Coronavirus Update: The US and the Grand Canyon State

 

I have another COVID-19 update for you today, focusing on the U.S. generally, and specifically on my home state of Arizona. I’ve heard that there has been some media focus on Arizona in recent weeks.

My last post was on June 23 (here), with my standard analysis comparing reported COVID-19 deaths in the US and Western Europe. I did run that analysis again today but am not going to post new graphs, as the trends are unchanged.

As usual, my data source is Johns Hopkins (here). My report today uses U.S. data through July 4, 2020, which includes daily reported deaths at the county level throughout the US.

If you’ve followed the COVID-19 outbreak in the US, you know that the early deaths were very highly concentrated in the New York City area, and in the states of New York, New Jersey, and Connecticut which contain the NYC metro area. My general sense is that recent media coverage has focused, for political purposes, I think, on increases in cases (and perhaps deaths) in other states. As a result, like me, you may have the impression that NY, NJ, and CT have the outbreak under control, while supposedly irresponsible red states like Arizona, Georgia, and Texas are having outbreaks.

I. New York City is Still Above the National Average

Here’s my first graph, showing the 7-day average of reported COVID-19 deaths per 100,000 population, for (1) the US as a whole, (2) NY-NJ-CT, and (3) the US excluding NY-NJ-CT:

You can see a huge jump in NY-NJ-CT (yellow line) on June 25, which is an anomaly in the data, due to New Jersey reporting about 1,800 prior deaths on that date (due to a change of classification criteria, I think). This doesn’t really matter, because the effect of this anomalous increase is passed in the 7-day average.

The notable fact is that the 7-day average for NY-NJ-CT (yellow) continues to be higher than the 7-day average for the rest of the country (purple).

II. Arizona’s COVID-19 Deaths In Perspective

It is true that Arizona has experienced an increase in COVID-19 deaths in recent weeks. Arizona remains well below the overall national average in cumulative COVID-19 deaths.

In addition to comparing Arizona to the country as a whole, I’ve included Los Angeles County in my calculations. LA County has a population of about 10 million, somewhat more than Arizona’s total population of about 7.2 million. LA County is supposedly under a strict lockdown, while Arizona is supposedly an irresponsible red state that rushed to reopen.

Here is a graph of cumulative reported COVID-19 deaths, per 100,000 population, for Arizona, LA County, and the US as a whole. I’ve also included the trend line for NY-NJ-CT (which is pretty much a proxy for the NYC area).

This graph makes it difficult to understand what all the fuss is about in Arizona. Our COVID-19 deaths, adjusted for population, are well below the national average, lower than locked-down LA County, and far lower than the NYC area.

Arizona is a large and complicated state, as I will address in Section IV below, which drills down to more local data at the county level. We have two large metro areas: (1) the Phoenix area, in Maricopa County, and (2) the Tucson area, in Pima County. Johns Hopkins does not report COVID-19 death counts in Arizona by city, but only by county. So I am using Maricopa County to represent the Phoenix area, and Pima County to represent the Tucson area.

Maricopa County has a population of almost 4.5 million, which is 61% of the state’s population. Pima County’s population is just over 1 million, another 14% of the state total. The counties are big geographically, each around 9,200 square miles, which is bigger than New Hampshire, Massachusetts, or New Jersey. (Maricopa County actually edges Vermont, too.)

Here is the graph of population-adjusted COVID-19 deaths for Los Angeles County, Maricopa County (Phoenix), and Pima County (Tucson), plus the total national figure:

As you can see, LA has suffered less than the national average; Pima County (Tucson) has suffered less than LA, and Maricopa County (Phoenix) has suffered least of all.

But Arizona has mandated masks, and closed bars, and I didn’t get to go to church again today. (The closing of church is not mandated by the state; it was the decision of my church leadership.)

It is true that the trend line in Maricopa County (Phoenix) has been steeper than the others in the past week or two. Here is the same data, showing the 7-day moving average:

To me, this does not look like a serious cause for concern. The recent increase in Maricopa County (Phoenix) is a bit troubling, and actions have already been taken (including ordinances requiring masks in public, which also apply in Pima County).

