I printed it out, and did some fiddling with the numbers. If I add the December numbers in the January report for "not in labor force" (75,631) and the number for "civilian labor force" (146,878), the total is 222,509, which matches the "as originally published" number in the Adjustments paper for "civilian noninstitutional population."

What does this mean? Beats the hell out of me, since the same two numbers in the Adjustments paper don't match the ones in the January report, even though they add up the same. And that's the BEST correlation I got.

I'm hoping somebody who knows how this works chimes in on this. Failing that, I suppose we can assume the February numbers will take all corrections into account.

Frankly, I think the BLS pulls a lot of this out of its ass. It's interesting to notice how many numbers for past months get drastically "revised" in later reports, although this rarely makes the news.