Well, we don’t know yet because, among other reasons:
- 2020 isn’t over yet
- Forecasting fatalities for the remainder of the year is more difficult than assuming that deaths occur at a constant pace throughout the year (because they don’t; they typically peak in the winter months)
- There’s a lag (“from 1 week to 8 weeks or more, depending on the jurisdiction and cause of death”) in reporting deaths
- The CDC report to which you’re presumably referring doesn’t include the first few weeks of 2020 (until the week ended February 1)
Let me make these issues with the data you’re probably referring to more tangible:
I was able to locate an archived version of the CDC’s “Provisional Death Counts for Coronavirus Disease 2019 (COVID-19)” report from November 16 and compare it to the
most recent, last updated December 18. These reports include weekly total fatalities from both COVID-19 and all causes combined, along with the percentage of expected deaths for the week.
In the November 16 report, the CDC estimated the number of fatalities from all causes between January 26 and November 14 to be 2,487,350. That’s 40 weeks and an average of about 62,183 deaths per week, so let’s extrapolate that out to include the 4 weeks since. That’d put our estimate for what we’d see in the current report at 2,736,085 deaths, for the period between January 26 and December 12. The current report has 2,800,974 deaths in that period.
Where did all of those extra deaths come from? Well, deaths pick up in the winter months, as I mentioned, which is part of it, but the CDC already factors this into their weekly “expected deaths” calculation. But also, as I mentioned, there’s a lag in reporting deaths. The number of COVID-19 fatalities for the week ended November 14 in the November 16 CDC report was 240 and the number of all fatalities was 6,716 — just 12% of the expected deaths for the week. In the December 18 version, the numbers for that week have been revised to 9,500 and 62,571, respectively, or 114% of the expected deaths for the week, which is pretty well in line with adjacent weeks in the neighborhood of roughly 110–115%.
Setting aside the past few weeks with incomplete data, the most recent week we had with 100% or fewer of the expected deaths was that ended February 15. Being more generous, we can bump the threshold to 105% or fewer. The most recent week where that was true was that ended March 21. That means that each week since that ended March 28, we’ve seen greater than 105% of the expected fatalities. Each and every week.