III. Lies, Damn Lies, and Statistics

Just for kicks, here’s a quick example of how Fake News works.

Say that I was a Wokeist propagandist journalist at a major news outlet. Let’s make it the New York Times because I really think that they are the very best at this. (Admittedly, this is a strange use of the word “best,” but you’ll see what I mean.) Here’s what I would do (and what I have done).

Pick a particular day around the peak of COVID-19 deaths in NYC, on which deaths in Maricopa County (Phoenix) were very low. Then pick a recent day on which NYC deaths were unusually low, but Maricopa County deaths were unusually high. Two great days happen to be April 7 (the peak in NYC) and July 1. Then I could “honestly” write a headline saying:

Between the peak of the outbreak in New York on April 7 and July 1, COVID-19 daily deaths in New York City declined 99%, while daily deaths in the Phoenix area increased over 1,000%.

Well, that’s true. On July 1, NYC reported 8 deaths, a rate of 0.10 per 100,000, while Maricopa County reported 46 deaths, a rate of 1.03 per 100,000. On April 7, NYC reported 814 deaths, a rate of 9.76 per 100,000, while Maricopa County reported 4 deaths, a rate of 0.09 per 100,000.

Then I could praise Gov. Cuomo and Mayor DeBlasio for their enlightened leadership while castigating that Trump-loving right-wing hack of an Arizona governor, Doug Ducey.

This would conveniently distract everyone’s attention from the following graph, showing the 7-day moving average of daily COVID-19 deaths per 100,000 population. Note that this is the exact same data as my prior graph comparing LA, Phoenix, and Tucson, except that New York City is added:

The green line is New York City. Notice that the death rate was so high back in April that the trend lines for LA, Phoenix, and Tucson are hard to differentiate.  That tiny yellow hump at the far left edge of the graph is the source of the criticism of Arizona and our supposedly reckless governor.

For the record, the latest 7-day average of reported COVID-19 deaths per 100,000, as of July 4, are:

  • Arizona (statewide): 0.47 per 100,000
  • Maricopa County (Phoenix): 0.44 per 100,000
  • New York City: 0.25 per 100,000
  • Los Angeles County: 0.24 per 100,000
  • Pima County (Tucson): 0.19 per 100,000

I don’t think that we have much of a problem here in Arizona.  I do think that it was sensible to take the relatively minor precautions that we adopted a couple of weeks ago while keeping almost all businesses open.

IV.  A Deep Dive in Arizona

Admittedly, it’s hard to do a deep dive in Arizona.  It’s hard to find any water at all, especially at this time of year.

I have one more graph for you, at a level of detail that requires a pretty good understanding of Arizona geography and population at the county level.  I’m not sure if this will be of interest to non-Arizonans, but I’m going to include a brief discussion of Arizona demographics.  Fortunately for you (or maybe not), I’ve been living here for 47 years, so I know the state pretty well.

Arizona is a pretty big place, over 113,000 square miles, which is just smaller than Italy.  It is the 6th largest state, with a population of almost 7.3 million.  As noted above, the most populous county is Maricopa County, which includes the Phoenix metro area, with a population of almost 4.5 million, over 60% of the state total.  The second most populous county is Pima County, my home county, which includes the Tucson metro area, with a population of a bit over 1 million.  There are about 1.7 million people in the rest of the state, spread over 13 more counties (a total of 15).

Arizona has an Indian population of over 270,000, second only to Oklahoma (source here).  That is just under 4% of the population of the state.  Many Arizona Indians live on a number of reservations, some of which are huge.  The Navajo Nation is over 27,000 square miles — bigger than West Virginia — and is mostly in northeastern Arizona, though it stretches into Utah and New Mexico as well.  We have three additional reservations bigger than Delaware: the Tohono O’Odham Nation, west of Tucson, and two Apache reservations, the White Mountain and San Carlos reservations, in east-central Arizona.  Over 27% of Arizona is tribal land (source here).

