Tuesday, May 5, 2020

A Very Grey Swan

Keynes and Knight famously simultaneously in 1921 identified the concept of fundamental uncerainty as a situation not understandable by using a probability distribution, an idea popularized by Nassim Taleb just as the 2008 crash happened as a "black swan."  Taleb defined white swans as situations describably by Gaussian normal distributions.  For situations not full uncertainty or white swans Taleb coined the idea of "grey swans," situations exhibiting "fat tails" and more generally lots of extreme outcomes, but yet possibly describable by probability distributions allowing more readily extreme outcomes.

I think this is what we are dealing with, although I claim no special expertise here.  Indeed, for me the issue is whether or not this current pandemic actually a black swan of fundamental uncertainty rather than "just" a very grey swan.  The reason it is the latter, despite the highly unusual outcomes on both health and economic outcomes, is that large numbers of experts have been warning for several years that we were facing with high probability a serious pandemic.  The current administration has in fact been officially faced scenarios not all that different from what has happened.  Among those is a largely ignored study in the president's own Economic Report for last year, where economists at the largely ignored CEA made such a study.  But, heck, who in this administration reads the Economic Report of the President, hahhahah!?

So, beyond my wisecracks, a very hard fact is that we do not yet really understand the statistical dynamics of this particular virus.  There is a lot more unexplained noise going on, which manifests itself at a superficial level in the wildly changing official projections of what is to come. To pick the greatest extremes, at one point our POTUS declared that while there were 15 cases in the US, this number would soon go to zero, but than not all that long later same individual at one point suggested, based on a model, that  deaths in the US could be in the millions.  But those numbers hae since been cut way back. But since then we have seen these forecasted numbers going all over the place.

In any case, I am struck by how much bizarre noise there is in all this.  So to get personal where I live, Harrisonburg, VA is a hotspot: with a 50,000 population we have 485 cases and over 20 deaths.  A city 25 miles to the southwest I once lived in and only slightly smaller, Staunton, has a whopping 14 cases and zero deaths so far.  I can think of many similar cases.  Why  has Greece had very few cases while fellow Med nations Italy and Spain have been among the hardest hit on the planet?

I think it was Einstein who said that the more we know, the more we understand how little we know.

Barkley Rosser

10 comments:

Anonymous said...

https://twitter.com/K_G_Andersen/status/1257512677611454464

Kristian G. Andersen @K_G_Andersen

India can do it. Slovenia can do it. Slovakia can do it. Greece can do it. New Zealand can do it. South Korea can do it. Australia can do it. Jordan can do it. Lebanon can do it. Croatia can do it. Norway can do it. China can do it. But apparently the United States can't do it.

Natalie E. Dean, PhD @nataliexdean

Kerala, India, is implementing “test, trace, isolate, support,” and it is paying off so far. They are working hard to keep case numbers low. What can we learn from them? A short thread. 1/5 

https://twitter.com/vincentrk/status/1257497158686826503…

Show this thread

11:29 PM · May 4, 2020

Anonymous said...

In terms of models being all over the place, one has disappeared from the CDC website - after being frequently cited by the Trump Administration, it recently projected an uptick in deaths and is no longer on the CDC website - here is a description of the model from the April 24 version and the associated weblink and the current weblink which shows nary a trace of the IHME model.

Institute for Health Metrics and Evaluation

Model name: IHME

Intervention assumptions

This model assumes social distancing stays in place until the pandemic, in its current phase, reaches the point when COVID-19 deaths are less than 0.3 per million people. Based on these latest projections, IHME expects social distancing measures to be in place through the end of May.

Methods

Non-linear mixed effects curve-fitting

April 24, 2020 version:

https://web.archive.org/web/20200501045214/https://www.cdc.gov/coronavirus/2019-ncov/covid-data/forecasting-us.html

Current (May 5, 2020):

https://www.cdc.gov/coronavirus/2019-ncov/covid-data/forecasting-us.html

Anonymous said...

perhaps Italy and Spain are more industrialized with larger cities than Greece? perhaps there are more chicken processing plants in the Harrisonburg area compared to Staunton plus Harrisonburg is right off I-81 and Staunton isn't.

rosserjb@jmu.edu said...

There are poultry plants near Harrisonburg although not in it. However, Staunton is also on I 81 as much as Harrisonburg is. Harrisonburg has also had a nursing home that got hit hard with over 100 cases ans most of the deaths in the city, plus we have more immigrants from various places than Stanton, and I have heard rumors of some of them having higher rates, and not just those who work in the poultry plants. But the poultry plants a re a lot of it. But still, Staunton's number is very low, and Waynesboro, a bit smaller, has only 10 cases.

Is it true that you live in Stephens City, VA, Anonymous?

GeorgeNYC said...

Does Staunton have a hospital? Might it be that the Coronavirus patients get sent somewhere else for treatment and this get counted with them?

Anonymous said...

Really nice thinking in this speculative essay.

rosserjb@jmu.edu said...

George,

Both Staunton and Harrisonburg have hospitals. Staunton also has nursing homes, but somehow none of them have had outbreaks like Corridus in Harrisonburg. As I sais, there is a lot of random noise in all thise.

Anonymous said...

No, but have been there and it is right off I-81 too! The interstate has been an economic engine, but in a way, now contributes to virus spread.

Anonymous said...

https://www.nytimes.com/2020/05/06/us/coronavirus-updates.html

May 6, 2020

Nearly 1 in 5 young children in the United States are not getting enough to eat, new research found.

Research released Wednesday shows a rise in food insecurity without modern precedent. Nearly a fifth of young children are not getting enough to eat, according to surveys of their mothers by the Brookings Institution, a rate three times higher than in 2008 at the worst of the Great Recession, reports Jason DeParle.

When food runs short, parents often skip meals to keep children fed. But a survey of households with children 12 and under by Lauren Bauer, a Brookings fellow in economic studies, found that 17.4 percent reported the children themselves not eating enough, compared to 5.7 percent in the Great Recession. Inadequate nutrition can leave young children with permanent developmental damage.

“This is alarming,” she said. “These are households cutting back on portion sizes, having kids skip meals. The numbers are much higher than I expected.” ...

Sandwichman said...

How about the concept of the Black and Blue Swan? Economic "science" has been abused and mutilated for so long for the benefit of those at the top that what we can "predict" is that things will get more brutal and nasty for most of us while the prospect of never ending bailouts will send the stock market soaring.