Sunday, March 8, 2020

You are 43 times more likely to be killed by Corona virus if you have a gun in the house

One of the most infamous "scientific" studies conducted on gun violence is the 1986 Protection or peril? An analysis of firearm-related deaths in the home by Kellerman and Reay.  Dave Koppel ripped this study to shreds (click through for link) but in a nutshell it boiled down to bad data.  One of the factors not considered by Kellerman and Reay was whether a felon lived in the house.  It is certainly plausible that gun crime is correlated at a much higher level with criminality than it is with firearm possession.  Political motivation likely was a major factor in this omission from the study.

I post all the time about the wretched state of the climate databases.  Once again, political motivation is the almost certainly present.

Without good data, your predictions for the future are junk.  And so to the Corona virus.  We simply don't have any idea of what's going on, or how bad it is.  Consider:

  • All of the data out of China is untrustworthy.  The Chinese Communist Party plays their cards very close to the vest, and information is always massaged prior to release.
  • Data from the CDC is wrong on a whole number of angles, but most egregious were protocols to track potential infected people ignoring people who hadn't been in China.  Since the virus can incubate non-symptomatically for a month, it could spread to or from people who were not recently in China.
  • The CDC completely screwed the pooch on Test Kits.  By keeping control of the creation and manufacture (and screwing that up, too), there were only a few hundred when other countries were making thousands or millions.  Use of a test kit is a major data component of the overall statistics, and what we know from the use of these kits tells us nothing about how many people are infected.
  • Reported death rates vary from a high of 3-4% to 1% to a low of 0.1%.  Sure, we know the death count with some level of reliability, but since we don't have any idea about how many people contracted the virus, there's no good way to predict how bad (or less bad) things might be.

Like with climate science or firearms research, if the data are bad your public policy will be bad.  Now look at politics - governments are strongly motivated to push a story they find useful for their purposes. Mass quarantine is useful for establishing control, and an atmosphere of emergency is, too.  A healthy skepticism of anything you hear from the Organs Of The State is A Good Thing®.

So what's going on?  Beats the heck out of me.  But I simply don't think that the data are reliable enough to support any conclusions about how bad things are.  We'll know a lot more in 60 days.

In the meantime, wash your hands and stop touching your face.  Have some food stores in case the government decides that a quarantine is convenient to their long term goals of gathering more power to themselves.

6 comments:

ASM826 said...

We know that nothing reported out of China is reliable. Same for Iran. It seems that nothing out of Washington can be relied on either. Since they are only reporting confirmed cases and there aren't enough test kits for a fraction of the possible infections, we have no data.

In the absence of that data, it is not unreasonable to presume that this is bad, and that is why it is unreported. Maybe not human extinction bad, but potentially as bad as a world war in terms of deaths and economic disruption.

My lily pad example from yesterday's post? If it doubled every day and the lake was covered on day 37, only half the lake was covered on day 36, 1/4 on day 35, 1/16 on day 34, and so on. A week before the lake was covered, the lilies would have hardly been noticeable.

If the mortality is as sharply skewed on age as has been reported, the one bright spot in all of this is that it will save the Social Security Trust Fund.

Old NFO said...

Sigh... Here we go again...

danielbarger said...

We can never know the true rate of morbidity. Because we will never know how many people actually get infected. For every person who presents as symptomatic there could be anywhere
from one to a dozen or more who are infected and either never have symptoms or don't get sick enough to seek medical care and be counted. Thus we can only know the morbidity rate as a percentage of those who SEEK CARE...not the rate as a percentage of those infected. And the same can be said for Influenza and most other respiratory communicable illnesses.

Richard said...

Wasn't it the CDC that published the bogus gun study you started with? I guess some things never change.

LSP said...

Black rain in Japan.

China has a lot to answer for.

Ted said...

Before you completely panic take a look at this site

https://www.arcgis.com/apps/opsdashboard/index.html#/bda7594740fd40299423467b48e9ecf6

See the curves on the bottom right. the top one represents mainland china.
notice that the number of cases in mainland china has plateaued in the number of cases. It is no longer increasing. It's unclear if this is because they have stopped reporting, or everything is shut down or they are just plain lying.

Going back to the lily pad analogy, sometimes there are outside factors at play that limit the grow beyond certain levels. the tempature above the deeper portions of the lake may exclude the gruth of lily pads so that once they cover the area of the cove where conditions are right they can't expand further.