To make matters more confusing, many of our counties have Indian names, some of which coincide with the names of tribes, but the tribal reservations are not necessarily in the county named for the tribe.  We have an Apache County (far northeastern Arizona) and a Navajo County (also northeastern Arizona, but west of Apache County).  But large sections of the Navajo Nation are in Navajo County, Apache County, and Coconino County (home of our friend Gary Robbins).  The White Mountain Apache reservation is in Apache, Navajo, and Gila Counties, and the San Carlos Apache reservation is in Gila and Graham Counties (and not in Apache County).

Got that?  I know, it’s complicated.  Here’s a map of tribal lands in Arizona:

You’re probably wondering why I’m telling you all of this.  The reason is my suspicion that a significant portion of Arizona’s COVID-19 death toll, and perhaps much of the current increase, is occurring on Indian reservations.

According to the Arizona Department of Health Services (here), 16% of COVID-19 deaths in Arizona — 298 of 1,809 total deaths — were Indians.  Remember, they are less than 4% of the state population.  The worst outbreak is in the Navajo Nation, which reports 377 deaths (here), though some of those may have been in New Mexico and Utah.  The White Mountain Apache reservation had 20 deaths as of last week, and more than 1/8 of the tribe were reportedly infected (here).

The deaths of Arizona Indians did not necessarily all occur on reservations.  There’s no law requiring an Indian to live on a reservation, which is as it should be.  The family next door could be Indian, though I don’t happen to have known many Indians during my time in Arizona (other than my time spent specifically on a reservation).  I did have a buddy in junior high who was a Tohono O’Odham, one of my best friends at the time, though we still called them Papagos back then.  (Papago was a name they didn’t like.  I don’t think that it was ever intended as an insult.  It was just a mispronunciation, based on what another tribe called the Tohono O’Odham.  So we palefaces properly agreed to call them by their rightful name, which means “Desert People.”)

On to the Johns Hopkins data.  Here is the graph of reported COVID-19 deaths in Arizona, by county, per 100,000 population:

Maricopa County (Phoenix area) is the yellow line, and Pima County (Tucson area) is the red line.  These are both pretty low.

The biggest problems, by a wide margin, are in Apache County (blue) and Navajo County (brown), with death rates that are about 5 times higher than the statewide average.  Coconino County (orange-brown) is third, with a death rate over 2 1/2 times the statewide average.

The Navajo Nation is located in Apache, Navajo, and Coconino Counties.  The Hopi reservation is also in Navajo and Coconino Counties, and the White Mountain Apache reservation is in Apache, Navajo, and Gila Counties.   (I’m particularly familiar with the White Mountain Apache Reservation, and know that very little of the tribe’s population is in Gila County.)

The next hardest-hit Arizona counties are Yuma County (light gray) and Santa Cruz County (light blue).  These are border counties, and I suspect that the higher levels here are associated with cross-border issues and perhaps illegal immigration.  I realize that if you look on the map, you’ll see that Pima County and Cochise County have lengthy borders with Mexico, but you have to understand the population distribution of these counties and the locations of the border crossings to realize that one would expect bigger problems in Santa Cruz and Yuma Counties.

My suspicion, therefore, is that the COVID-19 figures reported in Arizona are heavily concentrated on the reservations (especially the Navajo Nation) and in a pair of border towns (Yuma and Nogales).  There seems to be much less cause for concern in the rest of the state.

ChiCom and BLM delenda est.

Published in Healthcare
This post was promoted to the Main Feed by a Ricochet Editor at the recommendation of Ricochet members. Like this post? Want to comment? Join Ricochet’s community of conservatives and be part of the conversation. Join Ricochet for Free.

There are 41 comments.

Become a member to join the conversation. Or sign in if you're already a member.
  1. Gumby Mark (R-Meth Lab of Demo… Coolidge
    Gumby Mark (R-Meth Lab of Demo…
    @GumbyMark

    To clarify, the figure is from the AZ DHS site, which also included the note explaining the death certificate matching.  And yes, AZ, like many states (and countries) consistently sees a death count drop off (though not a case count drop off) on Sundays and Mondays which is why Tuesdays are generally the highest.  I also follow the Maricopa County site daily and it shows the same pattern.

