and 3 million jobs. So this one cost more than seven times as many.
HOST
Last leaf of this part: inflation. What happens to it in a recession? I'd guess it goes up. Times are hard, and everything feels expensive.
ECONOMIST
The data say the opposite. Since 1960, inflation has usually peaked close to the start of a recession, and then fallen during it. That's this leaf: inflation usually falls in a recession. The reason is weak demand. In a slump, firms have machines standing still and few customers. To sell more, they hold back their price rises, and inflation comes down. Careful with the words, though. Falling inflation does not mean falling prices. Prices usually still rise, just more slowly.
HOST
So in part one we saw how to measure the cycle: as a percent gap from potential output. Recessions push that gap below zero, they cost millions of jobs, and inflation usually falls during them.
ECONOMIST
Yes. Next part of the map: why a slump slows inflation, and what a boom does to it.
The Phillips curve
HOST
Part two of the map: The Phillips curve. The first leaf is its equation. What does it say in words?
ECONOMIST
The curve is named after the economist A.W. Phillips. Its equation says that the change in inflation from last year to this year equals a slope times short-run output. The slope is a number that says how strongly inflation reacts. So when output is above potential, inflation rises. Below potential, it falls. At potential, it stays where it is.
HOST
Why would a boom push inflation up?
ECONOMIST
Start with what people expect. Suppose prices on Islandia have been rising 2 percent a year, so everyone expects about the same again. Now a boom comes. Harbor Boatworks pays overtime, and its boats are hard to buy. So it raises its prices by about 3 percent. When every firm does that, inflation itself climbs to about 3 percent.
HOST
And in a slump?
ECONOMIST
Firms raise their prices by less than everyone expects, so inflation falls. And at potential, firms simply keep the usual rate. So inflation stays at whatever level it has, whether that's 2 percent or 10.
HOST
Next leaf: the U.S. slope, 0.35, and the textbook's value of about one third. How steep is the curve in real data?
ECONOMIST
Think of each U.S. year as one dot on a chart, placed by its output gap and by its change in inflation. Years with output above potential tend to have rising inflation. The best straight line through the dots has a slope of 0.35, close to the usual estimate of about one third. So if output stays 3 percent above potential for a whole year, inflation rises by about 1 percentage point. The dots form a loose cloud, though. Big jumps in oil prices pushed some years far above the line, because output is not the only thing that moves inflation.
HOST
Percentage point? Not percent?
ECONOMIST
A percentage point is the plain difference between two percentages. If inflation goes from 4 to 6 percent, it has risen by 2 points. In percent, it would be a rise of 50 percent.
HOST
Next leaf: output held 3 percent above potential. So policymakers keep output high on purpose?
ECONOMIST
It's a thought experiment. The economy starts at potential, with inflation at 2 percent. Then policymakers hold output 3 percent above potential, every single year. To keep the numbers simple, we use a slope of 1. For Islandia, we use the textbook's one third, 0.33, so this one is about three times as steep.
HOST
So inflation rises by 3 points in the first year, to 5 percent. But then surely it settles there, at a new, higher rate? A one-time cost for a stronger economy.
ECONOMIST
Not while the boom goes on. Next year, 5 percent is the rate everyone expects. With output still above potential, firms go beyond it again. So inflation rises by another 3 points, to 8 percent. In the third year it reaches 11 percent. It keeps climbing for as long as the boom lasts.
HOST
So inflation is like a running total.
ECONOMIST
Exactly. A running total is a sum you keep adding to. Every year, the gap adds to inflation in a boom, or takes away from it in a slump. So a boom that has ended still leaves inflation higher. With Islandia's flatter slope, the climb would be about 1 point a year instead of 3. Slower, but still without end.
HOST
Next leaf: cutting inflation by 1 point. If a boom pushes inflation up, a slump has to bring it down. How big a slump?
ECONOMIST
Let's build the number step by step for Islandia. Its slope is 0.33. So each year that output is 1 percent below potential, inflation falls by 0.33 points.
HOST
And we want it to fall by a whole point. So we need about three of those years.
