Trust the experts. Believe in science. These are literal mantras that are in the air today. But I have to be frank and admit that I do not trust the experts on everything. If I have a lump in my testicle, I’ll trust the experts. But when it comes to rare events, the experts seem to suffer from some biases. When it comes to COVID-19 I’m very skeptical of experts. The reason I’m skeptical of experts is many of them didn’t anticipate the coming pandemic when they really should have.
Looking back at my record, it’s mixed. I spent much of the second half of January trying to get my wife to stop telling me COVID horror stories, as she was tracking Wuhan-related social media. Looking at the evidence we did have on hand, it looked scary. But, previous worries had been unfounded, and the idea that the pandemic would get out of control was not something I could rationally refute, but it was something I really wished would just go away.
On January 26th I mentioned coronavirus, but tried to diminish the downside risk. On the 2nd of February, I admitted I didn’t think that COVID-19 would be a big deal in a few years. But…things were changing. After talking to Greg Cochran, I changed my priors. In the first week of February we went and stocked up and I withdrew some cash (the infamous toilet paper run!). The episode of The Insight on coronavirus in February was recorded on the 16th. You can tell, I think, that I was already becoming seriously alarmed. I had expressed worries in the last week of January to co-workers but was dismissed as an alarmist. So in general I kept my opinions to myself, though on the 18th of February I took my daughter to the park and talked her friend’s dad’s ear off about coronavirus (he texted us in March to thank us for putting the pandemic on his radar). On February 20th I had my last meal indoors outside my home for 2020.
I remember February 19th, because that was when COVID-19 broke out in Iran, and I flipped my shit. I nearly didn’t go to the lunch appointment I had the next day. I couldn’t hope that there was something special about China that made it vulnerable anymore. The pandemic was going to happen, I was 100% sure then. I was very angry when Donald Trump visited India on the 24th and 25th. On the 24th my friend Default Friend asked if she should be worried, and I admitted my hysterical level of alarm (many scientists have privately messaged me and said that this was their wake-up call, as they were prompted to do research after my public alarmism).
Where were the experts? Well, I can’t survey them, but Stat News has done a great job covering COVID-19. I used the Wayback Machine to look at its front page all through February. On the following dates, COVID-19 was not their prime feature story: 3, 6, 7, 8, 10, 11, 12, 15, 19, 20, 24, 25, 26. After the 19th, when major figures in Iran seem to have caught COVID-19, I have no understanding why Stat did not pivot in totality to COVID-19.
The point is that when it comes to your own health you should keep your own counsel. Many of the public health experts were relatively sanguine deep into February. They excused Black Lives Matter protesters because of ideological affinity. They were against masks before they were for them. They talked about how border controls don’t work, but later admitted they had just asserted something they wanted to be true (?). These are people who I don’t trust because they show that their concern is not with facts, but outcomes. Social outcomes.
On the topic of public health experts, saw Prof Balloux retweet Nate Silver yesterday on why certain current vaccine distribution proposal don’t make sense, with a comment that he never thought he’d see the day where he agreed more with someone from outside epidemiology than within epidemiology… Seems to have been deleted from the timeline. Instead retweeted someone from within public health that has made the argument against the “modest proposals”. Even the good guys have their limits, I guess (and he has been one of the good guys through it, to my eyes).
On vaccine distribution again, we have this Pennsylvania professor claiming that vaccine distribution should be ethnically restricted and deprioritized towards White Americans as an opportunity to “level the playing field a bit”. A euphamism for having them die younger.
The claim he also made at the same time being that “Older populations are whiter (because) (s)ociety is structured in a way that enables them to live longer.”. While of course, this is because of migration and demographic transition dynamics (Latin and Asian migrants are more recent and younger, Black Americans had somewhat higher numbers of children until convergence). Never mind that Latino Americans and Asian Americans have longer life expectancies than White, and particularly longer life expectancies adjusted for poverty (poverty adjusted gap between White-Latino may be larger than same White-Black!).