    That is why I am waiting to see if the AZ and Maricopa data follow their usual pattern and drop off somewhat from today’s report which would mean a continuation of the slow increase in deaths.

    Both AZ DHS and Maricopa have a lot of very good data on their sites, having greatly improved since March.

     

    • #31
  2. Steven Seward Member
    Steven Seward
    @StevenSeward

    Gumby Mark (R-Meth Lab of Demo… (View Comment):

    To clarify, the figure is from the AZ DHS site, which also included the note explaining the death certificate matching. And yes, AZ, like many states (and countries) consistently sees a death count drop off (though not a case count drop off) on Sundays and Mondays which is why Tuesdays are generally the highest. I also follow the Maricopa County site daily and it shows the same pattern.

    That is why I am waiting to see if the AZ and Maricopa data follow their usual pattern and drop off somewhat from today’s report which would mean a continuation of the slow increase in deaths.

    Both AZ DHS and Maricopa have a lot of very good data on their sites, having greatly improved since March.

    Please keep us updated.

     

    • #32
  3. The Reticulator Member
    The Reticulator
    @TheReticulator

    Steven Seward (View Comment):

    Gumby Mark (R-Meth Lab of Demo… (View Comment):

    As an update, Arizona reported a record 117 deaths today. HOWEVER, this figure does not represent current conditions. 52 of the 117 deaths were from death certificate matching going back as far as April 12. The “real” figure is 65 which while high (I believe the highest previous total, excluding certificate matching, was in the 40s) is not a step change like the 117 figure would lead you to think. From looking at county data my guess is a lot of the older cases added in are from Yuma and Pima counties. If pattern of previous weeks Tues-Sat data reporting holds, the next few days should have less than 65 deaths. If not, we are in changed circumstances.

    I assume you got that figure from the Worldometers website. I looked at their graph on Arizona Covid deaths. It looks like their reporting lapses are even greater than most states, where few deaths are reported on the weekends and holidays, and are made up for on the following Mondays and Tuesdays where the deaths are counted. Their dip in reporting seems to be mostly on Sundays and Mondays. For instance they recorded 4 deaths this past Sunday and only a single death yesterday. The previous Monday they reported Zero. Presumably today’s count of 65 is an amalgam of the last three or four days.

    I’m assuming that Arizona Patriots’s data from Hopkins has smoothed over those uneven reporting issues, though he didn’t chart a specific graph on daily deaths for all of Arizona. When I look at the smoothing line for the seven day moving average on the Worldometers graph, I noticed that Arizona really had only one short period from May 8th through May 27th where the number of deaths decreased. This may be due simply to the low numbers we are dealing with that do not make very smooth graphs. Still, the recent rise in deaths is not very substantial compared to most states.

    I sometimes wonder if the people who write the fearmongering headlines even look at the data. I look at those worldometer graphs every day now.

    • #33
  4. Steven Seward Member
    Steven Seward
    @StevenSeward

    The Reticulator (View Comment):

    I sometimes wonder if the people who write the fearmongering headlines even look at the data. I look at those worldometer graphs every day now.

    Many of them probably are looking at the raw data and just ignoring it.  One of the biggest disconnects I’ve ever seen between “news” and “real life” has been looking at the raw data over the Coronavirus and hearing what journalists are reporting.  Even supposed conservative writers like Jim Geraghty at National Review put a negative spin on all of this.

    Our local “conservative” Salem Radio affiliate, WHK often gives updates on the Coronavirus in the Cleveland area, Cuyahoga County which has 1.3 million people.  I’ve heard the news announcer say with grave voice and concern that “There have been 19 new cases of Coronavirus in the county over the last 24 hours.”  Only 19! You’ve got to be kidding.  A city block should be expected to have more than that.  It should be reason to celebrate, but no matter what the figures are, they are presented with the forbidding air of doom and gloom.

