When Will the Exploding Number of Cases Lead to More Deaths?


The number of Covid-19 cases is at a new pandemic high. Deaths will follow.

The 7-day average in cases peaked on April 9 at 34,756. The number of deaths per day peaked at 2,590 eight days later. 

To suggest a lag of about a week is wrong. 

Key Reasons the Lag isn't a Week

  1. States did not peak at the same time. Averaging the states together smoothed out the curve. 
  2. States had different reporting methods. Some states lumped probable deaths in with known deaths. 
  3. Some states purposely downplayed deaths. 

If is highly likely Florida artificially and purposely undercounted Covid deaths. 

The Tampa Bay Times reports Florida medical examiners were releasing coronavirus death data. The state made them stop.

When the medical examiners’ list was available, it showed more deaths than the state’s count.

The Palm Beach Post editorial says Gov. DeSantis still hiding crucial data on COVID-19 deaths from the public.

The number of cases are now too many to hide.

Fools' Analysis

The Wrong Way

Typical Reaction to Bad Analysis 

Federal Reopening Criteria

Distorted "Real World" Methodology Explained

Distorted View of the Real  World Methodology

The proper way to compare death rates is to pay heed to state-by-state variations and lag rates. 

Lag Time in Florida

New Cases va Deaths Florida 2020-06-26

It took over a month before deaths caught up to the original surge in Florida Covid-cases. 

Lag Time in New York

New Cases va Deaths New York  2020-06-26

New York had the smallest lag time. It was less than a week.

Lag Time in Arizona

New Cases va Deaths Arizona  2020-06-26

Arizona had so few cases that it is a stretch to point out a lag time. But an explosion of cases started at the end of May and on June 24 there was a meaningful new record high in the 7-day average number of deaths. 

Both of these lags are about three weeks. 

Lag Time in California

New Cases va Deaths California  2020-06-26

California is similar to Arizona. It has weak relationships as well, if not more so. But to dismiss the rising number of cases as meaningless is folly.

Lag Time in New Jersey

New Cases va Deaths New Jersey  2020-06-26

New Jersey is either a double top of two separate cases or a longer lag of about three weeks.

Texas Lag Time 

New Cases va Deaths Texas  2020-06-26

The initial lag time in Texas was a month. 

National Lag vs State Lags

  • Florida has a lag time of 5 weeks
  • Texas a lag time of 4-5 weeks
  • New Jersey a rolling top of 3 weeks. 
  • It is hard to say what the lag time is in California and Arizona. 
  • New York has a small lag of about a week.
  • Average all of this together you get a national lag time of 8 days. 

Comment From Nate Silver

Rise Not Due to More Testing

Stats That Allegedly Do Not Matter

Stats That Allegedly Do No Matter

Most of those are record highs. But hey, it cannot possibly matter.

Disbeliever Tactics

The disbelievers generally do not factor in that fact that lags can be a month or more.

Nor do they accept the proven fact that social distancing works.

We can debate the price tag but it is proven foolishness to suggest that social distancing does work. 

Governors in Florida, Texas, and Arizona casually disregarded and downplayed social distancing. 

So did mayors in various cities in California. And so did president Trump.

This Will Soon Go Away 

Please play that video for an amusing flashback.

Look what happened.

Panic in Texas and Other Responses

  1. Texas Shuts Bars at Noon as Covid Cases Surge
  2. Houston Exceeds ICU Capacity
  3. EU Lifts Its Travel Ban But Not For the US
  4. Florida Closes Bars in Response to Covid Spike and NBA Season Threatened
  5. Trump is Behind in Every Recent Battleground Poll and his Covid handling is one of the reasons.

Fake "Real World" vs Real "Real World"

In the fake "real world" we can ignore the exploding number of cases, just as Trump did in late February and early March. 

In the real "real world" things are just a bit more complicated. 


Comments (58)
No. 1-26

The average age of the person in the newer cases is dramatically lower, so 10,000 new cases may mean 2-5 deaths instead of 30 or 40 as in April and May.


Iran case graph profile matched the US. Their lag was 3 to 4 weeks. Lots of invulnerable young people exposed, deaths followed 3 to 4 weeks later. See the attached graph....


You cannot predict change in death rate without knowing the age of the new cases. If they are under 75, there will be no uptick in deaths. If they are asymptomatic with a positive antibody test, no uptick in deaths.

Just simple conditional probability. But people don't like math so more hysteria.

Said another way - you CANNOT predict future deaths with CONFIRMED CASE statistics if you don't adjust for (1) age; (2) age; (3) age; (4) in nursing home; (5) # of comorbidities.

A case on a person under 75 with no comorbids is 0.1% chance of death.



I want a chart of hospitalizations ve deaths over time but cannot find one



I said it a few weeks ago, and I’ll say it again: you’re slipping. I used to read you blog every day, now I rarely come by. You used to approach issues with a skeptical eye, but with anything virus related your approach lacks sound rational thought. Here’s but one example:

“The disbelievers do not factor in that fact that lags can be a month or more.

