No, there weren't more excess deaths in 2017 than in 2020.

Perhaps you've seen Samuel H. Preston and Yana C. Vierboom's paper, "Excess mortality in the United States in the 21st century," published this April in Proceedings of the National Academy of Sciences of the United States of America.

It was linked on the Junior Ganymede with the comment, "2020 is the deadliest year on record since . . . 2017." Vox Day, referencing the same paper, said, "Got that? Fewer people died in 2020 than in 2017." I more or less trust the JG and VD, but I clicked through to the original paper and read the abstract, which seemed to confirm their summaries:

In 2017, excess deaths and years of life lost in the United States represent a larger annual loss of life than that associated with the [birdemic] in 2020.

There's no excuse for my letting this slip past my BS detector -- I knew it completely contradicted the statistics W. M. Briggs faithfully reports every week -- or for failing to read the whole paper, which turns out to be really short. I'm ashamed to say I just saw it as ammo I could use with friends and family members who trust The Science and accepted it without asking too many questions. In fact, it turns out that the abstract is misleading, and the paper is garbage, publish-or-perish busywork that contributes nothing to human knowledge.

"Excess deaths in the United States" means, in this paper, not deaths in excess of what would be expected in a "normal" year in the United States, but rather the degree to which death rates in the U.S. in a given year exceed those of five selected European countries in the same year. The paper's meaningless and useless "finding" is that the difference between American and European death rates in 2017 (with American rates being higher) was larger than the difference between American and European death rates in 2020.

This has nothing to do with whether more people died in 2017 or in 2020, or with how serious the birdemic was. In fact, it has nothing to do with anything, and it's not entirely clear why anyone thought it worth publishing.

Thanks to my father for reading the whole thing and informing me that it doesn't actually say what it says on the tin.