    • #34
  5. Gumby Mark (R-Meth Lab of Demo… Coolidge
    Gumby Mark (R-Meth Lab of Demo…
    @GumbyMark

    Steven Seward (View Comment):

    Gumby Mark (R-Meth Lab of Demo… (View Comment):

    To clarify, the figure is from the AZ DHS site, which also included the note explaining the death certificate matching. And yes, AZ, like many states (and countries) consistently sees a death count drop off (though not a case count drop off) on Sundays and Mondays which is why Tuesdays are generally the highest. I also follow the Maricopa County site daily and it shows the same pattern.

    That is why I am waiting to see if the AZ and Maricopa data follow their usual pattern and drop off somewhat from today’s report which would mean a continuation of the slow increase in deaths.

    Both AZ DHS and Maricopa have a lot of very good data on their sites, having greatly improved since March.

    Please keep us updated.

     

    Today AZ deaths dropped from 117 to 36, which is about where the recent run rate for mortality has been.  Maricopa dropped from 55 to 23.  Because of the sheer volume of new cases, hospitalizations and ICU usage continue to increase which remains a concern.

    • #35
  6. MISTER BITCOIN Inactive
    MISTER BITCOIN
    @MISTERBITCOIN

    Gumby Mark (R-Meth Lab of Demo… (View Comment):

    Steven Seward (View Comment):

    Gumby Mark (R-Meth Lab of Demo… (View Comment):

    To clarify, the figure is from the AZ DHS site, which also included the note explaining the death certificate matching. And yes, AZ, like many states (and countries) consistently sees a death count drop off (though not a case count drop off) on Sundays and Mondays which is why Tuesdays are generally the highest. I also follow the Maricopa County site daily and it shows the same pattern.

    That is why I am waiting to see if the AZ and Maricopa data follow their usual pattern and drop off somewhat from today’s report which would mean a continuation of the slow increase in deaths.

    Both AZ DHS and Maricopa have a lot of very good data on their sites, having greatly improved since March.

    Please keep us updated.

     

    Today AZ deaths dropped from 117 to 36, which is about where the recent run rate for mortality has been. Maricopa dropped from 55 to 23. Because of the sheer volume of new cases, hospitalizations and ICU usage continue to increase which remains a concern.

    Are all the recent hospitalizations and ICU because of covid?  No

    People are seeking medical care for non-covid reasons.

    What is the median age of new cases?  How many are asymptomatic?

     

     

    • #36
  7. Full Size Tabby Member
    Full Size Tabby
    @FullSizeTabby

    MISTER BITCOIN (View Comment):

    Today AZ deaths dropped from 117 to 36, which is about where the recent run rate for mortality has been. Maricopa dropped from 55 to 23. Because of the sheer volume of new cases, hospitalizations and ICU usage continue to increase which remains a concern.

    Are all the recent hospitalizations and ICU because of covid? No

    People are seeking medical care for non-covid reasons.

    What is the median age of new cases? How many are asymptomatic?

     

    I was concerned a few days ago about rising hospitalization numbers here in Texas, but then found out that there may be some manipulation of the numbers and/or skewed presentation. Each week I seem to find more reasons not to trust numbers being presented by government officials. 

    Apparently at least some hospitals here in Texas are reporting as “Covid-19” patients any patient that tests positive for Covid-19, even if something else brought them to the hospital (like a car crash or a heart attack or some other medical issue) and Covid-19 is causing no significant health issues. Sorta like some of the problems with early death statistics. 

    Also, there is breathless reporting of hospitals “at 90% capacity,” presented so as to give the impression that there is an unprecedented shortage of beds. But, apparently many hospitals normally operate at 80 – 85% capacity so they can cover fixed costs and pay staff but still have room for emergencies. So they may be at higher than normal capacity, but not as dramatically as the headlines might lead the reader to believe. Many of the people in hospitals are people who would normally be there for heart surgeries, cancer treatments, joint replacements, etc., most of which they are still trying to catch up on because such treatments were prevented for two or three months. 