Nor do they accept the proven fact that social distancing works.”

  1. Grouping everyone who disagrees with you as a “disbeliever” is not only childish, but it paints a broad brush against people who might agree with you on some points, but has competing arguments on others.

  2. I think everyone should be wearing masks in indoor places, and crowded outdoor places. I believe that we’re not social distancing enough. That has zero to do with your first point, which happens to be wrong....

  3. Your third point is based on faulty reasoning. Look at a simple chart of cases to deaths. It lags, but the two lines have diverged considerably, more so than can be explained away by your conspiracy theory that deaths in Republican states are being hidden, or any of the other flimsy garbage you posted. Younger people are making up a larger percentage of cases in this round, as compared to NYC.


Mish, your chart of daily death has a spurious data point that doesn't belong there. It shows a jump to about 2400 a couple days ago that never happpened. The daily deaths the last few days have been:
June 22, 369
June 23, 871
June 24, 819
June 25, 653
June 26, 663

Oddly, there was a community post that claimed death had jumped over 2400 on Wednesday, and for documentation, he linked an article from the Washington Post. The article he linked was dated 4/25. Indeed, the deaths on Wednesday, 4/22, did top 2400, hitting 2409, but that's rather old news, and shouldn't be shows on the chart for this this week. (If you doubt that the Washington Post article was from 4/25, look at the URL.)

Edit - In further discussions, it appears that New Jersey went back and looked at all deaths since the start of the pandemic, and classified another 1900 deaths as "probable Covid deaths". The spike on your chart reflects those additional 1900 deaths. It's important to remember that they aren't current deaths, but rather, deaths from March-April that were retroactively classified as Covid.


The Russians and the Chinese are behind the latest increase in cases. I mean both Dems and Republicans can't be wrong right?


I suspect that deaths/overall cases will drop over time for a variety of reasons; including better treatment and more infections among younger people, where the death rate is lower.

Even with a lower death rate, there will still be many more deaths to come, over the next few months. Which is very sad as many of them could have been avoided.

Because the ideal way to reduce deaths is to reduce the spread of the virus. If there are no new infections, eventually there are no more deaths from the virus.

But the US is not reducing the spread, as it is currently hitting new record numbers of cases. Which is shocking because other developed countries have cut daily infections by over 90%. I don’t understand how the US is incapable of reducing the spread.

If you simply follow the task force recommendations of social distancing, masks, and sanitizing, you could reduce the spread like so many other countries have.

The only areas that need any sort of “lock down” are those where the health care system is overwhelmed. And you should be able to avoid any lock downs by simply following the recommendations.


What does it mean to test positive for covid? I know the potential illness/damage to ones body it can do but what about disclosure? Do you tell your friends/neighbors/co-workers/employers? And how will they treat you? If you are covid positive are you the new pariah?


The sad part about a lot of these comments is that it isn't understood that effects are not instantaneous. It's like turning asupertanker

The spike in cases is coming about a month after opening up.

And we all know that you don't die the instant you get CV.

The spike in deaths will be at least two to three weeks from now.

In one month from now we will all know for sure.

I will be rooting for no spike in deaths. But given the evidence from other places it is quite possible.


Mish, you are fortunate to have so many epidemiologists among your readership. /sarc

I wish I was as certain about anything as some of your readers seem to be about everything.

Keep up the good work. Illegitimi non carborundum .


This is why we can't have nice things....


My sincere apologies for a previous note wherein I disparaged tests. Please forgive me, I relent. Just today I heard that Dr. Fauci (I now use the Dr. in addition to the patronym or whatever he goes by) said he is about to institute a double test. The first will be a blanket test using a 'proven' methodology to, well, blanket the population (understandably the field grazed by the flocks), followed by the more trad swab test on those individuals 'captured' by the blanket effort. Oh! wonders! As my granddaughter says: two is more than one. Honey, two is better than one!


Mish, I would not worry till few "Hollywood Elites" crock of COVID-19 and,I read and confirm the report on TMZ, that when Meryll Creep,Robert D and few Pedophiles have final stages of COVID-19 in the Hollywood and come out and announce they are infected and are dying, that when I will start panicking, till then enjoy few shots of vodka.


Mish, that last sentence should end "does not work" (delete my comment after correction):

The disbelievers generally do not factor in that fact that lags can be a month or more.

Nor do they accept the proven fact that social distancing works.

We can debate the price tag but it is proven foolishness to suggest that social distancing does work.


In the particular case of Covid19, nobody seems to be paying attention.

If you believe that an infected individual becomes infectious after 2 days (3 days prior to showing symptoms) and if you believe the infection rate is 3 then the arithmetic looks something like this:

Day 1 - 1 infected person (we make abstraction here as to how this person got infected)
Day 3 - 4 infected persons (the original person infected + 3 new infected individuals)
Day 5 - 16 infected individuals (4 earlier infected individuals x 3 + 4 original individuals)
Day 7 - 64 (16 x 3 + 16)
Day 9 - 256
Day 11 - 1024

Given the assumptions, this progression gives you more than 80 million infected individuals in 31 days.