ECONOMIST
Right. To be exact, divide 1 point by 0.33, which gives about 3.03.
HOST
3.03 of what, though?
ECONOMIST
Of percent-years. One percent-year means output 1 percent below potential for one year. So cutting inflation by 1 point costs about 3.03 percent-years of lost output.
HOST
How could Islandia pay that?
ECONOMIST
With one sharp year, about 3 percent below potential. Or with three milder years, each about 1 percent below.
HOST
Then the gentle path is the cheaper way. Less pain in total.
ECONOMIST
Not in this model, and here's why. Each year, inflation falls by the slope times that year's gap. So only the sum of the gaps counts, not how you split it. Both paths add up to the same 3.03 percent-years. Policy chooses how to spread the pain, not how much pain there is. And a flatter curve means a bigger total slump, because each year of slump brings inflation down by less.
HOST
Last leaf of this part: Volcker. Who was he?
ECONOMIST
In 1979, the Federal Reserve was led by Paul Volcker. That's the U.S. central bank, which sets interest rates. Inflation was above 11 percent, so the Federal Reserve began to raise interest rates sharply. Inflation still peaked at 13.5 percent in 1980.
HOST
And then?
ECONOMIST
A deep slump. Borrowing became expensive, so firms and families spent less. In 1982, output was about 6 percent below potential. And inflation fell to about 3.2 percent by 1983. That's the direction the Phillips curve predicts. But it fell faster than the curve predicts. With Islandia's slope, the curve predicts less than half of the actual fall. So other forces helped too, such as falling oil prices.
HOST
But people lost their jobs. Was it worth it?
ECONOMIST
That is the hard policy question for any central bank: should it cause a recession to end high inflation? That slump cost almost 3 million jobs. The model says a slump is the only way down. The debate is about how fast to go, and about who pays: mostly the people who lose their jobs.
HOST
So in part two we saw that the output gap moves the change in inflation, not its level. A boom that lasts makes inflation climb every year. And bringing it down takes a slump. The slope decides how big that slump is in total.
ECONOMIST
Right. Next part of the map: what the gap means for jobs, and for the people who set policy.
Okun's law and policy
HOST
Part three of the map: Okun's law and policy. The first leaf is the equation of Okun's law. What does it say?
ECONOMIST
It links output to unemployment, and it's named after the economist Arthur Okun. First, the natural rate of unemployment is the rate an economy settles at in the long run. It's the job market's version of potential output. Now suppose Islandia's output falls 2 percent below potential. How far above its natural rate does unemployment go?
HOST
If firms make 2 percent less, they need 2 percent fewer workers. So I'd say 2 points.
ECONOMIST
Only 1 point. Okun's law says that for each percent that output is below potential, unemployment sits half a point above its natural rate. In a boom, it works the other way round. Why only half? Firms adjust in other ways first. Harbor Boatworks cuts hours, and lets some workers go, but keeps others on, ready for the recovery. And some islanders who lose their jobs stop looking for work. Only people who are looking for work count as unemployed.
HOST
Next leaf: the U.S. slope, minus 0.62, a tight cloud. Tight compared with what?
ECONOMIST
Compared with the loose Phillips cloud. Here each dot is a U.S. year since 1960, placed by its output gap and by its unemployment minus the natural rate. The dots sit close to a falling line, with a slope of minus 0.62. That's a bit more than half a point for each percent. One big exception is 2020. Unemployment jumped far more than the fall in output would predict, as businesses closed and let workers go.
HOST
Next leaf: growth below potential growth. Now, if output is growing at all, firms are making more. So unemployment must be falling, right?
ECONOMIST
Let's test that with numbers. On Islandia, potential output grows 2 percent a year. Suppose Islandia starts at potential, and this year its output grows only 1 percent.
HOST
Then it falls behind its potential by about 1 percent.
ECONOMIST
Exactly. Short-run output is now about minus 1 percent. Okun's law turns that into unemployment: half a point for each percent. So unemployment rises by about half a point, even though output grew. A lot of students think growing output means falling unemployment. But output must grow at least as fast as potential. That's also why a recession does not need falling output. A slump is about the gap, not about whether output shrinks.