    • #37
  8. The Reticulator Member
    The Reticulator
    @TheReticulator

    Full Size Tabby (View Comment):
    I was concerned a few days ago about rising hospitalization numbers here in Texas, but then found out that there may be some manipulation of the numbers and/or skewed presentation. Each week I seem to find more reasons not to trust numbers being presented by government officials.

    A simple, two-part rule covers most of the cases: 1) If the information comes from the government, don’t trust it. 2) If the information doesn’t come from the government, don’t trust it.

    • #38
  9. Steven Seward Member
    Steven Seward
    @StevenSeward

    Full Size Tabby (View Comment):

    Apparently at least some hospitals here in Texas are reporting as “Covid-19” patients any patient that tests positive for Covid-19, even if something else brought them to the hospital (like a car crash or a heart attack or some other medical issue) and Covid-19 is causing no significant health issues. Sorta like some of the problems with early death statistics.

    If this is anything like my wife’s hospital in Cleveland, they are now testing for Covid, every single person who is admitted to the hospital for other reasons.  This was not done before because test kits were scarce.  Now they are abundant.  They are catching a  percentage of cases that would have gone unnoticed before because the vast majority of these patients had no Covid symptoms.  This is certainly part of the recent rise in number of cases without the corresponding number of deaths, since these are mostly mild cases.  I don’t know what percentage of the new cases this makes up, but I would guess that hospitals everywhere are doing the same thing.  This is a variant of the “random testing” that everybody was hoping for two or three months ago, and the logical result is more cases.

    • #39
  10. MISTER BITCOIN Inactive
    MISTER BITCOIN
    @MISTERBITCOIN

    Steven Seward (View Comment):

    Full Size Tabby (View Comment):

    Apparently at least some hospitals here in Texas are reporting as “Covid-19” patients any patient that tests positive for Covid-19, even if something else brought them to the hospital (like a car crash or a heart attack or some other medical issue) and Covid-19 is causing no significant health issues. Sorta like some of the problems with early death statistics.

    If this is anything like my wife’s hospital in Cleveland, they are now testing for Covid, every single person who is admitted to the hospital for other reasons. This was not done before because test kits were scarce. Now they are abundant. They are catching a percentage of cases that would have gone unnoticed before because the vast majority of these patients had no Covid symptoms. This is certainly part of the recent rise in number of cases without the corresponding number of deaths, since these are mostly mild cases. I don’t know what percentage of the new cases this makes up, but I would guess that hospitals everywhere are doing the same thing. This is a variant of the “random testing” that everybody was hoping for two or three months ago, and the logical result is more cases.

    I think almost half of cases are asymptomatic

     

    • #40
  11. Gumby Mark (R-Meth Lab of Demo… Coolidge
    Gumby Mark (R-Meth Lab of Demo…
    @GumbyMark

    Steven Seward (View Comment):

    Full Size Tabby (View Comment):

    Apparently at least some hospitals here in Texas are reporting as “Covid-19” patients any patient that tests positive for Covid-19, even if something else brought them to the hospital (like a car crash or a heart attack or some other medical issue) and Covid-19 is causing no significant health issues. Sorta like some of the problems with early death statistics.

    If this is anything like my wife’s hospital in Cleveland, they are now testing for Covid, every single person who is admitted to the hospital for other reasons. This was not done before because test kits were scarce. Now they are abundant. They are catching a percentage of cases that would have gone unnoticed before because the vast majority of these patients had no Covid symptoms. This is certainly part of the recent rise in number of cases without the corresponding number of deaths, since these are mostly mild cases. I don’t know what percentage of the new cases this makes up, but I would guess that hospitals everywhere are doing the same thing. This is a variant of the “random testing” that everybody was hoping for two or three months ago, and the logical result is more cases.

    I think there is some of this happening but the degree to which it is differs from state to state and even from metro area to metro area.  Here in the Phoenix area, the number of  hospitalizations for covid is going up, as is mortality, though the rates are going down, because the volume of cases has increased by so much. EDs are now sending home covid patients who would have been hospitalized a month ago because they need to keep beds available for the sickest covid patients.  

    • #41
Become a member to join the conversation. Or sign in if you're already a member.