Streetwise Professor

June 30, 2020

Hide the Decline! Coronavirus Edition

Filed under: CoronaCrisis,Politics — cpirrong @ 11:03 am

Reported Covid-19 daily deaths (likely exaggerated for reasons I posted on months ago) have been declining inexorably since their peak, and are now about 10 percent of the maximum. Even that is overstated because of backdating in states like Delaware and New Jersey that resulted in large single day death reports that summed deaths that had occurred over periods of weeks or months.

We can’t have that, can we? How are we going to sustain the panic, as politically useful as it is, if we report good news?

No problem–just switch the metric! Whereas for weeks we were told the Grim Death Toll narrative, that has disappeared down the Memory Hole, to be replaced by the Skyrocketing Cases narrative, especially in Red States such as Texas, Florida, and Arizona.

I have been calling bullshit on the case numbers as a meaningful metric since March. It’s even more bullshit now.

A major reason case numbers (i.e., the number of positive tests) are BS is that testing is not a random process, but is endogenous. Moreover–and this is crucial now–the process driving who is tested is changing. Whereas before tests were focused on the symptomatic or the particularly vulnerable, testing is now more widespread. Some companies are requiring employees to be tested in order to return to work, or to remain at work if they have the sniffles.

As a result, more people are testing positive. Moreover, the average age of those testing positive has declined dramatically (because they were censored from the test population before). Most of those people are symptomatic, and those who are experience mild symptoms. Those under 60 exhibit little risk of death, or serious illness (especially if they do not have other serious health conditions). Those who are sick enough to require hospitalization are less likely to require ICU care, and those who do tend to recover at high rates (without ventilation), and have relatively short stays.

As a result, there has been a striking divergence between rising case numbers, and deaths.

But that doesn’t fit certain political needs. So we hear virtually nothing about deaths, but only shrieking about case numbers. This exploits the earlier misconception–misinformation, actually–that the death rate from the virus is high. Indeed, as positive tests accumulate, and serological studies accumulate, it is clear that the infection death rate is in the range of .1-.25 percent, far smaller than the earlier estimates that remain embedded in the memories of most.

The shrieking is particularly intense in–and at–Texas. Yes, Houston has seen a large increase in positive cases. But the deaths in Texas (and Houston) have never been large (up until now 2020 has seen fewer pneumonia-related deaths than in the typical year), and are not trending up . Not that you’d know from reading the media.

So there has been a reprise of the overwhelming the health care system narrative.

The worst sinner at this is my local POS newspaper, the Chronical. This article in particular, which insinuated that Governor Abbott (and no, I’m not a fan) had coerced Houston hospitals into covering up impending doom.

The article starts out with a lie, claiming that Houston ICU utilization had hit 100 percent. Actual data show this did not happen, and that Houston ICU utilization has been fairly constant over since April. Even throwing around scare numbers about 90 plus percent utilization is misleading. Of course hospital facilities are sized so that they do not have persistent unutilized capacity. That is wasteful, and inflates costs. As the data in the link show, moreover, hospitals–rationally–have the ability to expand capacity.

As I said in a very early post, capacity is not a destiny–it is a choice.

Yes, there as an increasing number of Covid patients in ICU. But this is clearly another manifestation of changing testing protocols, and most importantly, of the same problem that makes even the death data meaningless: lumping people in the hospital with Covid together with those who are hospitalized because of Covid. If the increased Covid numbers were there because of Covid, you would see ICU usage go up overall. You don’t. It’s oscillating around normal levels.

It should also be noted that there are reasons to believe that people who should have gone to ICUs, or to hospitals, did not because of Covid. This suppressed numbers and makes it dubious to attribute any increase in utilization to Covid.

As for the supposed coverup, the hospital systems did not stop reporting hospitalization/ICU data, but the projections of future usage.

The outrage! Yeah, because Covid-related predictions have always been spot on, right?

In fact, the only competition between projections is which is the most absurd.

The POS Chronical’s political agenda and utter hypocrisy is on full display:

Then, after reporting numerous charts and graphs almost daily for three months, the organization posted no updates until around 9 p.m. Saturday, sowing confusion about the hospitals’ ability to withstand a massive spike in cases that has followed Gov. Greg Abbott’s May decisions to lift restrictions intended to slow the virus.

Gee. What else happened around the same time. Let me think? Protests ring a bell?

But of course, the protests (and the massive George Floyd funeral) are sacrosanct, and out of respect the virus took a holiday and didn’t exploit the conditions (large crowds) that are supposedly the main source of contagion. (EG., MLB will be restarting–but without fans, because otherwise Minute Maid and other parks would be Covid Central.

