My meteorologist friend and sometime EconLog commenter Tom Lee had an excellent letter in today’s Monterey Herald. Here it is:
Humans get better at weather protection
“Study blames climate change for 37% of global heat deaths” reads the Herald headline, echoing a headline syndicated across the world. Sounds sinister, but surprisingly the study that generated the headline was not based on actual data on people who died because of heat. It was based on a computer model. If it were valid, one would expect it to agree with other data about heat-related deaths. Not so.
EPA-published records for heat-related deaths over the United states show no trend since about 1975. In fact, the most annual deaths occurred in 1980. If a warming climate were killing more and more people, wouldn’t we be seeing the most deaths now? The EPA heat wave index reaches back to 1895; the worst heat waves in the U.S. occurred in the 1930s with the Great Dust Bowl.
The article also claims an increasing toll due to global warming from storms, flooding and drought. But a plot of data from the EMDAT Global Disaster Database shows that climate- related deaths have plummeted in recent years. The reason: we humans have gotten much better at protecting ourselves from the weather. Increasing global wealth means that even the poorest are less vulnerable to the elements.
— Thomas F. Lee, Monterey
Interestingly, I posted on it on Facebook, but the post was apparently taken down without any message to me. So I did it again and it was taken down again. The third time is probably not a charm. The Facebook Stasi are apparently on their game. And what nerve Tom had in referring to data reported by that hotbed of libertarianism, the Environmental Protection Agency.
The picture above is of members of the Stasi.
Note: Before you get all hot and bothered, recognize that of course I know I’m exaggerating by referring to the employees of a for-profit firm as Stasi. I’m using the term they way the people on Seinfeld referred to the soup Nazi. In both cases, their private property gives them the absolute right to be petty, close-minded people who don’t want others to see alternative views, no matter how science-based.
READER COMMENTS
Thomas Lee Hutcheson
Jun 5 2021 at 5:59pm
I do not se anything inconsistent with the claim that x% of heat deaths would not have happened if wheatear patterns that would have prevailed with less CO2 in the atmosphere AND that we are learning better ways to prevent heat deaths.
I put it on my Facebook to see what happens.
Jon Murphy
Jun 6 2021 at 9:27am
Note that’s not the claim, though. The claim is: “Climate change [is to blame] for 37% of global heat deaths.” They’re making a factual claim, not a counterfactual claim.
Thomas Lee Hutcheson
Jun 6 2021 at 10:58am
“X caused Y means if it were not for X, Y would not have happened.
No increase in atmospheric CO2 => x% fewer heat deaths.
No increase in knowledge about how to prevent heat deaths => more heat deaths.
Jon Murphy
Jun 6 2021 at 11:31am
That’s an incorrect statement, though. I think you’re trying to make a counterfactual claim, but the logic is a bit messed up. X can cause Y, but lots of other things can cause Y as well. You’re implicitly assuming a uni-causal relationship here.
But note that it does not logically follow then that climate change is responsible for x% fewer heat deaths. The assumption necessarily for the first statement is violated by the second. Thus it is factually incorrect to claim that climate change caused a given level of deaths, QED.
Thomas Lee Hutcheson
Jun 7 2021 at 5:53am
I do not understand your point.
“But note that it does not logically follow then that climate change is responsible for x% fewer heat deaths. The assumption necessarily for the first statement is violated by the second. Thus it is factually incorrect to claim that climate change caused a given level of deaths, QED”
That climate change is responsible for x% of heat deaths is a factual claim. I may be fatally incorrect as you claim, but Henderson did not give any reasons to doubt it, nor do you.
And I stand by my view that whatever is going on in Henderson’s juxtaposition of the two models of factors contributing to heat deaths has not clear relevance to a model vs data distinction.
Jon Murphy
Jun 6 2021 at 11:39am
Casey Mulligan also has a great paper which addresses the error in reasoning you’re making.
Mulligan shows how the COVID lockdowns actually led to more deaths because they didn’t take into account changes in people’s behavior.
