Slaving away on my own personal make-work program, Kieran gets in on the fun we’re having with 1930s employment data.

Basically, the dispute is over how you compose your unemployment rate, and we could argue over that endlessly. Suppose we don’t? Suppose instead we discompose it, and see what’s going on with employment? Is the improvement in employment under the New Deal solely a product of “make-work”?


And combining nonfarm & farm employment isn’t a trick; in fact, farm employment stays nearly level but declines slightly.

Thanks, Kieran. Also, I clearly need to learn R.