The hysteria over case numbers reminds me of a phrase from the Climategate emails: “Hide the decline!” Just as in Climategate, there was a divergence between a number that mattered (actual temperature) and a bogus number (proxy data-based temperatures): actual temps were flat/declining when the proxy number was going up inexorably.

So the battlecry became: Hide the decline!

We are seeing the same thing now. Hide the decline in deaths by hyping irrelevant case numbers, misinterpreting those numbers, and making dire forecasts at odds with the actual data.

If you will recall, the entire justification for lockdowns was to “flatten the curve” to protect the healthcare system. The underlying rationale was that the virus’s spread was inevitable, but we need to control the rate. That is, suppression/elimination was an impossibility until herd immunity was achieved.

Based on that rationale, the surge in cases with low and arguably declining numbers of deaths and no data demonstrating an overwhelmed healthcare system is actually a good thing. It measures progress to herd immunity. Moreover, it’s better to have the spread now, in the summer, rather than when the flu season kicks in, and creates its inevitable increase in demand for healthcare resources–including ICU beds.

But the media and many politicians are completely invested in panic, for malign and dishonest reasons. So it is essential to hide the decline, and hype the spike.

Print Friendly, PDF & Email


  1. Not only would I pay only minor attention to case data, I wouldn’t pay too much attention to deaths purportedly caused by COVID, given the inconsistencies inevitable in attributing Cause of Death in old, ill patients. It’s probably best to look at excess deaths. Someday people may look carefully through that data trying to identify contributions, both positive and negative, to the total.

    “Someday” reminds me that we may yet be guilty of a rush to judgement – we’ll know more (if we are allowed to) in five years time.

    Meantime a well ordered polity would encourage those at low risk to resume economic activity and enjoy summer weather outdoors. If the Powers That Be have sacrificed all trust then a new pandemic, or a Spanish-Flu-like resurgence of this one, would be harder to manage.

    Two things I’ve learnt. (i) As I expected from early days, epidemiological modelling is essentially an intellectual fraud. (ii) Virology is a fluffier science than I had expected. It seems to be miles behind bacteriology. “The Science” that was apparently to be trusted is not very trustworthy.

    It might make sense to move research funding from (i) to (ii). And to hang a few of the bad’uns, if only metaphorically.

    Comment by dearieme — June 30, 2020 @ 11:42 am

  2. The English experience.

    Comment by dearieme — June 30, 2020 @ 12:12 pm

  3. @dearieme – I’m for hanging a few of the bad ‘uns literally, pour encourage les autre.

    Comment by Christopher Hunt — June 30, 2020 @ 4:20 pm

  4. The head doctor in charge of Vienna’s corona hospital, has recently said in a TV interview that mortality for ICU patients has gone down from roughly 50 percent in march to 10 percent now, due to improved treatment. This should have a significant effect on the IFR.

    Comment by viennacapitalist — July 1, 2020 @ 1:24 am

  5. Where are you getting your numbers, prof? Using 7 day moving average (because of the wild week-end swings in reported numbers) give a current daily deaths of around 600 or so and a peak of 2,200 for the US.

    But fundamentally I agree that the way numbers are reported are disingenuous (your suspicion I guess?) or ignorant/overexcited (my guess). The count of the number of cases is particularly bogus – these daily numbers are not even comparable for a single country over 2 or 3 days never mind between countries or over a longer period.

    The number of deaths is similarly inconsistent across time and countries because wildly different recording mythologies. But at least it can be compared with excess death statistics to get some idea of whether they make any sense or not. I’ve found it interesting to compare death stats using with reported corona virus deaths and there are some serious discrepancies – for example, it looks like the final tally for the UK to be significantly higher unless there is another unknown disease killing people off this year. In comparison, Ireland’s excess mortality has been close to 5 year averaged despite reporting significant number of coronavirus deaths. Unfortunately this normalised data is only available for European countries but there’s already plenty to puzzle over.

    Comment by derriz — July 1, 2020 @ 6:56 am

  6. “Whereas before tests were focused on the symptomatic or the particularly vulnerable, testing is now more widespread. Some companies are requiring employees to be tested in order to return to work, or to remain at work if they have the sniffles.

    As a result, more people are testing positive.”

    That doesn’t make any sense. More healthy people getting tested should lead to LESS positive tests.

