I've had several people bring up this claim recently, and I've seen it at least twice on RFM. You know who you are. It's a subtopic under "Covid-19 is all a hoax/no worse than the flu/falsely declared as cause of death" family of conspiracy theories.
I heard it enough times that I decided to do a little investigating to find out where this bit of nonsense is coming from. Learned a few interesting facts in the process.
First, let's establish a baseline. How many people die each year? What is the normal statistical variation from year to year? Is the death number increasing, decreasing, or static? How many people die from the flu each year? How widely does that vary? Where are numbers available, and how long before they finalize?
I didn't really know precise answers to any of those questions.
Year- US deaths - increase from previous year
2009..,2,437,163
2010..,2,468,435...31,272
2011...2,515,458...47,023
2012...2,543,279...27,821
2013...2,596,993...53,714
2014...2,626,418...29,425
2015...2,712,630...86,212
2016...2,744,248...31,618
2017...2,823,503...69,255
2018...2,839,205...25,702
2019...2,854,838...15,633
Source: wonder.cdc.gov , Underlying cause of death, 1999-2019. I can't give a precise link to this data because you have to click an "I agree" to their rules on use of the data before they let you in. The rules are things like you can't use the data to try and find out a specific person's identity.
the default report is for all deaths, 2018. You go through their blizzard of choices, find the year selector, and select the range of years you want. There are many many ways to slice and dice the data - by year, age, ethnicity, state, region, cause of death, yada yada.
OK, so the number of deaths increases every year, consistent with a rising population. It is just below 3 million right now.
The increase over the last decade was about 43,000 per year, range of 16K to 86K. That is actually a pretty narrow window of variation from year to year. Most of the variation depends on how bad the flu season was in a given winter.
BTW, the 2019 numbers were released just a week or two ago - Dec 2020.
OK, so how many flu deaths are there per year? Not as many as I would have thought. The CDC is very cautious about finalizing flu deaths. their table lists the last three years as still estimated numbers. Note also that the window of uncertainty on the number of deaths is pretty wide.
https://www.cdc.gov/flu/about/burden/index.htmlThe table is a bit complicated, so I'll only list the core number of deaths per year (thanks to Lot's Wife for typing them into another recent thread!)
https://www.exmormon.org/phorum/read.php?2,2349379,2350473#msg-2350473Influenza deaths, US
2010-2011 deaths: 37,000
2011-2012 deaths: 12,000
2013-2014 deaths: 38,000
2014-2015 deaths: 51,000
2015-2016 deaths: 23,000
2016-2017 deaths: 38,000
2017-2018 deaths: 61,000
2018-2019 deaths: 34,157
You'll note that the flu does not make a huge difference in the overall death total, and it is pretty much consistent with the yearly variations listed in the first table of overall deaths. Flu season tends to peak in January, so whether the season was early, late or normal affects which year will most feel the impact. One, two, or zero peak flu seasons can occur in a single given year, depending on the timing.
OK, so there's the baselines and past data points.
So what, if anything was wrong with the memes reporting that the death toll this year was the same as 2019?
I saw a report from Sarah Sanders dated Sept 24. Reuters investigated one from November. The one Reuters investigated had four errors, three of which they mention in their article, one that they missed.
Reuters found that the data was from a CDC report on Covid, which started with the week ending Feb 1, 2020, so January data was missing. Leaving out a month would of course lower the average death toll per month, and a projected death toll for the year.
the second error they mentioned was no allowance was made for normal reporting lag. CDC says there can be significant increases in reported deaths for up to 8 weeks, though the biggest changes happen in the first 2 weeks. If the state Vital Statistics office is very busy, or hospitals/doctors, or if autopsies need to be done, all can delay issuance of death certificates.
The third error was that the Reuters-investigated meme used an incorrect number for the number of deaths in 2019, that was too high by 45,000. I don't know where the author got the number, but the official number of 2.84 million was just released in the last week or two. The one the meme used was 2.9 million. That of course makes it easier to claim 2020 was not higher than 2019, since they were using a 2019 number that was too high to start with. I do not know where the 2.9 million number came from, but it was wrong.
The error Reuters does not mention is that the authors of the memes did a linear projection of the average death toll for the period, when the actual death toll over the course of the year is not linear. It peaks in January, and has a long, low period from mid spring through mid fall, when it starts heading up again to the next January peak.
This would not make a lot of difference on the meme created in mid November, but it would on the one from September. A straight line projection from September would miss all the increase in deaths for the last quarter of the year.
To recap, data was missing in January, perhaps deliberate, perhaps misreading of CDC reports on covid deaths. Either way, it was missing. Data was missing in the last month of each report because of normal reporting lag time. The 2019 estimated death count was too high. None of these errors were taken into account when calculating the projected death toll for 2020. All would produce an unrealistically low projected death count for 2020 compared to 2019.
https://www.reuters.com/article/uk-factcheck-chart-us-death-figures-2020/fact-check-chart-does-not-present-accurate-us-deaths-figure-for-2020-idUSKBN2872MVhttps://www.politifact.com/factchecks/2020/dec/11/facebook-posts/chart-comparing-2020-us-death-toll-previous-years-/https://apnews.com/article/us-coronavirus-deaths-top-3-million-e2bc856b6ec45563b84ee2e87ae8d5e7Is there a good place to see the "excess deaths" caused by covid-19?
Yes, CDC reports weekly on excess deaths, which are deaths above and beyond what would be expected for that week in the year. As noted in my first table, the expected deaths are quite stable from year to year, with much of what little variation there is (other than simple population growth) coming from flu season.
https://www.cdc.gov/nchs/nvss/vsrr/covid19/excess_deaths.htmThis is a rather long web page. About halfway down, there is a bar graph of the last 4 years, where the bars are weekly death counts. You can mouse over the bars and get detail stats for each week. The red line is the cutoff for excess deaths. It allows a little wiggle room for normal statistical variation, before they start considering deaths "excess deaths". Note that the 2018 flu peak was the only period where there were ANY excess deaths in the last 4 years, until late March, 2020, at which point there was a serious spike.
Note that the last few bars of that graph are also going to suffer from reporting lag, and are likely to increase over the next few weeks.
The final numbers for 2020 will likely take a year to be released, perhaps longer. Current estimates are that 2020 deaths will jump by 350,000 over 2019. That completely blows the doors off the growth numbers for the previous 10 years.
Covid is an actual disease. There are real people dying, and it is not old people who were about to die anyway. These are "excess deaths", above and beyond what is normal.
While this is not true for all of 2020, right now, Covid-19 is the number 1 cause of death in the US, barely beating out both heart disease and cancer.
I doubt if I killed that zombie meme, but I am pretty sure it is badly bruised. :-/