This isn’t the same argument made, exactly, in recent days, by the US CDC for vaccine distribution, but there does seem to be a kinship of ideas.
I wonder how if this was how people who voted for Biden in your election wanted to start of a proposed return to normality. You guys in the US are in the midst of a year-on-year decline in White relative life expectancy. This is the “Deaths of Despair”. While I am not the most sympathetic necessarily to the reasons for these deaths (and I should put that out there), it seems like public health may be putting systematically people into power who are unlikely for reasons of ideological disposition to actually take any ameliorative action around this. (Indeed, it may be very dark to think of it but perhaps they may look on this as a favourable decline in “privilege” and as a consequence of “White Fragility” that requires no action).
People may well die, and the ideology may well have life and death consequences. (Far from claims that this is all just student politics froth with no consequence, no power.)
On a broader topic, it seems like the US needs some kind of legal reform in healthcare. I’m not totally convinced whether or not you guys are getting more or less stuff or less effective stuff from your healthcare per unit of expenditure, or that expenditure is out of control for your income levels… but the billing and pricing system you have seems complicated and crazy (where as described, if perhaps sensationalized, it seems to be almost a lottery or a Kafka-esque system in how the pricing relates to distribution of benefits). That’s something which the incoming US administration probably needs to do. (I may be wrong as very distant from experiencing your system, but that’s how it looks, broadly).
Now, is anyone going to trust people on the same political “side” as those who openly prioritise these sorts of ideological goals with that sort of reform? Are they gonna be able to work across the aisle to do it?
Matt, here is an explanation for why health care costs so much in the U.S:
https://marginalrevolution.com/marginalrevolution/2020/02/random-critical-analysis-on-health-care.html
As for why it’s not very effective, I think you need to get Hansonian: health care is not about health. For centuries doctors didn’t really have effective treatments for basically anything and were in fact worse than doing nothing, but people went to them anyway.
SARs 2003, H1N1 2009, MERS 2012, CCP virus 2020. They need to prepare for back to back pandemics.
@TGGP, thanks, always appreciate your opinion, though yeah, I’ve read that analysis – he’s sort of who I’m alluding to when I say it’s credible that Americans get about as much and as much quality healthcare, in aggregate at a national level, as expected for their national income level. (There seems really like there isn’t anyone else who even tries to make a well founded case for US expenditures being roughly in line with expectation!). And readily agree that higher expenditures on healthcare seems to happen with rising country income despite giving increasingly low returns.
But I’ve never really understood RCA’s claim that the US is roughly as redistributive across varying income levels as the rest of the developed world, when it comes to the payment of its system’s costs.
Take for example, this claim by Saez and Zucman, offered up for the UK in the Guardian – https://www.theguardian.com/commentisfree/2019/oct/25/medicare-for-all-taxes-saez-zucman): “The main difference between the insurance premiums currently paid by American workers and the taxes paid by workers in other countries is that taxes are based on ability to pay. The income tax has a rate that rises with income. Payroll taxes are proportional to income, at least up to a limit. Insurance premiums, by contrast, are not based on ability to pay. They are a fixed amount per covered worker and only depend on age and the number of family members covered. “.
Is that simply untrue (effectively a lie to a UK audience who know no better!), or explained by how much US healthcare insurance is really employment linked or via Medicare, or is there something else I’m not getting?
You’ve also got things like this: https://www.verywellhealth.com/why-an-ambulance-costs-so-much-4093846 – where charges written described here as routinely well in excess of collections and costs, and how much is actually collected seems to vary wildly from circumstance to circumstance. (And there’s another example of crazy billing practices linked by Tabarrok at your link there – https://www.vox.com/health-care/2018/12/18/18134825/emergency-room-bills-health-care-costs-america).
