Tuesday, December 22, 2020

Maybe Getting 74 Million Votes Is Not Particularly Impressive

Note to readers: Before turning to my column for today, I want to provide a link to Professor Dorf's new Verdict column, which was published this morning: "Odysseus, Avocados, and Election Litigation Timing."  The column discusses yet another after-the-fact Supreme Court filing by the losing Trump campaign, with Professor Dorf focusing in particular on the doctrine of laches, which denies relief when a case was filed unreasonably late.  He also argues that the opposite doctrine, ripeness, should be loosened, for reasons that I encourage readers to discover by reading the piece.

I cannot resist adding a comment that is consistent with Professor Dorf's analysis.  The Trump team's current claim is that they are not too late because Biden has not yet been inaugurated.  Only then, they say, will it be too late.  But why should we take them seriously?  They spent the first five weeks after Election Day saying that they had plenty of time before the Electoral College's vote on December 14, at which point it would be too late.  Then they moved on to January 6, the day that both houses of Congress will formally accept the electoral votes.  Not even waiting until that failed, they now say that January 20 is the day.
I fully expect them to argue after January 20, however, that their inevitable further challenges are even more timely, because the country will at that point have a president who is -- in their fevered minds -- not legitimately elected.  No laches apply, they seem sure to argue, because every day under Not-President Biden is another day in which Americans must suffer the consequences of a China/Venezuela-inspired stolen election.  No day will be a drop-dead date for these people, and they will want to re-litigate this forever.

By Neil H. Buchanan

If it is healthy to stop and ask whether one's most sincerely held beliefs need to be reassessed, it must also be healthy to ask whether one's reassessment should itself be reassessed.  That is where I find myself today with regard to my response to the number of people who voted for Donald Trump in the 2020 election.  In 2016, his vote total was 62,984,828, just a hair under 63 million.  In 2020, his total was 74,216,747, or almost eleven-and-a-quarter million more than last time.  That is, to put it mildly, disheartening.

Many have noted that Trump seems obsessed with how big his number was this time around.  When Senator Mitch McConnell at long last recognized President-Elect Joe Biden's win, Trump frantically tweeted that McConnell should remember Trump's "75 million votes."  Not only is that a generous round-up of the total (consistent with adding imaginary floors to Trump Tower and imaginary people to Trump's 2017 inauguration non-crowd), but it is among the weirdest one-sided invocations of data that I have ever heard.

Late-night comedian Seth Meyers captured it best, saying that this would be like Trump hearing a word problem in high school algebra but not waiting for the whole question.  Q: "Suppose a train leaves the station at 120 miles per hour, ..."  Trump interrupts: "Oh, wow, that's a fast train.  There are no other trains that are that fast."  Q: "... and a second train leaves the station at 130 miles per hour."  Trump: "Oh, forget about that other train.  My train is fast!"

That is both hilarious and accurate in lampooning Trump, of course.  Joe Biden received 81,268,867 votes, 7+ million more than Trump.  But Trump is impressed with his 74.2 million, and he is convinced that this is what McConnell should care about.  My question today, however, focuses on the question of whether Trump's number -- which is the second-highest number of raw votes ever received by an American presidential candidate -- is meaningfully "big" in the way that most pundits quickly decided that it is.
Short answer: Probably not.  There is a statistical apples-to-oranges problem that is leading people to overstate Trump's influence in a way that could distort political discussions going forward.
To be clear, I am one of the people who initially was stunned by Trump's vote total.  On November 12, I wrote a column here on Dorf on Law in which I struggled to understand what I had misunderstood about people who would vote for Donald Trump this year.  I simply never imagined that he could receive that many votes, and I wanted to know how it was even possible.

Taking Trump's total votes, which on November 12 stood at 72.4 million (because votes were still being counted), I asked how he could have received more than 9.4 million additional votes above his 2016 total.  Today, the difference is even larger, which suggests that my initial sense of being chastened ought only to have intensified.

Or perhaps the right word is not "chastened," because the conclusion that I drew was more cynical than my previous theory.  Going into November 2020, my working hypothesis had been that Trump's 2016 total of 63 million votes was the upper limit of what he could receive in 2020.  His base is older, which means that they were dying off more quickly than anti-Trumpers (even before COVID-19).  More importantly, he had done everything possible to alienate people by the millions, meaning that he had most likely driven at least some people away.  ("Wow, I thought I'd give him a chance, but was that a mistake!")
Yes, his remaining base of supporters continued to be highly motivated, but it seemed to be a finite, shrinking group.  The pre-pandemic economy was nowhere near as strong as Trump and the Republicans claimed, and post-pandemic economic pain -- to say nothing of hundreds of thousands of deaths -- made it even less likely that he could get anywhere near his 2016 total.  The pre-election polls seemed to bear this out.

