Why Every Recession Prediction Has Been Wrong (So Far)
In 2022, economists predicted a recession by end of 2023. In 2023, they moved the forecast to 2024. In 2024, the consensus said Q1 2025. We're now in early 2026, and the US economy is still growing, employment is near historic highs, and consumer spending hasn't collapsed. What keeps getting it wrong?
The inverted yield curve problem
The most cited recession predictor for the past three years has been the inverted yield curve, when short-term Treasury yields exceed long-term ones. It predicted every recession since 1970 with no false positives. Then it inverted in 2022, stayed inverted for the longest stretch on record, and nothing happened.
This doesn't mean the indicator is broken. It means the lag is longer and more variable than the historical average suggested. It also means that when you have 40 years of data and one extreme outlier (a global pandemic followed by an enormous fiscal stimulus), your models may not be telling you what you think they're telling you.
The labor market that refuses to break
Every recession model has some version of unemployment as a central input. The theory: when rates rise, companies stop hiring. When they stop hiring, consumers cut spending. When consumers cut spending, the economy contracts.
What the models didn't account for was a labor market permanently altered by COVID. The labor force participation rate for prime-age workers is at a 20-year high. Companies that went through the brutal experience of hiring in 2021-2022 have been extremely reluctant to lay off workers, because they know how hard it is to get them back. This 'labor hoarding' behavior has kept unemployment low in conditions that would historically have caused it to rise significantly.
The fiscal elephant in the room
Government spending has been quietly doing a lot of work. The CHIPS Act, the Inflation Reduction Act, and ongoing defense spending have injected hundreds of billions of dollars into specific sectors, including semiconductors, clean energy, and defense manufacturing, all running at full capacity while other parts of the economy slow.
This is unusual. In previous rate-hiking cycles, fiscal policy wasn't actively fighting monetary policy. The Fed raised rates and the government wasn't simultaneously spending at wartime levels. The result is an economy that's slowing but not collapsing, which makes every recession call look wrong six months out.
So when does it happen?
Maybe it doesn't, not in the traditional sense. What's increasingly likely is a rolling slowdown that hits different sectors at different times rather than a synchronized contraction that shows up cleanly in the GDP numbers. Real estate is already in recession. Trucking went through one last year. Consumer discretionary spending is weak. But tech employment recovered, manufacturing is strong, and healthcare never slows down.
The forecasters who've been calling recession keep being right about the weakness and wrong about the timing. That's not a forecasting failure. It's a sign that the economy is more complex and more segmented than the models assume.