Thread by @JustinWolfers

Teamed up with @qdbui to write for @nytimes about what today's massive initial jobless claims numbers mean for the economy.

Spoiler: These numbers are less informative than you might think.
nytimes.com/interactive/20…

Here's the key point: This recession is different.

Normally, it takes months or years for job losses to accumulate as businesses slowly learn there's a recession that'll hurt them.

This time, it's like an umpire whistled all the players off the economic playing field.

So today's numbers simply can't be compared with past recessions, because the *speed* of the downturn is so different.

We can't (yet) tell whether the big number is because this downturn is unusually compressed and synchronized, or unusually deep.

Some simple numbers: The Great Recession led to about 26 million extra jobless claims, spread over 5 years.

If the current rate of initial claims continues for 8 weeks, we'll also be at 26 million extra claims.

That would suggest this downturn is roughly as bad.

But if initial claims continue at this rate for only half as long (4 weeks), then this downturn is only half as bad.

If they continue at this rate for twice as long (16 weeks) then this downturn is twice as bad as the Great Recession.

We won't know which it'll be for many weeks

Maybe it's best not to compare this downturn with past recessions, but rather, past natural disasters.

Here's how initial claims in Louisiana responded in the wake of Hurricane Katrina.

The rise in initial claims in Louisiana following Katrina was larger than this morning's national rise. Claims in LA rose to twenty times its long-run average levels. Today, the US rose to more likely ten times its long-run average.

I calculated that Katrina boosted initial claims by a total of 350k, spread over 14 weeks.

Scale that up for the U.S.—the national economy is 75 times that of LA—and it says that a Katrina-like shock would add up to 26 million extra jobless claims. (Same as the last recession!)

But again, so far the coronavirus shock hasn't yielded an immediate jump in jobless claims as large as that in post-Katrina Louisiana, so perhaps the outcome won't be that bad.

But maybe the coronavirus shock will persist longer, so perhaps it'll be much worse.

Who knows.

Or perhaps this shock will be more like how Texas responded after Hurricane Harvey.

There, the initial shock to initial jobless claims was also large, but it receded pretty quickly.

If the U.S. follows this trajectory then the coronavirus recession will be relatively mild.

Point is, today's massive spike in initial unemployment claims is not unprecedented. What matters is how long it lasts.

As we write in NYT: "early data from a highly synchronized economic catastrophe provide extremely imprecise signals about the depth of the ensuing recession."

Depressing conclusion: We're not going to know how bad this downturn is for at least several weeks.