    Comment by lib — July 1, 2020 @ 7:18 am

  7. @lib
    It makes perfect sense. If you test more people you will get more positive results. You will get a lower proportion of positive results because you have changed the cohort from the vulnerable to the less vulnerable.
    Some patients who have had covid have been tested after recovery and show no sign of antibodies, indicating (to no one’s surprise) that the human body has a multiplicity of means to overcome infections.

    Comment by philip — July 1, 2020 @ 8:06 am

  8. Facebook’s USP is to track internet chatter to target ads. Many other companies offer similar services.
    Indeed, the WHO employs crawler bots to monitor disease outbreaks by searching for key words in internet chats. Particularly useful in countries where data collection is patchy or unreliable.
    Such a case is China and it beggars belief that the WHO was not monitoring traffic. So they were almost certainly aware of an impending epidemic back in2019, even before Taiwan blew the whistle.
    The WHO may be even more corrupt than we thought.

    Comment by philip — July 1, 2020 @ 8:14 am

  9. @philip – agreed. I assumed that the author meant proportionally more positive results…. but maybe I misread it!

    Comment by lib — July 1, 2020 @ 10:05 am

  10. In any case, the proportion of positive test readings in Texas is trending upwards, so you have more positive readings on an absolute and relative basis.

    Comment by lib — July 1, 2020 @ 10:11 am

  11. “mortality for ICU patients has gone down from roughly 50 percent in march to 10 percent now”: good day, Vienna.

    “due to improved treatment”: doctors always say that irrespective of its truth content. Hell, they claim all credit for the great fall in heart attacks in middle aged men, even though the decline long predates any change in medical practice. I dare say that there is improved treatment. Hurray! I challenge anyone to demonstrate that it is solely – or even principally – the cause of a five-fold decline. Maybe he means “we’ve stopped killing them with ventilators now”.

    Comment by dearieme — July 1, 2020 @ 11:06 am

  12. @dearieme
    Don’t get me started on ventilators!
    What, btw, are they going to do with the excess? There must be millions of these lethal surplus gadgets in storage.

    Comment by philip — July 1, 2020 @ 11:57 am

  13. @ dearieme
    It would be a sad state of affairs for medical science, if there were no lessons learned over time – which is why the “constant integral under the curve” agument never made sense to me.

    Although, I want to point out that his is anectodal evidence due to the recently low case count in Austria. However, I have seen (tentative) analysis based on US data supporting a decrease of this magnitude.

    …“we’ve stopped killing them with ventilators now”…

    the doctor admitted as much. The main reason being, if understood correctly, that the lungs are compliant far longer(that’s the term he used) compared to pneumonia with similar symptoms, i.e.intubation not indicated at all, in most cases – a big insight.

    Further, he pointed out that Remdesivir is of great help, as it significantly reduces hospital stay (mortality somewhat). They also know better what risk factors to look out for.

    I personally believe better Vitamin D provision in the northern hemisphere must play a major role in reducing case severity, as some studies suggest.

    Comment by viennacapitalist — July 1, 2020 @ 12:04 pm

  14. Were any random testings made? I remember the news that someone in the administration proposed it, but I didn’t hear anything about this since.

    Comment by mmt — July 1, 2020 @ 12:29 pm

  15. Just want to take a few moments to highlight Alberta’s rather good handling of this pandemic. We didn’t overreact and go to an absurd level of lockdown, sure indoor social gathering spaces were hardest hit, but I could go to the hardware store, get curb side pickup from lots of stores and he’ll I bought a house and had movers pack and move us during the height of our “lockdown”. Bars and restaurants opened up again last month. Outdoor seated events can have up to 200 attendees. Our testing capacity is high and testing is open to any who want it for the last month or so.

    Yes our overall daily numbers are up but Alberta’s chief medical officer continues to beat journalists over the head about not reading too much into simple positive case numbers. In fact the main stats they are looking at for the reopening is hospitalization rates and ICU admissions.

    They have been clear that this virus is not going anywhere and the response has all been risk based not kneejerk based. The initial measures were about ensuring capacity in hospitals incase things went to shut like Italy. This didn’t happen because most people being sensible and following the fairly simple guidance of stay home if sick, increased hygiene around high touch surfaces and avoid crowded spaces.

    So the message hasn’t changed, as long as the healthcare system has capacity and the numbers stay steady get out and enjoy our all to short a summer.

    Oh and the province also has distributed masks for any that want them for those unavoidable situations where a mask is sensible (oh yea mask wearing isn’t mandatory because that would be stupid)

    Oh and from day one they were very open with the data

    Oh and a story that will surprise no one, incentives matter.