This is basic math.

If there is any debate to be had, it is whether or not this progression is correct. Anything else only leads to bickering and recriminations.

Is the arithmetical progression of the infection correct or not? Nothing else matters.

If the math is correct, then:

As we are finding out that this virus has been circulating since fall 2019 or even summer 2019 the math tells you a number of things.

The first thing you can infer is that, given the progression as outlined above, the virus is already well spread throughout the global population.

That being the case, the first realisation is that blanket confinements of the population 3 months or even 1 month after realising a new bug is on the scene is, arithmetically speaking, not very useful.

The second thing you can infer is that, given the progression as outlined above, the population must already have some degree of immunity. We can infer so by the fact that if the entire population was susceptible to this virus, we would have had significantly more hospitalisations and deaths; we are talking orders of magnitude here. This would have been particularly so aboard all the ships where Covid19 was detected. Yet, of all the ships identified as being infected with Covid19, the Diamond Princess is the one that reported the highest infection rate with just under 800 infections out of a population of 3500. But the more interesting case is highlighted by The Zaandam. The Zaandam was stranded at sea for a total of 23 days. During this time, first flu-like symptoms manifested on day 14. It is only on day 15 that the captain issued confinement orders. By the end of the whole sorry saga, 23 days had elapsed. 23 days during which the population of The Zaandam had been completely isolated, at sea, far from anyone anywhere in the world and where, during 14 days straight, passengers and crew mingled happily without a care in the world. Yet, The Zaandam presents us with infections of fewer than 100 out of a population of 2000 passengers and crew.... and 4 fatalities of which only 2 tested positive for Covid19.

The third thing we can infer from the above is that regardless of what you think of masks, given the arithmetical progression as above, wearing masks will change little.

Finally, although the spread of the virus may, to a degree, have been slowed by confining the population, three things should be glaringly obvious:

1 Confinement does not eradicate the virus. Confinement can merely slow down the contagion. So, albeit at a slower pace (i.e. Ro of 1.4), contagion progresses. Due to the compounding nature of contagion however, confining the population merely buys you a handful of days time as compared to not mandating blanker confinements. But progress the virus will nonetheless.

2 Wearing masks does not eradicate the virus. So, albeit at a slower pace, contagion progresses. Due to the compounding nature of contagion however, wearing masks merely buys you a handful of days time. But progress the virus will nonetheless.

3 Even assuming we could put out a vaccine today, it will be many months, if not years, till the entire global population can be vaccinated to a reasonable degree. In the meantime, the arithmetical progression as outlined above, continues.

So, unless you can dispute the arithmetical progression as outlined above, confinements and mask wearing may make you feel good. The reality however is that it makes little difference.

So the entire debate should hinge on whether we do or do not have the arithmetical progression right.

Everything else only results in unnecessary bickering and recriminations.

So, can we or can we not agree on the arithmetical progression of the infection?


I echo the other commenter who critiqued your use of the term 'disbeliever.' As you yourself know better than most - because it is your skillset - sorting through the data available is difficult. The standards for such data are all over the map. You say CV19 is under-reported, many others say it is over-reported. These things can go round and round forever.

The root problem that the term 'disbelief' correctly indicates is that it all DOES come down to belief. Again, given that precise definitions for what constitutes a covid death are lacking and so counting criteria vary widely. Efficacy of tests vary widely as well (there is occasional reference in the literature about how the false positive rate is far higher than generally admitted, and now also occasional reports about how there may be far more widespread immunity already in the population from previous covid episodes - which may explain in part Asia's greater success numbers-wise along with being prepared after the last SARS whereas we were not.

In theory, we should have excellent data so we wouldn't have to throw around such terms as 'disbeliever' and 'believer' but as it is, that's what we are left with. One of the greatest failures of Team Fauxi, imo.


@Mish: Isn't this a typo?

" it is proven foolishness to suggest that social distancing does [NOT] work. "


The view from Ventura County:


The only death the Federal government seems to be worried about is the financial system.


It is obvious that masks and social distancing help.

There is little-to-no price to using masks in crowded spaces <period>.

One can make the argument that the price of social distancing is too high as younger people as up to 40s are hardly impacted.

However, there are social activities which are less critical. I'd argue that we can live w/o movie theaters and bars. Schooling of pre-K and elementary is quite critical. Middle and high school could manage online - maybe with one day per week attending the actual school facility in person. This might not suffice for kids with poor conditions at home, though. Etc.


A four week lag seems right. Even if this wave has more young people, they will infect their parents and grandparents soon enough. Family spread is something that all successful countries dealt with and the US is doing nothing so family spread will blow out the death rate once again.

Studies will be done on long term health problems from covid for survivors.


Mish, saw this at CNBC today and wow is it damning for red states that ignored warnings:

Notice that every state in the sharp increase category is a red state and the the blue states that did impose sharp restrictions are all now contributing only a small fraction of the uptick.


The "We can debate" sentence is missing a neagation.

Global Economics