If that’s an accurate summary, that seems potentially very inequitable, if there is in practice somewhat of a lottery in who pays and who doesn’t. Also massively distorting to public views of real costs and their ideas of healthcare’s profit margins, to the direct detriment of confidence in the system. If people think healthcare is way more profitable than it is, and it becomes very much a rallying point of these ideas that a corrupt and exploitative business elite is bleeding people dry for profit (both to critics of the US inside and outside the US!). Seems like you would want to fix that?
Again, yeah, I don’t really have that much problem with finding RCA persuasive at an aggregate, country level of per capita expenditures and the per capita amount and quality of procedures and drugs etc that the US gets for its expenditures. It’s the “Who pays?” bit where he starts to lose me a bit.
The only thing I am confident of about this pandemic is that it will be years before we know what happened. The demographic impact of the disease is so focused that it is entirely possible that total deaths will drop below trend for the next couple of years.
Comparing this pandemic to historical pandemics shows that we have taken extraordinary measures to thwart a disease whose impact is really quite limited.
Forget the Black Death, and the waves of Eurasian diseases that rolled through the Americas in the 16t and 17th Centuries destroying ancient civilizations. The 1918 Flu Pandemic killed about 650,000 Americans out of a population one third our current size. Many of them were otherwise healthy prime age adults and very young children. https://www.cdc.gov/flu/pandemic-resources/1918-commemoration/1918-pandemic-history.htm
In the current pandemic less than .05% of the deaths have been children under 18 https://data.cdc.gov/NCHS/Provisional-COVID-19-Death-Counts-by-Sex-Age-and-S/9bhg-hcku
“The owl of Minerva spreads its wings only with the falling of the dusk.” G.W.F. Hegel
Talking about expertise, many deaths counted are totally unrelated to Covid-19. I recently heard from relatives about a case in which the victim of a car crash was counted as Covid-death! You can die “with Covid” and get counted, probably even with a knife in your back…
I didnt trust the experts when they played the pandemic down in 2019/2020, said its “just a flu”, and there were some here, from big scientific institutions, while I was alarmed early on, after reading first Chinese reports, and I don’t trust “experts” which demand measures which will just bring us closer to the totalitarian, global surveillance & control state “because of Covid” as they say.
There is always an agenda and what expert opinion being pushed or suppressed in a given context by the mass media and institutions varies, just like agendas shift.
If anything, the current situation should make that clear to anybody.
For many agendas you can find the “right expert” if you search and pay enough. Then you just need the communication channels to spread the message you want. In the last months we could see that often enough on various issues.
This is one of the many reasons why I do not trust the medical establishment and place no stock whatsoever in their recommendations. I believe in DIY medicine (see longecity.org as example). particularly with anti-aging life extension. As a DIY life extensionist, I’ve known since day one that I am out on my own when it comes to medicine and my long term personal future self. The COVID-19 business has exposed the corruption and ineptitude of essentially all large scale societal institutions. Something that I was vaguely aware of, but never aware of the scale. This is why my libertarian instincts have been reinforced a hundred times over. I realize that I and only I can ensure my personal future. I cannot rely on external institutions to do so.
I took cues from your tweets (and Balaji Srinivasan, the Silicon Valley guy). Sincere thanks for that! I managed to buy some *precious* PPE around Feb 20 based on the videos from China and Japan. PPE was very easily available on Amazon then. I timidly bought only two boxes (but have been reusing them). I wore it to grocery stores starting around Feb 20. Stocked up for about a week-long camping trip. I even filled up on gas, with plans to drive west sorta like Cormac McCarthy’s The Road. We let a musical ticket lapse in late Feb. In early Feb, I tried unsuccessfully to get some retiring relatives to sell off manufacturing stocks.
Somehow today, again things feel a lot like March 2020: the lack of caution (domestic air travel is at 40% of 2019 Xmas), the news from UK, etc.
Hope the vaccines work!
Matt, I find RCA’s and Tabarrok’s reasoning and graphs to be compelling, but they do not rebut your intuition about US health care. As one pending example, my daughter’s name is mis-spelled on a bill I received from a major lab testing corporation, so our insurer denied coverage. I’m on the hook for $316. My wife had a similar set of tests done: sticker price $341, insurance-contracted price $105, co-pay $21.
