Lesson: Is a Recession Coming? Reading the Warning Lights

One class period on the three indicators people watch before a recession is called, what each one has gotten right and wrong, and what they say now. For teachers; the student handout is a separate page.

At a glance

Opening hook (5 minutes)

Ask: "How would you know if the country were in a recession right now?" Take answers. Many students will say "two quarters of falling GDP."

Then tell them: that isn't the official rule. A committee of economists at the National Bureau of Economic Research makes the call, usually months after a recession has started. In 2020 it dated the start to February and announced it in June. So people watch warning lights before the call. Today each group becomes the expert on one light.

Expert groups (20 minutes)

Split the class into three groups (or six, two per light). Each group opens its chart, reads its light's paragraph in What a Recession Is, and fills in its box on the handout: what the light measures, the signal, how many recessions it caught on the chart, its false alarms, and what it says now.

Tip for every group: the gray bands on each chart are recessions. Count a hit when the signal comes on within about two years before a gray band starts, or a few months after; count a false alarm when no band follows. The yield curve warns earliest, so its hits sit furthest ahead of the bands.

Report back (15 minutes)

Each group gets three minutes: what the light measures, its track record, and whether it's lit now. Build a table on the board as they go. Then ask the class: "Put the three together. Is a recession coming?"

Discussion questions (10 minutes)

  1. Reading it. Why does the official call come so late?
  2. Each light measures something different. Name what each one watches: jobs, layoffs, or what?
  3. Placing it. The yield curve stayed inverted from late 2022 to late 2024 and no recession followed. Does that make it useless? What might explain it?
  4. Why do lights based on the job market tend to come on at the start of a recession rather than long before it?
  5. Judging it. A news site writes "Recession signal flashes red" the week one light comes on. What would you want to know before believing it?
  6. If you had to watch just one light, which would it be, and what would you give up?

Answer key

Group boxes (from What a Recession Is):

Sahm ruleJobless claimsYield curve
MeasuresThe rise in unemployment's three-month average over its low of the past yearNew filings for unemployment benefits, four-week average10-year Treasury yield minus the 3-month yield
Signal0.5 or moreA rise of about 50,000 from its low of the past yearBelow zero (inverted)
RecordCrossed within four months of the start of every recession since 1970Within about a year of the start of all eight recessions since 1967Inverted 11 to 23 months before each of the four recessions since 1982
False alarmsLate 1976 (one month) and 2024About ten, including after Hurricane Katrina (2005) and in mid-2023Five days in 1998, then the 2022–24 inversion with no recession
Now0.00 (October 2026)about 198,000 (the week ending November 14, 2026)0.90 points (October 2026)
  1. The committee waits for data that come out late and get revised, and it wants to be sure a downturn is broad and lasting before naming it. Being right matters more to it than being fast.
  2. Sahm: the unemployment rate. Claims: layoffs, weekly. Yield curve: what bond buyers expect the Fed to do with interest rates, which reflects what they expect of the economy.
  3. Not useless: before 2022 it came on ahead of every recession since 1982. The page's argument section gives the explanations on offer: the pandemic scrambled indicators built on older patterns; companies that saw the warning trimmed hiring and investment early and softened the downturn (Campbell Harvey's view); or the Fed's rate cuts in late 2024 softened it. The honest answer is that the next recession will be the test.
  4. Unemployment and layoffs are part of a recession itself. They rise as it starts, so they confirm more than forecast.
  5. Which light, how far past its signal, for how long, whether the data will be revised, and what the other lights say.
  6. Any answer that names its trade-off works: the curve warns earliest, but its big miss was 2022–24; the Sahm rule comes late but has few false alarms; claims are weekly and fast but the noisiest, with about ten false alarms, and they miss hiring freezes.

Sources

The weekly digest

One email on Monday mornings: the week's new articles and charts, the numbers that moved most, and what comes out in the week ahead. No ads, and no tracking of opens or clicks.

Past issuesHow addresses are kept