Does this cohort data also exist by income group and if so how have these rates evolved? And how does the top performing cohort compare to other places in the world? Say OECD countries.
This is a good question -- so, we actually have some interesting related info.
The Society of Actuaries have a couple relevant research projects on this. One is based on general population mortality trends, and while death certificates (for obvious reasons) don't have the decedent's income on it, we do have death trends by county.... and all sorts of socioeconomic info by county. So the SOA has stratified counties by socioeconomic quintile and made a dashboard:
Second, the SOA has some experience studies for retirement plans, and they've got some stratification by employment type (blue collar/white collar) and some other stuff like that
But looking at the reports I see, it will give you a snapshot at one moment in time, and generally the trend tables they apply the same improvement scale to all tables.
A good example of cohort life expectancy is that if you were a Japanese male born in 1925, you had a much lower cohort life expectancy (due to soldiers dying in World War II) than if you were a Japanese male born in 1930, and too young to fight in that war.
I hope so, HOWEVER -- drug overdose and motor vehicle accident deaths have increased by a lot, and those affect younger adults far more than older ages.
So even with COVID mortality going away, I could see a depressed period life expectancy result in 2023 compared with 2019, as I don't see the drug overdose trend being dealt with.
In the video, I mentioned that life expectancy improvement stalled out in the mid-2010s. That was due to the increase in drug OD deaths, which has had a bad trend for years, and only got worse during the pandemic.
First -- that sort of regulation is done at the state level, so that can differ state-by-state.
Second, they definitely can ask about smoking tobacco, and alcohol use, in all states.
Third, I have seen some stuff about marijuana use & life insurance, as many states are approving marijuana use at various levels (though the federal government still criminalizes use, which is a weird legal situation, but that's federalism for you.)
Fourth, whether or not the insurers ASK about the use, it can be a part of the blood/urine tests as part of the medical screens in applications. And insurers definitely can look up criminal histories as part of underwriting not only for life insurance but for other types of insurance coverage... but again, that can differ by insurance type and by state.
For the record, I have been involved in discussions with non-U.S. insurance regulators, and trying to explain the "weird" situation of U.S. insurance regulation.
It was very amusing to me to have to explain that no, I did not think the state insurance regulators would be interested in giving up their power.
But for those of us who only care about retiree mortality, I’m thinking it will bounce back. Who knows; maybe the older folks who survived COVID are healthier than the average old person years ago (survival of the fittest?) and life expectancy for them will increase beyond 2019 numbers.
Yes, I've discussed some of that with others at the Hartford Actuarial Club (and elsewhere) -- given the drop in mortality rates for age 80+ in 2021, it looks like some of the "dry tinder" got burned up in 2020. (Also, it didn't help that some like Gov. Cuomo in NY and Gov. Whitmer in MI sent COVID positive folks back to nursing homes)
So maybe some of the most vulnerable seniors died in 2020, and then less vulnerable seniors are now left.
Does this cohort data also exist by income group and if so how have these rates evolved? And how does the top performing cohort compare to other places in the world? Say OECD countries.
This is a good question -- so, we actually have some interesting related info.
The Society of Actuaries have a couple relevant research projects on this. One is based on general population mortality trends, and while death certificates (for obvious reasons) don't have the decedent's income on it, we do have death trends by county.... and all sorts of socioeconomic info by county. So the SOA has stratified counties by socioeconomic quintile and made a dashboard:
https://tableau.soa.org/t/soa-public/views/USPop2020-Full/DBbyCounty?%3Aembed=y&%3AisGuestRedirectFromVizportal=y
(I believe I did a video on this once, but I should do one just on this dashboard)
Second, the SOA has some experience studies for retirement plans, and they've got some stratification by employment type (blue collar/white collar) and some other stuff like that
But looking at the reports I see, it will give you a snapshot at one moment in time, and generally the trend tables they apply the same improvement scale to all tables.
So here are for private plans:
https://www.soa.org/4aa609/globalassets/assets/files/resources/experience-studies/2019/pri-2012-overview-highlights.pdf
Public plans:
https://www.soa.org/resources/research-reports/2019/pub-2010-retirement-plans/
A good example of cohort life expectancy is that if you were a Japanese male born in 1925, you had a much lower cohort life expectancy (due to soldiers dying in World War II) than if you were a Japanese male born in 1930, and too young to fight in that war.
If COVID-19 deaths fall to a very low level in 2023, would you expect life expectancy to rebound close to where it was in 2019?
I hope so, HOWEVER -- drug overdose and motor vehicle accident deaths have increased by a lot, and those affect younger adults far more than older ages.
So even with COVID mortality going away, I could see a depressed period life expectancy result in 2023 compared with 2019, as I don't see the drug overdose trend being dealt with.
In the video, I mentioned that life expectancy improvement stalled out in the mid-2010s. That was due to the increase in drug OD deaths, which has had a bad trend for years, and only got worse during the pandemic.
In the US, are insurance companies allowed to ask drug abuse history?
First -- that sort of regulation is done at the state level, so that can differ state-by-state.
Second, they definitely can ask about smoking tobacco, and alcohol use, in all states.
Third, I have seen some stuff about marijuana use & life insurance, as many states are approving marijuana use at various levels (though the federal government still criminalizes use, which is a weird legal situation, but that's federalism for you.)
Fourth, whether or not the insurers ASK about the use, it can be a part of the blood/urine tests as part of the medical screens in applications. And insurers definitely can look up criminal histories as part of underwriting not only for life insurance but for other types of insurance coverage... but again, that can differ by insurance type and by state.
For the record, I have been involved in discussions with non-U.S. insurance regulators, and trying to explain the "weird" situation of U.S. insurance regulation.
It was very amusing to me to have to explain that no, I did not think the state insurance regulators would be interested in giving up their power.
But for those of us who only care about retiree mortality, I’m thinking it will bounce back. Who knows; maybe the older folks who survived COVID are healthier than the average old person years ago (survival of the fittest?) and life expectancy for them will increase beyond 2019 numbers.
Yes, I've discussed some of that with others at the Hartford Actuarial Club (and elsewhere) -- given the drop in mortality rates for age 80+ in 2021, it looks like some of the "dry tinder" got burned up in 2020. (Also, it didn't help that some like Gov. Cuomo in NY and Gov. Whitmer in MI sent COVID positive folks back to nursing homes)
So maybe some of the most vulnerable seniors died in 2020, and then less vulnerable seniors are now left.