Insurers negotiate contracts as a steep discount from cash prices. So visiting oil sheikhs and the uninsured owe mind-blowing sums that bear little relationship to the costs of goods and services.
A somewhat free-market system with no price transparency. It speaks to a broader dysfunction that this small Gordian knot has yet to be cut.
@AMac76, thanks for the story. It seems that RCA may be onto something in rebutting the usual arguments made from plotting US GDP/capita against healthcare expenditure and then gesturing that somethings up, *but* that this just does not really rebut a lot of dysfunction in the system at the level of process.
…
General Covid related comment that seems to fit better here than in open thread: I’ve commented before about how the second wave of Covid-19 has hit Eastern Europe really hard. This seems to me to be an underreported story, still. However only by looking at individual countries, which could potentially be overscrutinising small countries.
So I took the OWiD coronavirus data and aggregated death figures, then worked out subregional figures for different European regions.
Plots here: https://imgur.com/a/4PuM87R
If you separate the EU member states into East and West, deaths/million in the East is about 85% of the West, up from around 10% in July, with the bulk of the converge in the period since 05/10/2020.
If you consider the regions of Europe as East EU, West EU, non-EU West and non-EU East (including Russia), then non-EU West is worst hit. Primarily for non-EU West being mainly constituted by United Kingdom, as still one of largest states which were badly hit (though there are of course many smaller states with higher rates). While EU West still has Germany as a large state that’s a success story relative to Europe (though Germany is now converging a bit as well).
Non-EU East is the most protected, *but* this is primarily due to Russia doing relatively well in this wave, if we believe their official line (and perhaps we do and perhaps we don’t).
If you remove Russia and simply count the remaining Eastern European states and Western Europe states, then in the past week the Eastern European states have collectively surpassed the Western states in death rate/million… Some convergence.
We’ll see if this so called new fast spreading strain changes the picture (even if that’s true as I hope not, it may not change the relative spread of the disease much, since it seems to be spreading across Europe).
Another visualisation of the trend to convergence of the epidemic across Europe- https://twitter.com/simongerman600/status/1341413280926101508
Very few, fairly small holdouts against the trend.
Iceland is one of those countries. Recent article on Iceland’s approach – https://www.nature.com/articles/d41586-020-03284-3
Although I’m a bit of a skeptic on most claims of capacity and strategy in outcomes (it seems less likely in many instances that “luck”), will hold out some hope that deCODE Genetics sheer testing capability is important in Icelandic relative success (compared to other plausible explanations of them simply being a relatively distant island, with a cultural kindred to Scandinavia and a homogenous high trust population).
If sequencing costs drop enough, a deCODE Genetics to cover every Iceland sized population might be an aspiration that could really help control infectious diseases without resorting to any PRC style coercive measures that are at odds with our political order (or the ineffective stay-at-home orders that don’t do much more than government advice).
Discussed the pandemic early on with some folks, I made the case that Iceland’s relatively low CFRs gave a truer picture of the epidemic’s IFRs, as they had the world’s best testing setup (and quite frankly it seems they still do despite improvement in larger countries). The counter generally “Oh, Iceland is still too early in the epidemic, and their epidemic too small and subject to randomness; their CFRs will rise as deaths catch up to cases. We’ll see it rise to well above 1% when their wave completes.”.
Well, Iceland have now crushed two waves of cases+deaths – completed waves – and have the highest testing capacity, and their CFR (OurWorldInData) is still only 0.5%. This is set against an estimate of 0.8% for Iceland’s age structure assuming an European average of 1% IFR – https://www.sciencedirect.com/science/article/pii/S0305750X20302977. As in the above article, latest paper by Steffanson’s team still puts true IFR at 0.3% (https://www.nejm.org/doi/full/10.1056/NEJMoa2026116). If you assume that Iceland are undercounting infections and not deaths by about 50% that’s kind of plausible, and still would be a lot less of an undercount that other nations, whose cases probably only capture 20-30% infections or really less if the IFR is lower than commonly thought.
