"I'm a Fronkensteen"

There is a movie, "Bad Day at Black Rock", and any objective economist (there aren't any, and that includes Your Humble Servent) must think that today is such a day. The Monthly Unemployment report has blown the lid off of the data. How could a lackluster increase in employment yield a 4 tenths of a percentage point *decrease* in the unemployment rate? Yeah, Dude, what's up.

I'm here to reveal the awful truth.

First, contrary to widespread belief, you're not counted as out of the workforce when your UI runs out. And you're not counted as newly hired when you get a job. All the numbers you're hearing today, and likely for some time, are generated by *surveys*; two in the case of the monthly report. This is one. This is the other. And here is a description of sausage making. The last is a rather long, and detailed, report. I'd wager that few economists, data scientists, and certainly not reporters, have ever read one front to back; I did, but that's when I was in grad school, so it shouldn't count.

These, to be technical, are stratified random samples. Just as Gallup and Harris tell you what the most watched TV show is with a handful (relative to the nation's population) of viewers, the BLS and Census tell you what's going on in the economy with a handful. Don't get fidgety, though. The private sector prognosticators do the same thing. The country is too big, both in bodies and square miles, to do a census each month on citizens and businesses.

But, I will admit, that some news spewers have made a point of the 315,000 drop in the labour force to explain how it is that so-so new jobs leads to .4 point drop in the unemployment rate. Same thing happened during Reagan, but no one's made that point.