The graph doesn't look like pregnancy rates precede recessions at all, it looks like they occur at the same time (and thus aren't good for prediction). What's

The graph doesn't look like pregnancy rates precede recessions at all, it looks like they occur at the same time (and thus aren't good for prediction). What's going on is that we have a goofy rule for recessions that require continuous retraction for a recession to be official. In most of these recessions it looks like the economy peaked the same time as pregnancy rates, but a tiny upward blip two quarters later prevented us from being officially in a recession. If you smoothed the idea of a recession into something like "contraction in 3 out of 4 quarters" you wouldn't have tiny blips delaying the official recession, and you wouldn't see pregnancy rates preceding anything, they would just be following the rest of the economy.