(And on IFRs and public health experts again, those claims that Africa would face IFRs at something like 50% and South Asia 90% of the European IFRs, as here – https://gh.bmj.com/content/5/9/e003094 – are looking wrong. Morbidities and ‘health care system’ capacity don’t seem to be as important as the some PHEs thought they must be, in making their arguments for relatively young countries to impose lockdowns.)
As an addition to the above, where I’ve said cumulative Eastern European deaths per capita might be held down relative to cumulative Western European deaths if we consider Russia within Eastern European (against Eastern European deaths overtaking Western Europe, excluding Russia), note the following article today: https://medicalxpress.com/news/2020-12-russia-virus-cases-deaths.html
“They also registered 635 deaths, increasing total fatalities to 53,096 since the beginning of the pandemic.
Russia’s death rate is much lower than that of other badly hit countries, raising concerns that authorities could be downplaying the scale of the outbreak.
Data published by the country’s statistics service earlier this month indicated excess deaths of nearly 165,000 year-on-year between March and October, suggesting virus deaths could be much higher.”
Excess deaths in somewhere in the neck of 300% of official coronavirus deaths is, needless to say, probably quite a bit higher than the EU (where it is typically less than 50%), and would probably take death in Eastern Europe including Russia decisively above the West (and would make Russia generally comparable with the EU and the United States)…
More pandemic geekery; thought this might be interesting. The OurWorldInData file I have with Coronavirus deaths per million, cumulative up to 21/12, has the fraction of the population >65. So that is a decent test of whether countries are just protected by youth, adjusting the death rate by purely >65.
E.g. if a population has 2x the pop over 65, naively assume a 2x death rate, as a very simple demographic adjustment (of course risk accelerates even more than that with increasing age, but as a simplification)
Here’s how this compares with death rate/million: https://pasteboard.co/JGJPhwh.png
Then some plots of deaths/population>65: https://imgur.com/a/Dj6JPlk
Cross plotting deaths/population aged >65 against consumption level (as index of living standards) and average temperature.
One of the interesting things I see in those is that average temperature has now no correlation at all with this age adjusted death rate.
Caribbean, Central America, South America, West Asia, Arabian Peninsula, all hard hit for their age structure despite warm climates. South Asia, Mainland and Insular Southeast Asia all protected.
In other words, there is still a regional “Asian vs Western” pattern that holds that is not explained by temperature or age. While this is a simple age adjustment, the same would be true in a complex age adjustment (this is, if anything, favourable to young countries). Any ongoing success of SEA most likely cannot be chalked up to either climate or age, at this point (must be some other effect, policy, or an areal reduction in R0 through “dark matter” or something else).
Although climate explains things within countries and regions over time, it probably does not explain the divergence between Europe+Americas+West Asia+North Africa and Asia very much.
Because the age-adjustment from my above comment was a bit crude, here’s another way of doing age adjustment. Bommer and Vollmer published a set of estimated age-structured adjusted IFRs for countries, using study data and age structure (https://www.uni-goettingen.de/en/606540.html). These correlate pretty well (near perfectly) with the fraction>65 in OurWorldInData’s set: https://imgur.com/a/LSmmz8g
So we can then use these to predict IFR for all countries in OurWorldInData and then adjust the real record total death rate per million by this, to allow us to determine which countries have had large relative outbreaks and show up lower down the list of deaths/million due to being protected by young age structure. (E.g. “What would their death rate be with roughly US/EU age structure?”).
Result: https://imgur.com/a/wxiSazj
Using the IFR adjustment method doesn’t really vary very much from the simple “deaths per population>65” method (other than in overall magnitude, because >65 only a low fraction of total population), R^2= 0.965, and the countries which are adjusted largely the same.