2020 Excess Deaths Analyzed

I have read many things about excess deaths in 2020, and how they relate to COVID. Some people claim there are no excess deaths. Others claim that excess deaths are much higher even than the COVID deaths. Being a numbers guy, I decided to run the numbers myself. But I wanted to look at the numbers for each age group, as that seemed to be an important point. One would expect if the excess deaths were from COVID, that the excess deaths would be similar to the covid deaths for that age group.

The data I used came from the CDC. Specifically, I used Figure 3. Age-specific death rates for ages 15 years and over: United States, 2017 and 2018 from the Mortality in the United States, 2018 brief that can be found at https://www.cdc.gov/nchs/products/databriefs/db355.htm for the death rates and the Provisional Death Counts for Coronavirus Disease 2019 (COVID-19) that can be found at https://www.cdc.gov/nchs/nvss/vsrr/covid_weekly/index.htm. I selected the 2018 death rates, as those death rates are statistically lower than 2017. I have not found 2019 death rates available, but I am making the assumption that lower death rates were a trend. I went for the more conservative death estimates that would maximize excess deaths in my calculations.

Lets start with COVID deaths by age group. I start with ages 15-24, as that is where the data for death rates starts. I threw in a calculation of percent of deaths from COVID. As you can see, the number of deaths from COVID, as well as the number of total deaths rise as age rises. That would be expected. More interestingly, the percentage of COVID deaths per all deaths starts very small at the youngest ages, but increases with age till about the 65-74 age group. At that point, the numbers basically stay statistically similar through the rest of the age ranges. This point is definitely worthy of further research.

This next chart is the deaths per 100,000. The 2018 figures were taken directly from the first CDC data reference. The 2020 figures were calculated by dividing the population by the total deaths in the second CDC data chart and multiplying by 100,000. The last column is calculated by dividing the 2020 numbers by the 2018 numbers, giving a percentage. At first glance, one might assume that since most percentages are less than 100, 2020 has a lower death rate. And one would not be the first to make that mistake. However, the data used to calculate the 2020 death rate only accounts for 46 weeks of data. Based on that, the percentages should be about 88 for an even death rate. In fact, the second data set is provisional, so would probably be slightly lower. When I get to normalizing the death rate to calculate excess deaths, I will actually use 84 percent for my calculations.

Now we get to the fun stuff. Here we have calculated excess deaths. We did this by first normalizing the total deaths to 2018 levels. I did this by dividing the population by 100,000 and then multiplying that by the 2018 death rate per 100,000 population. I then multiplied that by .84, to allow for the shortened and provisional data as discussed above. Next I subtract total deaths from this number. This gives an estimate of excess deaths. Interestingly, and maybe coincidentally, the excess deaths are fairly close to the total COVID deaths. Coincidental, because the number of excess deaths by age group do not correlate with the number of COVID deaths of that age group. The excess deaths are much larger than COVID deaths for the youngest of ages, with the difference decreasing as we approach 55. From 75 and up, the excess deaths are smaller than COVID deaths. The age group of 65-75 is interesting in that the number of excess deaths seem to jump up only for this age group, defying the trends we see with the other age groups.

While we can, and will, come to some hypotheses about the trends that this data shows, I want to caution first that these hypotheses are just that, and further research with more complete data would be needed to establish the validity of these hypotheses. The first hypothesis we come up with is that since excess deaths exceeded COVID deaths through age 75 by 41,610, there must be some other cause(s) of death that is elevated. Perhaps we will find that this reflects the negative effects of the COVID interventions. The second hypothesis is that the 23,497 COVID deaths above excess deaths in the 75 and up age groups are from deaths that would have occurred anyway. We may also find that deaths from COVID interactions also are reflected in the excess deaths in the 75 and up age groups, as well as that some deaths would have happened with or without COVID in the under 75 age groups. More careful scrutiny of the actual causes of death would be needed to prove or disprove either of these hypotheses.