Real GDP for 2020 II

U.S. real GDP declined at an annualized rate of 32.9% in the second quarter.

As pretty much everyone knows, that is by far the worst performance for any quarter going back the 70+ years to when we think we can first estimate this data.

I’ve been very reticent about commenting on the macroeconomy over the last 5 months, because, as a macroeconomist, my considered opinion is that none of us knows wtf we’re talking about. This is unprecedented.

In fact, I’d hesitate to call this a recession or depression. Here’s why.

A recession (or depression) is a period in which economic events cause broader cultural, political, and societal problems. But with a pandemic, it’s the other way around: broader cultural, political, and societal events are causing economic problems.

In my Principles classes, I stress that part of the problem in studying business cycles is that every one is different from the ones before. I hope 2020 proves my point to you.

Prima Facie Causality

6/20: Trump rally in Tulsa without social distancing.

6/29: Herman Cain tests positive for COVID-19.

7/30: RIP.

************************

In stats classes we all learn that correlation does not imply causation.

That’s always true in cross-sectional analysis.

But it’s not quite so always true in time-series analysis.

And, do note most importantly that the converse is true: causation does imply correlation (although you might have to correct for a lot of interference from other data to uncover that).

Going back to Bertrand Russell (and others), before there was a field of time-series analysis, it was established that temporal ordering was an important feature of causality: the cause must precede the effect.

Correlation between past events that could theoretically be causal, and future events that could theoretically be effects, is known as prima facia causality.†

And it’s a substantially stronger indication of (true) causality than mere correlation.

In this case, the 9 days between the rally and the positive test is right in the middle of the current CDC range for exposure to onset of symptoms.

† My 1992 dissertation did quite a bit of work on this in the context of monetary policy.

A Bad Sign with Local Cases

Six days a week the Southwest Utah Public Health Department publishes information on local cases. This always gives a breakdown by county.

For 8 days now, they have been unable to clear out a backlog. This shows people who've tested positive, but whose county has not yet been designated.

This points to a mismatch between people testing positive and people keeping track of those who test positive.

In some sense, this is not surprising: tests run has doubled over a little more than a month, and positive test results have doubled over a little less than a month. Yes, more people will be needed to process that information. Having said that, it's not like this growth isn't steady and predictable.

Something's up. Occam's razor suggests this is an early sign the epidemic is getting out of hand in the local region.

Arizona

Once scaled for population (which is the reasonable way to do it), Arizona's COVID-19 outbreak is now worse than that of NYC in April.

This is showing the rate of new cases (positive test results) averaged on a daily basis over the last week.

Utah is not shown on this chart, but currently we'd be right below (the faint) label on the right for Tennessee.

Washington County is at just about the same spot.

Iron County is a bit lower, at around 120.

Legacy Media Bias In the U.S., COVID-19 Performance

If it isn’t scaled, beware. This is the truth:

Here'e some perspective on how the US is faring versus Europe in the spread of Covid

On a per-capita basis, of the European majors only Germany has done better than the US in death rate pic.twitter.com/1UvwtVqoDT

— Chris Anderson (@chr1sa) July 4, 2020

Not sure why Spain is included with the big four, while Poland is not. Prejudice probably.

Do keep in mind that all of these countries have more socialized healthcare provision than the U.S., and that Germany is known for probably having the least socialized provision of these five.

Georgia Tech Faculty Revolt

Faculty at Georgia Tech are up in arms about the state of Georgia’s plans for how universities will run in the fall when they re-open.

GT is going to try to do face-to-face, with professors required to wear masks, but students not required to wear masks.

A supermajority of faculty signed the statement.

Do keep in mind that the dominoes of university closings in the spring started with a single one. That was the University of Washington, in a state with an early outbreak that never got out of hand.

Currently Georgia Tech has students living in fraternity housing over the summer, and has had about a dozen test positive.

Via Marginal Revolution.

(Implicit) Event Study for Utah

An event study is a technique in finance, where you take data before and after an event (and sometimes during), and check whether the data behaved differently around the event.

Johns Hopkins now has a graphical device for doing something like an event study with public policy choices related to COVID-19 for individual states.

******************************************

WARNING: Proceed with care. Doing event studies with trending data is not straightforward. You have been warned. There's a reason I am not writing up something too sophisticated about this (and it's not because I think I understand why this looks the way it does).

******************************************

So here's a time series showing new cases:

Obviously cases are trending up. Part of that is our state isn't one of the ones that's doing OK right now. But another one is that as we test more and more, we find more cases (which implies we should see some arcing up to the right, but it should be fairly smooth ... and it's not).

Anyway, what's cool about this is that the interactive graphic (and you should go to the site to play with it) shows public policy events: towards lockdown are in orange-ish, towards opening are in bluish-green, and neutral ones are in gray.

QUICK UPDATE: I chose April 28, a neutral one, for zero reason ... just showing an example.

The problem is, with trending data, especially when we don't understand the underlying mathematical process (and we're sure it can't trend forever), that You. Can. Not. interpret the orange moves as keeping cases under control, and the green ones as taking the lid off the pot.

You. Can. Not. But everyone does anyway. And they're full of BS. Do note that I'm not denying that this may be the true story, just that I know better than to open my big mouth with this sort of data.

And here's a chart showing new deaths:

Same issues here, although it helps that they're not trending, and don't have an obvious kink like above. Even so, the warning still applies.

Hat tip to MJ.

UPDATE: As always, if growing data isn't logged, beware. In this case, it's a double problem. Really, we ought to start with the series we know is growing — total cases or deaths — then log that, and take differences to get something like what they're trying to get at here. But, by skipping two steps, my guess is that these cool charts are just junk in this form.

Hospitalization Data

We need better hospitalization data, and it's been getting better. But there's a way to go. This page from the COVID Tracking Project entitled "To Understand the US Pandemic, We Need Hospitalization Data - and We Almost Have It" touches on some of the details.

Didja' Notice Millard County?

Not sure what towns the outbreak is hitting, but Millard is now near the top of the list of counties getting hit.

Now, there are roughly 3,000 counties in the U.S., and Millard is about 120th (the list I'm going from isn't numbered, so that's an estimate).

For scale, rates are based on per 100K people. Millard is at 36, while Salt Lake is at 24, Washington is at 22, and Iron is at 14. Obviously though, Millard does not have anywhere near 100K people, so the total number of cases is less than 36.

FWIW: I have relatives in the suburbs of Buffalo, New York. The current rate for Iron County is about 4 times higher than there.

More Southwestern Utah Deaths

We had a 12th death announced today. There was an 11th the other day, but I didn't not it here.

Ummm ... that means we doubled in 21 days (the 6th death was on or around June 12th).

First it was 28 days to double (1 to 2), then 25 (2 to 4), then 23 (4 to 8), and now 21 (6 to 12).

What People Should Notice About Cedar City

It has been 21 days. There has been no uptick in case rates in Cedar City since the Collin Raye concert. That's a good thing.