Nice links! I personally think that the idea that the shingles vaccine preventing dementia is most likely just a confound, like so many other things in dementia research. Here's an intriguing take on the topic, probably too strong but I'd guess most likely directionally correct, i.e. the effect is probably largely just a confound. https://www.drvinayprasad.com/p/shingles-vaccines-and-dementia-the
Interesting. Essentially, getting shingles gives you more contact with the medical system, making it more likely to receive a diagnosis. And people who don't get the vaccine are more likely to get shingles.
I think it's more likely the healthy vaccinee effect. I.e. healthier, more health-engaged people get vaccinated and also have lower dementia rates for reasons unrelated to the vaccine.
Regression discontinuity designs are designed to neutralize healthy vaccinee confounding by comparing people just on either side of an eligibility birthday cutoff, who should be near-identical in baseline health. But Prasad points out that even here, the speed of the effect is implausible. This brings me back to my prior, which is that there is some sort of confound. I've just seen too many observational dementia findings blown up in actual RCTs (e.g. aspirin, ginkgo, vitamin E, NSAIDs).
So are property taxes. {Hypothesis: the advantage of LVT over PT is that _in practice_ PT do NOT value undeveloped land at its opportunity value and LVT does try to.
Nice links! I personally think that the idea that the shingles vaccine preventing dementia is most likely just a confound, like so many other things in dementia research. Here's an intriguing take on the topic, probably too strong but I'd guess most likely directionally correct, i.e. the effect is probably largely just a confound. https://www.drvinayprasad.com/p/shingles-vaccines-and-dementia-the
Interesting. Essentially, getting shingles gives you more contact with the medical system, making it more likely to receive a diagnosis. And people who don't get the vaccine are more likely to get shingles.
I think it's more likely the healthy vaccinee effect. I.e. healthier, more health-engaged people get vaccinated and also have lower dementia rates for reasons unrelated to the vaccine.
Regression discontinuity designs are designed to neutralize healthy vaccinee confounding by comparing people just on either side of an eligibility birthday cutoff, who should be near-identical in baseline health. But Prasad points out that even here, the speed of the effect is implausible. This brings me back to my prior, which is that there is some sort of confound. I've just seen too many observational dementia findings blown up in actual RCTs (e.g. aspirin, ginkgo, vitamin E, NSAIDs).
"Land value taxes are highly progressive"
So are property taxes. {Hypothesis: the advantage of LVT over PT is that _in practice_ PT do NOT value undeveloped land at its opportunity value and LVT does try to.