The Sahm rule fires when the three-month average unemployment rate rises 0.5 percentage points above its trailing 12-month low. It triggered in July 2024 at 0.53 points above the low. Every prior trigger going back to 1970 was followed by a recession within months. This one was not. Real GDP grew 2.5% in 2024 and is tracking near 2.0% for 2026.
The rule was built on an implicit assumption: unemployment rises because demand collapses. Firms stop hiring, then lay off, then households spend less, and the feedback loop tips into recession. Under those mechanics, a 0.5 point rise off the cycle low is a tripwire because it can only happen if demand is breaking.
What happened in 2023 to 2025 was different. Foreign-born labor force participation rose by roughly 3 million workers between early 2023 and mid-2025, largely driven by post-pandemic immigration flows that the CBO estimates added about 1% to the labor force annually for three years. Unemployment rose because the denominator grew faster than firms could absorb new entrants, not because firms shed existing workers. Layoffs (measured by initial jobless claims and JOLTS separations) stayed near multi-decade lows the entire time.
This is a supply-side normalization. The vacancy-to-unemployed ratio fell from a peak of 2.0 in early 2022 to 1.1 by late 2024, almost entirely through the unemployed numerator rising, not through vacancies collapsing. Wage growth decelerated from 5.8% to 3.8% over the same window, consistent with a market clearing rather than a market breaking.
The practical implication: indicators built on the assumption that labor market loosening signals demand destruction are, for now, miscalibrated. The Sahm rule, the Conference Board LEI, the yield curve inversion signal all carry the same embedded model. None of them anticipated supply-driven softening because the post-1970 sample contains no example of it. The Fed appears to read it correctly, which is part of why policy held steady through the 2024 unemployment uptick rather than cutting in panic.
What still works: layoff data, real consumer spending, and the credit channel. Initial claims have been below 240,000 for most of 2026. Real personal consumption is growing 2.2% year-over-year. Bank lending standards loosened modestly in the April senior loan officer survey. None of these are flashing recession. The labor market got softer through more workers showing up, not through fewer jobs being available.