So, let’s alter your claim a little bit:
We can accept both of these to be true, but they are mutually exclusive. Absent more knowledge and information, lockdowns prevent COVID deaths (Claim A). But, if Claim B is false (ie, there is an increase in knowledge), then the necessary conditions of A no longer hold, and thus the conclusion does not hold. Indeed, the conclusion becomes the opposite! Lockdowns => x% greater COVID deaths!
So, again, we have to rely on data, not models, to make factual claims. Models are not data.
Knut P. Heen
Jun 8 2021 at 10:10am
A numerical example may be helpful.
Suppose 1000 deaths of which 370 are due to climate change.
Suppose we prevent 900 deaths. Wouldn’t we expect to see 100 deaths and 37 due to climate change?
I put little confidence in the 37 percent coming out of their model, but that is not what this discussion is about.
Thomas Lee Hutcheson
Jun 5 2021 at 8:06pm
I also do not understand in the context of contrasting “model” vs “data.”
There is a model linking heat deaths and CO2 emissions (probably taking prevention technology as given) and another linking heat deaths and knowledge of how to prevent heat deaths (probably taking weather as given). Both models are based on data and do not produce obviously contradictory results.
Jon Murphy
Jun 6 2021 at 9:30am
“The most that can be expected from any model is that it can supply a useful approximation to reality: All models are wrong; some models are useful.” -G.E.P. Box
Models are not reality. They are an attempt to think about reality, cause and effect, and make predictions. Models use assumptions, but they are not themselves data.
Yes, models are (ideally) built off data and they are tested by data. But they are not data themselves.
Thomas Lee Hutcheson
Jun 6 2021 at 10:24am
I agree with what you say but it seems irrelevant as Henderson is discussing two different models.
Jon Murphy
Jun 6 2021 at 11:32am
No. See my response to you immediately above.
Incidentally, confusion about modeling seems to be very prevalent among scientists. See my latest paper with Abigail Devereaux, Nathan Goodman, and Roger Koppl.
Thomas Lee Hutcheson
Jun 7 2021 at 6:06am
I see your response, It fails to address in what way Henderson’s bringing together two different models of factors affecting heat deaths (either of which could be correct or not) is showing us anything about a “models” vs “data” distinction.
john hare
Jun 6 2021 at 2:17pm
I find it more distressing that an apparent public forum is dismissing hard data than that there might be different interpretations of it as discussed in most of the comments. Anything can be “proved” if no dissent is allowed.
Philo
Jun 6 2021 at 3:25pm
My attempts to post this blogpost on Facebook have failed.
David Henderson
Jun 6 2021 at 4:50pm
Thanks for that datum, Philo.
TMC
Jun 7 2021 at 10:41am
From webmd.com ” there were 1,935 cold-related deaths and 70 heat-related ones. The cold caused 94% of temperature-related deaths”
Logically then, the additional heat that ‘37% of global heat deaths’ would have also saved twenty times more lives from the cold.
Matthias
Jun 8 2021 at 9:28pm
Maybe they were a bit sloppy with their words?
From what I’ve heard, people blame global warming for more extreme weather in general.
TMC
Jun 9 2021 at 11:54am
The claim, and my refutation (not mine originally) have been around for quite a while.
Look up Judith Curry as for extreme weather. Both data and theory suggest GW reduces it.
Dan Culley
Jun 9 2021 at 9:55am
Have you appealed your take-downs? It’s doubtful that any human being looked at your post and made a conscious decision. The appeals do get reviewed by humans and also help better train their algorithms to avoid this happening in the future. My guess is that you would very likely have your post restored. And, who knows, if not you might even make it to their mythical Supreme Court. You’d probably at least enjoy it.
David Henderson
Jun 10 2021 at 2:46pm
I haven’t appealed. And it turns out that I don’t need to. The posts showed up 2 days ago.
TMC
Jun 9 2021 at 12:01pm
David,
I do enjoy this site, but I find this ironic given the number of deleted comments here.
Comments are closed.