HOST
Next leaf: the chain from the interest rate to output, and on to inflation and unemployment. So far, policymakers simply picked the gap. How do they really move it?
ECONOMIST
That's the chain the next chapters build. Suppose the Bank of Islandia, the island's central bank, raises its interest rate. Borrowing gets more expensive. Harbor Boatworks delays its investment, and islanders buy fewer boats on credit. Spending falls, so output drops below potential. Then the Phillips curve takes over: inflation falls. And Okun's law takes over too: unemployment rises.
HOST
Last leaf: potential output is estimated, not observed. Why is that a problem?
ECONOMIST
Because the whole chain runs on the gap, and the gap needs potential output, which nobody can see. One explanation of the 1970s in the United States goes like this. Policymakers thought potential output was higher than it really was. So actual output seemed to be below potential, and extra spending looked right.
HOST
But if potential was really lower, the economy was running above it.
ECONOMIST
In that explanation, yes. Inflation then climbed, as in our boom experiment, and oil price jumps added to it. So what should a central bank do? This is debated. Some say: use the best estimate of the gap anyway, because inflation responds to it. Others say: pay more attention to what we can measure, like inflation and unemployment. Okun's law also works backward. If unemployment is 1 point above its natural rate, output is about 2 percent below potential. But the natural rate is also an estimate, so this check is not perfect either.
HOST
So in part three we saw that unemployment follows the gap, about half a point for each percent. It rises whenever output grows more slowly than potential. And policy must be guided by a gap that nobody can see directly.
ECONOMIST
Exactly. Let's sum up.
Wrap-up
HOST
Let's go back to our three goals. First, how do we measure booms and slumps?
ECONOMIST
With short-run output, the percent gap from potential. In a U.S. recession, it falls below zero, millions lose their jobs, and inflation usually falls.
HOST
Second, why does a boom that lasts make inflation rise year after year?
ECONOMIST
Because of what people expect. Take our thought experiment. Everyone expects 2 percent, and the boom pushes inflation to 5. Then 5 becomes the expected rate, so with the boom still going, it climbs to 8. To bring it down, you need a slump, so that firms raise prices by less than expected.
HOST
And third, jobs.
ECONOMIST
Each percent of output below potential adds about half a point of unemployment. So here is the big message. Output swings around potential. A boom pushes unemployment down and makes inflation rise year after year. So bringing inflation down takes a slump, and the slope of the Phillips curve sets the total size of that slump.
HOST
Here's a question to think about. Islandia's Treasury, the government's finance ministry, spends more and pushes output 2 percent above potential for two years, just before an election. After the vote, the Bank of Islandia wants inflation back where it started. How big must the slump be, in total? And would a steeper Phillips curve make that slump smaller?
ECONOMIST
Pause the episode and think about it.
HOST
Okay, we're back. What's the answer?
ECONOMIST
Inflation is a running total of past gaps, so to end where it started, the gaps must add up to zero. The boom was 2 percent above potential for two years, so 4 percent-years. The slump must add up to 4 percent-years too, for example 1 percent below potential for four years. And no, a steeper curve does not help.
HOST
Let me check with a slope of one half. The boom adds 1 point a year, so 2 points in all. Each year at 1 percent below potential takes off half a point. So four years again. But didn't we say the slope sets the size of the slump?
ECONOMIST
It does, when the inflation is already there and we must remove it, like the high inflation Volcker faced. But here, our own boom made the inflation. A steeper curve makes more inflation in the boom, but also removes it faster in the slump. So in the end, the slope does not change the total size of the slump.
HOST
Got it. And where can students try all this?
ECONOMIST
Open the chapter's Colab notebook. The Model Apps let you move the slope of the Phillips curve without any code, and the Data Apps show the data for many countries.
HOST
And next time?
ECONOMIST
In Chapter 10, we use these tools on two huge shocks. We ask how a fall in house prices pushed banks toward failure, and how the pandemic slump was different.
HOST
That's it for today. Thanks for listening, and see you next time.