[dragonwriter]

> This is more like "how will we know when a recession has already started without waiting for a down GDP for two quarters." That's 6 months in.

It's at least 6 months in: NBER recession dating even for recessions that include at least two down quarters (which, while a popular conception is not the official definition, though I don't think any two-quarter consecutive downturn has failed to be included within a declared recession) is often before the first down quarter in a consecutive pair. (E.g., the Great Recession is timed from the Q4 2007 peak, even though there was only one quarter of decline immediately following, then a dead-cat bounce quarter, then a four quarter consecutive decline. If you looked for a pair of down quarters, you'd detect the Great Recession a year after it started; 4-5 months moves that up a lot.)