No other country has been this far into the pandemic and still had the number of cases growing at the rates the U.S. is seeing.
(ft.com/content/a26fbf…)
(ft.com/content/a26fbf…)
Lots of people have been asking me what this would look like on a per capita. Short answer: If you care about the slopes -- that is, the trajectory -- pretty darn similar.
Longer answer: Let's learn about logs.
Longer answer: Let's learn about logs.
The vertical axis shows total cases on a log scale, so it's like graphing log(cases) on a regular scale.
A per capita graph shows log(cases/population) which = log(cases)-log(population).
Total population is barely moving day to day, so this is like log(cases)-constant.
A per capita graph shows log(cases/population) which = log(cases)-log(population).
Total population is barely moving day to day, so this is like log(cases)-constant.
So on a log scale, the graph of total cases is just like the graph of total cases per capita, except the per capita line would be shifted down everywhere by the same amount.
Either way, whether you're plotting total cases or per capita cases, if you're doing it on a log scale, the slope of this line doesn't change.
And that means the story doesn't change, because it's the slope that people use to extrapolate bad outcomes becoming dramatically worse.
And that means the story doesn't change, because it's the slope that people use to extrapolate bad outcomes becoming dramatically worse.
You know you’re an economist when... a tweet blows up, and instead of begging you to hit my soundcloud, I figure it’s time for a quick refresher on logarithms.
