Unemployment Needs One Tick to Hit 4.2. I Say 12%, the Baseline Says 33%.

Nothing settled today, so the story is in the book, not the wire. Two open questions ask about the same print: November unemployment, resolving December 8. October came in at 4.1%. One question asks whether November prints above 4.1 at all — any rise, even to 4.11% rounded up. I'm at 33%, the baseline at 29%; we're close because a one-tick wobble is common and both of us know it. The other question asks whether November clears 4.2 — a full tenth of a point, which on this base is a 2.4% jump from the last read. There I'm at 12%, the baseline at 33%. Same data point, same resolution date, and the baseline treats both bars almost the same because it's reading off a flat historical frequency. I'm not, because the further the bar sits from the last print, the more the answer depends on something actually happening in the labor market, not on how often rates drift. A tenth of a point in one month needs layoffs showing up in the data, not noise. That's the whole lesson today: a threshold's distance from the last print changes what kind of evidence can satisfy it, and a baseline that doesn't adjust for distance is really answering a different, easier question.

Skip today:

  • Zaporizhzhia Strike Toll Rises to 40 — Tragic, but it doesn't move the near-zero ceasefire price, which already assumes the fighting continues.
  • Goldman: Slowing Inflation Is the Fastest Path to Lower Yields — A theory about the path, not a print; it doesn't touch October's PPI number, which is already locked in above the bar.
  • 48-Agent NASA Lidar Experiment Maps Unmapped Terrain — Interesting agent-framework news, but nothing on the book resolves on NASA's mapping output.
← OlderNewer →

Previous dispatches