Are we actually going to have swing-state polls this year with asterisks attached to the 65+ group? E.g.:

65 and older: Biden 51, Trump 47 ***ASSUMING 95% SURVIVE

I’m not ready for election models that incorporate the risk of mass death into the forecast. But if we’re going to be so ghoulish as to contemplate this, it’s important to bear in mind that Biden, not Trump, currently leads among senior citizens. I wrote about that yesterday. It’s the single most consequential demographic shift from 2016 to now, at least so far. If, God forbid, COVID-19 claims a few hundred thousand seniors this year, one would think right now it would take more of a bite out of the Democrats’ vote share than the president’s.

Academic researchers writing in a little-noticed public administration journal — Administrative Theory & Praxis — conclude that when considering nothing other than the tens of thousands of deaths projected from the virus, demographic shifts alone could be enough to swing crucial states to Joe Biden in the fall…

Johnson and his colleagues Wendi Pollock and Beth M. Rauhaus projected that even with shelter-in-place orders remaining in effect, about 11,000 more Republicans than Democrats who are 65 and older could die before the election in both Michigan and North Carolina.

In Pennsylvania, should the state return to using only social distancing to fight infections, over 13,000 more Republican than Democratic voters in that age category could be lost.

“The virus is killing more older voters, and in many states that’s the key to a GOP victory,” said one of the authors. Yes, normally that’s true, but not this year, at least not so far. I haven’t done a deep dive into the study itself but the authors seem to just assume that older voters will vote Republican because that’s what they did in 2016. (“According to the Pew Research Center, when data from the 2016 U.S. Presidential election was analyzed, the Republican candidate (Donald Trump) had an advantage with voters fifty and older…”) Politico notes that they also used a worst-case-scenario model of the death toll from the COVID epidemic, one that forecasts a number of death “orders of magnitude higher than what’s borne out so far in battleground states.”

Let’s do some back-of-the-envelope math to see if we can gauge a more realistic electoral effect from seniors dying. The worst-case scenario in the White House’s more plausible model of the epidemic imagines 240,000 Americans dying. Let’s assume that 75 percent of those deaths, or 180,000 people, comes from the 65+ group. Let’s also assume for ease of analysis that deaths among senior citizens from coronavirus are evenly distributed across the population. That would mean that Michigan, which has about three percent of the U.S. population, would see about 5,400 people aged 65 or older die of COVID-19. Now assume that either Trump or Biden wins the 65+ age group in Michigan by a comfortable 55/45 margin, a realistic best-case scenario for either man. (Trump topped Hillary in MI by four points in 2016.) That would mean that the winner of the senior demographic would have lost 2,970 votes due to coronavirus deaths while the loser would have lost 2,430 votes.

A grand difference of 540 votes. Michigan’s not going to come down to 540 votes. I think. I’m pretty sure.

Even if the authors are right that Trump will ultimately rebound and win the 65+ group, I’d imagine a lot depends on geography. Biden’s base of seniors is more likely to live in higher-density places like cities, Trump’s in lower-density rural areas. That leaves Biden’s base more exposed potentially. It’s hard to see how Republican voters in swing states will end up with more bad outcomes than Democratic ones.

Look at this new Fox News poll of Michigan and tell me that Trump, rather than Biden, has more votes to lose potentially due to deaths among senior citizens.

Biden’s 8-point advantage over Trump, 49-41 percent, is slightly larger than the poll’s margin of sampling error. However, both candidates remain below 50 percent and 10 percent of voters are still up for grabs…

The former vice president owes his lead to women (+20 points), non-whites (+63), and voters over age 65 (+19)…

But likeability could be the key. By a 10-point margin, more have a favorable (53 percent) than unfavorable (43 percent) opinion of Biden. Views of Trump are net negative by 8 points (44 favorable, 52 unfavorable).

Biden’s lead among women, a core Democratic group, is nearly identical to his lead among senior citizens, a core Republican group. That’s mind-boggling, and fatal to the president’s chances if he doesn’t fix it. A separate set of polls from Reuters today finds Biden leading in all three midwestern swing states, by three in Wisconsin, by six in Pennsylvania, and by eight in Michigan, the same as Fox has it. For all the criticism of Biden for having disappeared from American TV sets over the past month, his vanishing act seems to be only helping him. It’s turned the race into a pure referendum on Trump, with Biden as the proverbial generic Democrat, and voters haven’t liked what they’ve seen lately from the president in handling the COVID epidemic. If I were Biden I’d continue to lie low unless and until Trump’s fortunes rebound. Every interview he avoids that would have led him to lapse into incoherence if he’d participated is a net win.

Exit question: Is Obama going to be Biden’s attack dog against Trump on the coronavirus response? I’ve been skeptical, but after reading this tweet I’m more convinced.

While we continue to wait for a coherent national plan to navigate this pandemic, states like Massachusetts are beginning to adopt their own public health plans to combat this virus––before it's too late. https://t.co/Eb2Hz8H8vU

— Barack Obama (@BarackObama) April 22, 2020