There has been a suggestion that people here enter into bets about what they are asserting will happen about C19 going forward. Some of the assertions are put more mildly, or are only implied, certainly, but numbers have a way of bringing clarity. to statements. The advantage will be going back and checking, bringing the items forward. I am not a betting man, but I should at least make some assertion to get us started. I am 90% confident there will be 100,000 American deaths by the end of the year. I am 30% confident there will be 200,000.
You can address an entirely different question if you like, such as penetration into the Western states, or number of overall cases or whatever.
A month from now I might be hanging my head in shame, but that's where I am starting, here at the end of April.
15 comments:
Merely by framing your prediction as a range you are discouraging 95% of the predictions that people might make. That wouldn't be a bad thing. My prediction is: Who knows, time will tell.
@Jonathan, well that seems to be one prediction you can make to ensure you are never incorrect on any topic.
The Slate Star Codex blog got me in to the idea of no stakes personal betting to start calibrating my own intellectual instincts (he does yearly predictions), because it is good when you start getting confident or critical to look back at your track record and say "ah, yes, but I tend to jump the gun on topic x" or "I lowball it on topic y". I note he put up some predictions up about this recently, but I'll do mine first:
1. We reach 100,000 US deaths by year end - 90% confident
2. We reach 150,000 US deaths by year end - 70% confident
3. We have at least one drug or drug cocktail shown to significantly (in the statistical and clinical sense) by year end - 40% confident
4. We have a vaccine (new or old) that is shown effective and is being used in at least 10,000 people in the US - 20% confident
Obviously I hope I look back and find my confidence in those first two was misplaced and the numbers were much lower. And I hope I look back and discover my confidence in those last two should have been higher because they happened.
SSC predictions here: https://slatestarcodex.com/2020/04/29/predictions-for-2020/
I bet total US deaths (all causes) will be lower in 2020 than in previous years.
@ stevo - that's an interesting one, because lockdowns have multiple side effects, including traffic deaths and other infections
I wager that worldwide death from the economic effects will be larger than from the virus itself. I wager that will be true even though I’m actively trying to stop it from being true.
@Grim - I'd be interested to hear if you have any thoughts on how to parse out numbers for that prediction, as it seems there's going to be a lot of grey areas in what gets reported. For example, I think you mentioned UN reports recently. I looked some up and couldn't find the one you mentioned, but did find one that talked about the massive impact in the third world of halting childhood vaccination programs due to COVID. What was not clear is if these were being halted because the resources were being redirected to treat COVID, they were stopped in attempt to stop the spread of COVID, or if it was because hospitals now lacked resources to fund their vaccination programs. Seemed like the kind of thing ripe for classifying either way depending on what you wanted to prove.
In the long run it seems like there's really 5 categories of deaths here:
1. Virus related deaths that will happen regardless of what path we take because novel deadly viruses are awful and hard to handle
2. Non-virus economy related deaths that will happen regardless of what path we take just because novel deadly viruses are awful and hard to handle
3. Virus related deaths that are preventable if we act in certain ways to minimize them.
4. Non-virus economy related deaths that are preventable if we act in certain ways to minimize them.
5. Non-virus deaths due to healthcare system issues that may or may not be economic in nature.
I'm absolutely terrible at economic type data, but I am making an assumption here that economies take a hit when bad unforeseen things happen regardless, so parsing #2 and #4 seems key for decision making, and where you file #5 is going to shift things substantially.
In a year, population density and weather will be more statistically significant to the number of COVID fatalities in a state than social distancing practices.
In the next few months there will be a significant spike in the number of cases and deaths in the northern tier urban areas around the Great Lakes.
If I were running a hospital in rainforest Africa, COVID would not be where I'd put what little resources I had.
My prediction: 50% sure that I'll get the disease [in some form] sometime in the next three years, assuming I don't die of something else first.
@bs king:
The report I'm especially thinking of is here: https://www.un.org/sites/un2.un.org/files/policy_brief_on_covid_impact_on_children_16_april_2020.pdf
Here's a section from the summary:
"Threats to child survival and health:
Economic hardship experienced by families
as a result of the global economic downturn
could result in hundreds of thousands of
additional child deaths in 2020, reversing
the last 2 to 3 years of progress in reducing
infant mortality within a single year. And
this alarming figure does not even take into
account services disrupted due to the crisis –
it only reflects the current relationship
between economies and mortality, so is
likely an under-estimate of the impact. Rising
malnutririon is expected as 368.5 million
children across 143 countries who normally
rely on school meals for a reliable source
of daily nutrition must now look to other
sources."
Here is another report from the UN estimating contraction effects of 5, 10, and 20 percent. In the best case scenario, 85-135 million people are put into extreme poverty. In the worst case scenario, half a billion people end up in extreme poverty.
https://www.wider.unu.edu/sites/default/files/Publications/Working-paper/PDF/wp2020-43.pdf
How you count the edge cases only matters if the effects are close. These numbers are orders of magnitude worse.
@ Grim - these are among the most excellent points for Americans taking other peoples into account in our estimation of risks and benefits. Keep it coming, because it is powerful. However, we don't know how much of this is already baked in. The effect of the virus out of China on the poor of the world might already be 50, 75, 90% sealed, whatever we do. We know for certain that the cost is high even if Americans did a rash reopening tomorrow and restarted at full speed.
Your numbers sound about right to me.
@ Grim - a comment of yours farther back. You used the phrase "economic effects." I think those effects will be larger, as you say, so long as we include poorer countries. But those are not the same thing as "government effects" in general, much less US government/state government effects. There has been an increasing tendency among conservatives over the last month to treat voluntary moves by businesses (some of which fear being sued if their patrons get sick) as entirely the result of government actions. Many Christian denominations will restart services well after the time the government gives them permission. This will hold true for any large-gathering enterprise. Still more will institute modifications for social distancing that will reduce the number of customers, and that has economic effect as well.
So, say it is 90% baked in and assume that the UN numbers are right. (There’s no reason they should prove better than climate models or Covid models, but for the sake of argument). If we can effect only a 10% improvement on a 20% contraction, that’s 50 million people. They won’t all die from being forced into extreme poverty, but they will all suffer; and many will die.
Say it’s 10% of a 5% contraction. Now you’re ‘only’ saving maybe as few as 8.5 million people; but by holding it to 5% instead of 20%, you’re also saving ~490 million from the risk of extreme poverty.
I’m ok with your second argument. Government may have a role here, even if it’s largely limiting liability. That’s the real major effect of Trump nationalizing the meat packing system. Now they can’t be sued, so they’ll keep producing meat if they can continue to staff plants. Good.
Essentially agree. Over a huge worldwide population, small percentages matter. But it's still going to be terrible for everyone in the third world, and setting up the standard that we have to fix that to have succeeded is going to have bad political consequences, partly for the UN or the left blaming the US and the free market system which is the biggest engine of what can save them, or...
...for the next crisis or crises. We should open up. But we should also expend our energy on the narrative now, or more will die later.
Post a Comment