How might that affect the unemployment rate?

The graph from Monitoring Real Activity in Real Time: The Weekly Economic Index:

The sudden drop at the recent end of the graph is mostly the result of new initial unemployment claims in response to the pandemic. From the article at Liberty Street Economics:

The effects of the pandemic become visible in the week ending March 21. This week saw an unprecedented 3.28 million UI claims, a sharp decline in consumer confidence and fuel sales, and a more modest decline in steel production, but also a countervailing surge in retail sales, as consumers took to stores to stock up. In spite of this positive signal from retail sales, and the relatively mild declines in some other series, the UI release drove the WEI to a level not seen since the 2008 financial crisis.

Their graph goes back just far enough in time to show that the decline of 2008-09 was a relatively slow one, as compared to the drop of the week ending March 21st. Even one week before 21 March, the graph shows no indication that such a drop was coming. Yes, we might have known. But even in closeup, the Initial Claims data offer only the slightest hint:

Initial Claims for the week ending 21 March

What struck me about this is that there is not something wrong with the economy. Back in '08, when there was clearly something wrong with the economy, it took time for the crisis to develop. I've seen a few graphs that show the peak as far back as 2005 or 2004. Here, initial unemployment claims were decreasing thru the seventh of March; they show what would ordinarily be a substantial (but still ordinary) increase on the 14th; and then the increase of over three million the following week.

There was something wrong with the economy 15 years back, which led to the financial crisis of 2008. But, given the "new normal" where 2% growth is called "good", there was nothing wrong with our economy in mid-March.

I think that's good. I think it means we should be able to have a rapid recovery, once the pandemic is behind us, because there's nothing wrong with the economy. Economic declines are always more rapid than recoveries, of course, because recoveries require rebuilding. Declines don't: All we have to do is stop.

Our economy could not possibly recover in a week. But it should be able to recover (back to that new normal) much more quickly than it did after the financial crisis, if all goes well.

What could go wrong? We could default on our debts, enough to start a chain reaction which creates another financial crisis. (I'm sure there are other things that could go wrong, but my mind always goes back to debt.) If our debt was low -- not the Federal debt, our debt -- the chance of a chain reaction would be less. The Federal debt back at the end of World War Two was more than GDP, but debt for the rest of us was only about half the size of GDP. Our debt today is more than twice the size of Federal -- and two and a half times the size of GDP:

Federal Debt (blue) and Everyone Else's Debt (red) as Percent of GDP

If our debt went from half of GDP to two and a half of GDP, the risk of a chain-reaction debt implosion became what, five times worse?



The "Initial Claims" for 21 March indicates that 3,283,000 people filed for unemployment benefits that week. How might that affect the unemployment rate?

Sometime back I figured out how the unemployment rate relates to the number of people unemployed: Multiply the Unemployment Rate by the size of the Civilian Labor Force, and divide by 100 (to get rid of the "percent" thing) to get the number of people who are Unemployed.

You could go the other way and divide the number Unemployed by the Civilian Labor Force, then multiply by 100 to get the Unemployment Rate.

To preserve my sanity I will assume no change in the size of the Civilian Labor Force. I will also assume no one was hired, so that the Initial Claims number is the net change in the number of people who are working. That's not perfect, but I think it must be ballpark.

The count of people who were unemployed in February was 5787 thousand, or 5,787,000.

The Civilian Labor Force in February was 164,546 thousand, or 164,546,000.

5787 divided by 164,546 equals 0.035, so 3.5%. This matches the Unemployment Rate given in the FRED data.

//

Now take the count of people who were unemployed as of February, and add the 3,283,000 who lost their jobs during the week ending 21 March:

5,787,000 + 3,283,000 = 9,070,000

Divide 9070 by 164,546 to get the new estimated unemployment rate: 0.055, or 5.5%. Based on that one week's data alone, the unemployment rate could rise from 3.5% to 5.5%.

That's an increase of two percentage points. Divide 2.0 by the Initial Claims number of 21 March, and multiply by a million to see how much the unemployment rate changes for every additional million people unemployed: a 0.6 percentage point increase in the unemployment rate, for every million people.

To preserve my sanity I assumed some things that won't turn out to be true. To make up for that, take the 0.6%-per-million and call it half a percent. For every million people added to the unemployment rolls, unemployment will go up an additional half a percent. Maybe more.

Three million additional unemployed adds 1.5% (or more) to the February rate of 3.5%.

Due to some quirk in the timing of data collection and reporting, or so I hear, the 3,283,000 who lost their jobs will not be reflected in the Unemployment Rate number for March. I dunno. But 3.5% unemployment plus 1.5% more unemployment comes to 5% unemployment. And that's a change that happened in one week. The last time we had a 5% unemployment rate was September 2016.

Economic declines are always more rapid than recoveries.


Well it's no longer March. After I got this thing written, I checked FRED's Initial Claims data and (sure enough) the number is out for 28 March: 6,648,000. That's on top of the three million who lost their jobs last week. Six or seven million people at a half-percent higher unemployment rate per million comes to an increase of 3% or 3.5% -- and that's on top of the 5% rate after the week of 21 March. So, say an unemployment rate of 8 or 9 percent, after the week of 28 March. How high it goes after that, I can't say.

Initial Claims for the week ending 28 March
(Note the slight increase in the number for Week Ending 21 March)

Usually I prefer line graphs. But here, a line graph made it look like we had a three-million increase in unemployment last week, and another three million this week. That's not it.

Three million last week, and another six million this week. Closer to seven.

These increases occurred when they did, not directly because of the pandemic, but directly because of the response to the pandemic, the closing of businesses and all. Unemployment will not keep rising the way the number of Covid-19 cases keeps rising.

I can't guess when unemployment will stop the insane increase, but that will come first, and then the pandemic will show signs of retreat. Let's call it "soon".