4 ms·

> Current predictions put the economic impact per life saved by social distancing at ~$190 million. That sounds, on the face of it, implausibly high. To run s

> Current predictions put the economic impact per life saved by social distancing at ~$190 million. That sounds, on the face of it, implausibly high. To run some numbers - estimates of how much of the population would be infected without countermeasures run 40-70%, and death rates run 1-2% (this would be much worse if hospitals were overrun by an estimated 8 million ICU patients over a few months, but we'll use very conservative numbers). The US population, with 40% infected and 1% of those dead, leaves us with about 1.3M deaths, times your $190M per life is $25T, or about 15 months of US GDP. Is that really the claim here? The equivalent of bringing GDP to 0 for 15 straight months?

manfredo 7y ago
The death rates so far have been dropping considerably once further testing becomes available. Right now, most tests are being performed in response to patients presenting symptoms. This creates artificially high results. The majority of COVID-19 infections do not result in symptoms, and these people would likely never be reported. Nor would a significant portion of people who did present flu-like symptoms but did not need medical care. Italy and Denmark have initiated random testing have revised death rates to be far lower, less than once percent. Random testings substantially increases the denominator, since it captures the vast majority of people who did not have serious symptoms. The infected cruise ship had a death rate of 0.7%, which is even more surprising giving the older average age of cruise passengers. The fact that places are reporting higher death rates than a population disproportionately older should have raised eyebrows. Again, the data itself is not suspect it's failing to account for the selection bias it imposes. But this does result in some very skewed conclusions on your end: > The US population, with 40% infected and 1% of those dead, leaves us with about 1.3M deaths, times your $190M per life is $25T, or about 15 months of US GDP. Is that really the claim here? The equivalent of bringing GDP to 0 for 15 straight months? 1.3 million deaths is an order of magnitude more than what is being projected. Earlier this month, the range given was 100,000 - 240,000. Now this projection has been lowered first to 80,000 and now to 60,000: https://www.cnn.com/2020/04/08/health/us-coronavirus-wednesday/index.html https://www.cnn.com/2020/04/08/health/us-coronavirus-wednesd... https://www.usatoday.com/story/news/investigations/2020/04/09/coronavirus-deaths-u-s-could-closer-60-k-new-model-shows/5122467002/ https://www.usatoday.com/story/news/investigations/2020/04/0... https://www.bloomberg.com/news/articles/2020-04-09/fauci-says-u-s-virus-deaths-may-be-60-000-halving-projections https://www.bloomberg.com/news/articles/2020-04-09/fauci-say...
- CydeWeys 7y ago
  Yes, those projections are that low precisely because the strict social distancing is worsening, and we are managing to keep things within the capacity of the healthcare system. Reopen everything immediately and you'd see the death toll shoot right back up into the previously estimated, much higher ranges.
  - manfredo 7y ago
    The "much higher" ranges are still an order of magnitude less than your 1.3 million scenario.
- thedufer 6y ago
  > Right now, most tests are being performed in response to patients presenting symptoms. This creates artificially high results. You don't think anyone else has noticed this? The estimates have been based on lower numbers than the raw data. > The infected cruise ship had a death rate of 0.7% I agree that the Diamond Princess was particularly interesting because of the high testing rate. Your number is very clearly wrong, though. The actual numbers reported was 13 deaths in 712 cases, or 1.8%. They also have 55 unresolved cases, and someone who has been sick for this long is looking at worse odds than the general population. > 1.3 million deaths is an order of magnitude more than what is being projected. Earlier this month, the range given was 100,000 - 240,000. Now this projection has been lowered first to 80,000 and now to 60,000: First of all, those numbers are going down precisely because of the measures you're complaining about. Second, they're clearly implausible. The 60,000 estimate is from 9 days ago. Meanwhile, the current official count is just shy of 40,000, and the excess deaths have consistently been double or more of the estimate for COVID-19 deaths (Italy, for example, is reporting 6x [1], while NYC was until a recent change in their reporting showing 2x [2] - but we haven't seen these changes anywhere else in the country). [1] https://nymag.com/intelligencer/2020/04/coronavirus-is-only-part-of-the-excess-fatality-mystery.html https://nymag.com/intelligencer/2020/04/coronavirus-is-only-... [2] https://gothamist.com/news/surge-number-new-yorkers-dying-home-officials-suspect-undercount-covid-19-related-deaths https://gothamist.com/news/surge-number-new-yorkers-dying-ho...

