Mask Mandate Analysis

By August 31, 2020 Commentary

A few people have asked me about this and I am working on it, but want to be rigorous.  Nonetheless I will give you a preliminary peek.  The state’s lag between the number of cases they report on a day in the daily report and getting the results into the right date for cases by specimen collection day is fairly long, so I just have to be pretty sure a specimen collection date is pretty complete before I consider the analysis to be using accurate data.  That dump of 19,000 tests by one lab on August 18 really screws it up.  The state is still working through that backlog.  It is going to add more cases to days included in the analysis.  But I can tell you with a pretty high degree of certainty right now that the mask mandate is not going to show any effect on cases, in fact they are going to be up.  And I will also predict that when confronted with this fact, the state’s lame response will be “well if it weren’t for the mask mandate they would be up even higher”.

I want to again set the background.  You can find the videos on YouTube and probably other sources, so you can verify exactly what the Incompetent Blowhard in particular said.  He said that forcing people to wear masks would cut cases dramatically.  He cited only a couple of horrible studies that have no credibility and ignored all the other evidence of no effect.  Those were his justifications for the mandate.  He promised cases would go down significantly.  As I explained at the time, this was incredibly stupid, but what else would you expect from an Incompetent Blowhard.

Here is why it was dumb and if I know this so should his supposedly so expert advisors and officials.  In a multi-factorial situation, it is very hard to prove that any one thing is responsible for a change in either direction in an outcome.  When the measure you choose, cases, is at a very low prevalence, showing a statistically significant difference is about impossible.   When your intervention, making people wear masks, is already voluntarily occurring at a high level, how much difference is a mandate going to make?  That is magnified when several of your largest population centers already had mask mandates.  When much of your transmission doesn’t occur in settings where people wear masks, why would they make any difference?  And when all the evidence tells you that masks don’t actually slow community transmission, why would you put yourself on the line with a stupid promise?  The answer to the last part is self-evident–because you are Incompetent.

Notwithstanding that I know why the mandate never could and never will make a difference in cases, we are going to hold the Incompetent Blowhard to his word.  We are going to look for a significant reduction in cases.  And I don’t think significant is 5% or 10%, I think it is at least 25%.  Some studies claim masks lead to much greater reductions.  How long will it take to see the trend?  The Health Department officials said two to three weeks and weaseled even more lately, because they know it isn’t making a difference.  The CDC says the average time from infection to symptoms is less than a week, so let’s assume a person would seek a test when they have symptoms.  Lots of cases are found from testing of people who are asymptomatic, so really, a “significant”effect on cases should be seen in week or so.  But we will cut the Blowhard some slack and wait for two or three weeks.  One more twist is that cases should probably be normalized to testing levels, even though the state didn’t specify this as a condition.  I will do it both ways, although matching testing to cases by specimen collection date accurately is impossible the way the state reports data.

To do this analysis you have to establish the trend, what was happening for several weeks before the mandate–were cases going up, were they flat, were they declining?  If the mandate effect doesn’t show up for two or three weeks, whatever date you pick when the intervention should show an effect, everything before, even if after the date of the mandate, goes into the pre-intervention measuring period.  Then we are just going to keep tracking and tracking.  I aggregated this by weeks, you could do daily averages before and after, but it all will give the same result.  Barring some further massive case dump, which will make it even worse for the Blowhard, here are weekly case numbers based on today’s table of cases by specimen collection date.  I did not start with the week of July 4, since that was likely an off week due to the holiday and the riots, I mean protests.  That is doing the Blowhard a favor because it was a low week.  I will get this in a table and chart for the real analysis.  Dates are day the week ended.

July 11–4323 cases.        July 18–4491 cases        July 25–4585     August 1–4683 cases    August 8–4308 cases

Now the first weeks we would see an effect according to the state,  August 15 and August 22.

August 15–4333 cases     August 22–4722 cases.    Cases will go up in these weeks, as further results are added.

Let’s throw in the week of the 29th, even though it is far from complete, August 29, for example, right now has 2 cases for that day.   Week of August 29–3784.   I threw that in because today the state added over 900 cases, which have yet to be farmed out to the table.  Based on relative completion factor, August 29 will be a 5000 case week.  Even without any fancy statistical analysis, you can see that the Blowhard is screwed.  No obvious decline.  Cases vacillating around the same general area.  In fact the week ending on August 22nd was the highest week in the last seven, and by then the officials tell us we would definitely be seeing the effect.  And that week is not nearly complete.  It too will end up being a 5000 case week.  Is it possible masks actually increase transmission?



Leave a Reply