    Off to enjoy a well earned Canada Day beer or three.

    Comment by Steve — July 1, 2020 @ 1:06 pm

  16. @mmt. It’s scandalous–there haven’t been any random tests, and in particular no random serological tests. I said very early (in February, if memory serves) that this should be priority one. And it still hasn’t happened, at either the Federal or state level. Appalling.

    Comment by cpirrong — July 1, 2020 @ 3:10 pm

  17. @Steve–Thanks for the update. Glad that someone was reasonable. Happy Canada Day. Have a beer for me, eh.

    Comment by cpirrong — July 1, 2020 @ 3:12 pm

  18. @phillip–Very early on I lamented that there wasn’t a ventilator futures contract I could short.

    Comment by cpirrong — July 1, 2020 @ 3:13 pm

  19. Re: Texas (and probably also more generally). You won’t see articles like this in the mainstream media.

    Comment by ColoComment — July 1, 2020 @ 4:55 pm

  20. @ mmt
    @ prof
    There have been plenty of random serological tests all over Europe, about 2 months ago. The most comprehensive one was conducted in Spain with a sample size of 60k with a disappointingly (from the viewpoint of herd immunity) low prevalence of about 5 percent, if memory serves well. Even in Sweden, if I remember correctly, it was quite low (About 10 percent).
    I think that’s the reason random serological testing has gone out of fashion.

    Comment by viennacapitalist — July 2, 2020 @ 3:39 am

  21. “I personally believe better Vitamin D provision in the northern hemisphere must play a major role in reducing case severity, as some studies suggest”

    I’ve been taking those pills for months now. Sometimes the supermarkets are out of stock so it seems I am not alone. My decision was based on the imbalance between “these might help” and “there’s no sign at all of any nasty side-effects”. The latter isn’t surprising if all I’m doing is raising my level to what it would have been in the past when I spent plenty of time in the summer out in the sun.

    I used to commute by bike: I suspect that even two daily cycle rides in the sun were enough for me to synthesise plenty of Vit D in summer. Not in our grey winters though.

    Comment by dearieme — July 2, 2020 @ 5:45 am

  22. Meanwhile, elsewhere in the US, de Santis is accused of manipulating the data down to make things look better than they are. These Red states eh?

    As for blaming the BLM protests for the spike, well didn’t you all want the lockdowns to be eased way back? You can’t have it both ways. Also, if you didn’t want people to engage in mass protests, perhaps you ought to reign in elements of your police force? Just a thought.

    @Vienna – any idea what the current IFR is thought to be? If, as is suggested here, the US should still pursue a goal of herd immunity, you still could be looking at high hundreds of thousands of deaths, assuming 60%+ of people need to be exposed to it in order to achieve this.

    The fact remains that we are still at the early stages of this pandemic and there is much to learn.

    Comment by David Mercer — July 2, 2020 @ 5:56 am

  23. @ David
    Dr. Streeck who analyzed the entire Population in one of the first german clusters (Heinsberg) has found an IFR of 0.37
    Alternetively there is this from nature summarzing serological studies

    However, this might be overestimated in case of better Treatment/ Vit. D as per above

    Comment by viennacapitalist — July 2, 2020 @ 7:34 am

  24. OK Prof
    I’ll see your “lamented that there wasn’t a ventilator futures contract I could short.”
    and raise you
    Funding will be cut off before a vaccine goes into general use.

    Comment by philip — July 2, 2020 @ 9:14 pm

  25. @ David Mercer, “ perhaps you ought to reign in elements of your police force? Just a thought.”

    What an extraordinarily mindless comment.

    Comment by Pat Frank — July 3, 2020 @ 1:01 am

  26. Just to say, “Hide the decline” referred to truncation of a tree-ring time series so-called proxy for temperature because it inconveniently went down after 1960, while the air temperature record went up.

    So, they chopped it off there, and grafted the temperature record on, and then smoothed to hide the join. Hiding the decline QED.

    Steve McIntyre at has thoroughly and analytically documented the deceit. No one has ever paid a price for the worst crime a scientist/sarc can commit.

    Comment by Pat Frank — July 3, 2020 @ 1:09 am

  27. “No one has ever paid a price …”: one of the few consolations if the Arsehole Tendency manages to bring down our civilisation is that it will presumably be possible for individuals to wreak their own revenge on the scientific scammers.

    Comment by dearieme — July 3, 2020 @ 5:53 am

RSS feed for comments on this post. TrackBack URI

Leave a comment

Powered by WordPress