Yes, simply taking account the average monthly winter death rates compared to the minimum during the summer shows that US winters kill about 120,000 each year. Most of the extra deaths occur in the elderly or those with a severe chronic health problems. Winter weather can significantly increase the risk of death even if it is not the only factor.
Statistics shows that death rates have a significant seasonal component as shown in the graph below. Deaths go up in winter weather and reach a minimum in warm weather.
For the US, the minimum death rate occurs in the August with a about 93% of the annual average. The peak occurs in February with about 111% of the average. About 2.7 million people die each year so the excess mortality during December, January, February and March compared to August works out to about 120,000 people.
If you say this figure sounds ridiculously high consider this report from the UK, which estimates that 48000 people died as result of winter weather last winter. The UK has about 1/6 of the US population, so the figure would correspond to about 300,000 deaths in the US. My guess is that the the greater seasonal effect in the UK may be a result of the lack of central heating in the UK compared to the US.
This type of analysis was recently used to estimate the number of deaths from recent hurricanes. My understanding is that most of the increase comes as a result of damage to electrical power and other infrastructure and most of the deaths are elderly or those with chronic health problems.
For example, consider an elderly person who has smoked for 50+ years and has a lung condition that requires oxygen. Is it fair to say that power failure from a storm caused the death?
What about decades smoking, old age, and lung disease? Aren’t they also contributing to the death as well?
How do you assign blame when there are logically multiple causes for a particular death?
Yes, it is using “excess mortality” compared to what would have been expected to have occurred without the storm. That same approach says we could save 120,000 lives if we could magically keep US weather like it is in August year round.
Is that a fair way of evaluating cause of death? A smoker with terminal lung cancer may be recorded as a death from a storm-related power failure, but that ignores the fact that the smoker had life expectancy of only a few months to begin with.
You prepare for winter and the aftermath. You prepare for a hurricane and the aftermath. You then decide if the outcomes were acceptable. If not you make changes. You’re trying to confound things.
An issue is that you logically end up with multiple cause for the same death.
Someone may arguably die from combination of decades of smoking, lung cancer, obesity, genetic predisposition, and a power failure caused by a storm. The risk of dying is arguably reduced by eliminating any of these factors.
Another issue is that if a storm, such as Maria, hastened the death or relocation of people most likely die in a storm, then a repeat of the same or bigger storm is likely to have a much lower death rate. Would that mean that a Maria would actual save lives by reducing the estimated fatalities in a second storm?
The what ifs get very confusing and can quickly get into philosophical issues.
Cohort 1: Puerto Rico mortalities in non-hurricane 6 month intervals.
Cohort 2: Puero Rico mortalities in hurricane year 6 month intervals.
You have two variables, year and hurricane. Causes of death not related to the hurricane cancel out from both cohorts. Any increase in a specific cause of death between the two cohorts would be considered secondary to one of the two variables.
Taking cancer related deaths for example. If the non-hurricane cohort had 5,000 deaths over a time interval and the hurricane cohort had 5,500 deaths, you’d have an increase of 500 deaths. You’d have to show that the prevelance of cancer had also increased to offset those deaths (time variable) or you’d assume it was an effect of the hurricane.
It still implies that the hurricane was responsible for any increase in deaths.
There is an old saying, “earthquakes don’t kill people, buildings do.” The same thing is true with hurricanes. Properly built structures can withstand hurricanes with minimal damage.
In the case of Puerto Rico, people who I have talked to from the island say that the power system was notoriously unreliable long before the storm. The power would frequently go down after even a small thundershower or for no reason at all. On top of that the power company is bankrupt.
If the lack of power is the main factor in increased deaths, is the ultimate cause the hurricane or simply the long-standing problems with management of the electric company?
Winter weather is hardly unexpected in most of the US. Neither are hurricanes in the Caribbean. Is weather really the cause of the deaths?
Would the building have killed someone if the earthquake didn’t happen? No. The building killed someone because the earthquake happened, causing the building to collapse.
If a hurricane happens, which leads to power outtages, lack of water, whatever, the root cause is still the hurricane. You can say there was poor baseline infrastructure, but without the hurricane, those issues would not have happened.
Obviously we have to ignore the poor building codes put in place by democrats, otherwise it isn’t fair and we can’t blame a Republican President. Yes, we know.
Instead of blasting the 3,000 death study, Trump should have hired professionals to do their own analysis. The main point being that even if you can attribute the deaths to the hurricane you cannot compare those 3,000 deaths to the deaths of Sandy and Katrina. You would have to do a comparable study of each of those on a similar basis. It may be that, using this method, Sandy killed 10,000.
Trump bungled the point but the point remains, the media beat him up with a study using a new metric for measuring deaths in a hurricane. As usual though, he was too stupid to make the actual point and fumbled an easy lay up. and yes I freely mix and match my sports meataphors.
