COVID-19: Why US situation is far worse than Europe's, and why this may not be immediately evident
TL;DR: In assessing relative risk status, the future must be considered, not simply the present.
An HN thread[0] discusses whether the US or Europe are experiencing a worse Covid situation. The question contains nuances and pitfalls, though the general answer seems to be:
- The EU's situation is generally several weeks advanced relative to the US. As with the Jan--Mar 2020 interval, situations in different regions can be generally considered as time-shifts of one another rather than distinct dynamics.
- Instant measures of current case or death rates fail to account for built-in and likely future impacts and risks. Ignoring these is a category error, though a common one.
- The European daily trends are slowing or reversing. US trends are accelerating. The US future looks far bleaker than the European future. This contrasts with the blinding bias of considering only immediate present measures such as daily mortality.
Covid and population here come from Worldometers.
The thread begins with Aperocky's comment asserting, correctly, that "The worst hit place right now is the ~United States of America."
Responding, esja asserts "this is not true", though doesn't clarify their redefinition of "worst hit", for another two rounds of discussion, finally settling on "deaths today".
That basis is fatally (so to speak) flawed as it entirely dismisses the facts that:
Cases today translate directly to deaths in the 2--4 week future, at a best-case rate of 0.5% CFR and far more plausibly 1.5--3% CFR, based on present reported cases.[1]
US new cases per capita are at least on par if not worse than Europes's.
Europe's daily case rates are trending at worst flat, and are generally decreasing.
US case rates are rising, at an acellerating rate.
The US today reports 158,363 new cases (7-day average), and a 3% CFR. In ~2--3 weeks, likely daily deaths will be 2,375--4,750, or 7.5--15 per million.[2]
Germany, to use esja's favoured example, reports 18,363 new cases (7-day average), and a 2% CFR. In ~2--3 weeks, likely daily deaths will be 367--550, 4.4--6.6 per million.[3]
All Europe reports ~220,000 new daily cases (16 Nov 2020, not smoothed). in ~2--3 weeks, likely daily deaths will be 3,300--6,600, 4.4--8.8 per million.[4]
In all cases, baked-in future daily US mortality rates will be roughly twice those of Europe, adjusted for population and are trending still further worse. The US 'benefits' only by having begun its annual seasonal coronavirus peak some 4--8 weeks later than Europe, with an European inflection beginning in September--October and a US inflection beginning October--November.
To provide an analogy, esja is laughing at Europe being in a ditch whilst the US is racing toward a cliff's edge. Assessments of present health or wealth must include obvious future consequences or risks. Critics of EU response entirely ignore these, and reframe the initial criterion to do so.
Such analysis suffers from presentism and risk blindness and is utterly flawed.
Adapted from HN comments to the thread linked above.
Notes:
Beginning here: https://news.ycombinator.com/item?id=25113115
I'm ignoring the fact that reported fatalities undercount true COVID-19 fatalities as demonstrated by overall excess deaths by about 30% per an August 2020 New York Times report and other independent studies and data. This is a largely global bias, doesn't affect inter-regional comparisons, simplifies analysis, and strengthens my argument as the case I present, bleak as it is, is less severe than the actual reality
Using 1.5--3% CFR.
Also using 1.5%--35 CFR, despite Germany's lower experienced CFR.
Worldometers does not provide continental/regional plots or smoothed trends, though law-of-large-numbers helps somewhat. Again at 1.5--3% CFR, based on reported values, whic undercounts recoveries, experienced CFR is ~4%. Using a non-smoothed current high-point number further overstates total European future mortality relative to the US.
#covid19 #UnitedStates #europe #CriticalThinking #FlawedArguments #HackerNews #risk #worldometers