manfredo 7y ago

The death rates so far have been dropping considerably once further testing becomes available. Right now, most tests are being performed in response to patients presenting symptoms. This creates artificially high results. The majority of COVID-19 infections do not result in symptoms, and these people would likely never be reported. Nor would a significant portion of people who did present flu-like symptoms but did not need medical care. Italy and Denmark have initiated random testing have revised death rates to be far lower, less than once percent. Random testings substantially increases the denominator, since it captures the vast majority of people who did not have serious symptoms. The infected cruise ship had a death rate of 0.7%, which is even more surprising giving the older average age of cruise passengers. The fact that places are reporting higher death rates than a population disproportionately older should have raised eyebrows. Again, the data itself is not suspect it's failing to account for the selection bias it imposes. But this does result in some very skewed conclusions on your end: > The US population, with 40% infected and 1% of those dead, leaves us with about 1.3M deaths, times your $190M per life is $25T, or about 15 months of US GDP. Is that really the claim here? The equivalent of bringing GDP to 0 for 15 straight months? 1.3 million deaths is an order of magnitude more than what is being projected. Earlier this month, the range given was 100,000 - 240,000. Now this projection has been lowered first to 80,000 and now to 60,000: https://www.cnn.com/2020/04/08/health/us-coronavirus-wednesday/index.html https://www.cnn.com/2020/04/08/health/us-coronavirus-wednesd... https://www.usatoday.com/story/news/investigations/2020/04/09/coronavirus-deaths-u-s-could-closer-60-k-new-model-shows/5122467002/ https://www.usatoday.com/story/news/investigations/2020/04/0... https://www.bloomberg.com/news/articles/2020-04-09/fauci-says-u-s-virus-deaths-may-be-60-000-halving-projections https://www.bloomberg.com/news/articles/2020-04-09/fauci-say...

CydeWeys 7y ago
Yes, those projections are that low precisely because the strict social distancing is worsening, and we are managing to keep things within the capacity of the healthcare system. Reopen everything immediately and you'd see the death toll shoot right back up into the previously estimated, much higher ranges.
- manfredo 7y ago
  The "much higher" ranges are still an order of magnitude less than your 1.3 million scenario.
thedufer 6y ago
> Right now, most tests are being performed in response to patients presenting symptoms. This creates artificially high results. You don't think anyone else has noticed this? The estimates have been based on lower numbers than the raw data. > The infected cruise ship had a death rate of 0.7% I agree that the Diamond Princess was particularly interesting because of the high testing rate. Your number is very clearly wrong, though. The actual numbers reported was 13 deaths in 712 cases, or 1.8%. They also have 55 unresolved cases, and someone who has been sick for this long is looking at worse odds than the general population. > 1.3 million deaths is an order of magnitude more than what is being projected. Earlier this month, the range given was 100,000 - 240,000. Now this projection has been lowered first to 80,000 and now to 60,000: First of all, those numbers are going down precisely because of the measures you're complaining about. Second, they're clearly implausible. The 60,000 estimate is from 9 days ago. Meanwhile, the current official count is just shy of 40,000, and the excess deaths have consistently been double or more of the estimate for COVID-19 deaths (Italy, for example, is reporting 6x [1], while NYC was until a recent change in their reporting showing 2x [2] - but we haven't seen these changes anywhere else in the country). [1] https://nymag.com/intelligencer/2020/04/coronavirus-is-only-part-of-the-excess-fatality-mystery.html https://nymag.com/intelligencer/2020/04/coronavirus-is-only-... [2] https://gothamist.com/news/surge-number-new-yorkers-dying-home-officials-suspect-undercount-covid-19-related-deaths https://gothamist.com/news/surge-number-new-yorkers-dying-ho...