How would you know if a recession is just collective laziness?
Our results suggest that in a recession Canadians sleep an average of 2 hours and 34 minutes more per week, or 22 minutes more per day.
Well … yeah … obviously.
The researchers who did this report are not looking at the collective laziness question. They’re just interested in the sociological aspects of a recession: sleep is related to performance and mortality.
But, let me stretch your brains a bit. I’m not claiming that recessions are collective laziness. But, this is a claim I would have laughed at 20 years ago, and now I’m not so sure.
The trick is to convert per capitas to other aggregates. Let’s do the math. They find that the average person sleeps 22 more minutes per day during a recession. That’s an average over a week, so 154 minutes per week. If you work a 40 hour work week, that is 2,400 minutes. Now, if the 154 minutes of sleep come out of what would otherwise be worktime, that’s like you being unemployed for about 6.4% of your time that you weren’t unemployed for before. That 6.4% is in the range of the increase in unemployment in Canada (where the study was done) during this recession – it’s actually quite a bit on the high side since Canadian unemployment only went up by about 2.5%, indicating that increased sleep more than accounts for the increased unemployment.
This is what in econometrics would be called a reduced form explanation. It tells us about associations, but it isn’t a theory of why this happens. That would be called a structural explanation. The problem is that reduced form explanations are usually consistent with more than one structural explanation. Here’s two: 1) a collapse in confidence causes a recession which gives people more time to sleep, or 2) people are tired and need more sleep, so they do less work, and we measure that as a recession. Separating out those structural explanations requires a deeper model, along with all the inherent biases that the researcher builds into it.