So why the big upward jump?  Where did he find people to replace all of those he had lost over the last four years, plus a net increase of eleven-plus million Americans who voted for this year's loser?

As I explained in that column last month (but not using this metaphor), I soon realized that the illogic of being a Trump supporter is a free-floating virus that could infect not only those who were laid low in 2016 but could take out other susceptible people at any time.  In short, it never made sense in the first place that those 63 million people went to the polls for Trump in 2016, so I was mistaken not to think that any number of additional people might suddenly come down with a case of Trumpitis now.
After all, it is not as though we could be certain that he had picked up every gullible, racist or racist-comfortable vote in 2016.  Maybe his toxic actions over the last four years had somehow made sense to eleven million or so Americans in ways that his pre-2017 toxic actions did not.  There was no vaccination for this toxic virus, and if some people who in 2015 could never have imagined voting for Trump ended up supporting him in 2016, then certainly some number more could reach the same inexplicable conclusion in 2020.

"Chastened," then, truly is not how I felt.  I was confused, infuriated, scared, befuddled, and more.  No matter the label, my worry as of November 12 was that somehow Trump's support had grown in four years, even though it should have shrunk.  (But again, it should not have existed at all, in anything like a decent or minimally rational world.)  And like a lot of other observers, I found this very large and expanded number of Trump voters to be even more worrisome than the 2016 total was (which was plenty worrisome).

About a month later, however, I saw a Washington Post op-ed that has changed the way I think about this year's election.  The column, "Let’s take a closer look at Trump’s supposedly intimidating 74 million vote total," was written by Matt Bai, whose name on a byline does not make me sit up and say, "Ooh, I bet this will be good!"  That is not to say that Bai is in the category of a George Will, Ross Douthat, or Frank Bruni (all anti-Trumpers who nonetheless manage to continue to reset the bar for insufferability), or anything close to it.  I am saying only that I was not inclined to agree or disagree with the column when I began to read it.

In a nutshell, Bai points out that this year's election was categorically different from all previous elections, which should obviously have led us at least to think that historical norms are not reliable benchmarks to which we can compare this year's results.  This year is simply different.

Bai sets up the story vividly:

"Liberals have been obsessing on this 74 million vote number — far more than they have on Biden’s historic showing — since the election. To them, it’s shocking evidence that Trumpism only grew in popularity over the four years in which Trump himself was conducting a national seminar in mismanagement and bigotry.

"If Trump could earn more votes than any sitting president ever, the thinking goes, then clearly his movement is here to stay, and he could even get himself reelected in 2024."

Again, I am one of those liberals to whom Bai refers.  But as he goes on to point out, many Republicans are similarly obsessed, "trembling like a bunch of chihuahuas in the face of Trump’s awesome power."  There is a reason that the next-most-powerful Republican in Washington, Mitch McConnell, never quite said directly that the people had chosen Biden.  True, a big part of that was strategic, keeping Trump's voters on the hook for the Georgia runoffs, but that was hardly the whole story.

Bai then makes the crucial point that the raw vote totals might not -- in fact, almost certainly do not -- mean what people think they mean, because "this was the first election that featured multiple ways of casting a ballot pretty much everywhere, including early and by-mail voting. As a result (and also because of heightened emotions in the electorate), turnout was the highest it has been in more than a century, clocking in at more than 66 percent."
Remember that we are now talking about raw numbers.  74 million is supposed to be more impressive than 62 million, even though both totals were less than Trump's opponents' totals.  Crucially, both candidates' raw totals went up in 2020, even though Trump disparaged mail-in voting: "Every time another batch of mail-in votes showed up, bringing Biden that much closer to victory in key states, Trump’s haul grew as well."
Continuing to quote extensively from Bai's column:

"According to the University of Florida’s Elections Project, about 20 states reported the number of mail-in ballots by party registration. In those states, about 10 million Republicans sent back completed ballots, as did about 9 million nonaffiliated voters.

"Even if a sizable chunk of those independent voters didn’t support Trump, and even if we assume that some percentage of Republicans would have shown up to the polls if they hadn’t had an option, that’s probably millions of votes that Trump wouldn’t have gotten in a traditional election — and that doesn’t include data from giant states such as Texas, Ohio and New York."

Note also that California, which went heavily for Biden (11+ million votes), nonetheless delivered 6 million to Trump -- about the same as Alabama, South Carolina, Kentucky, Louisiana, and Arkansas combined.
Bai adds: "Absent the nontraditional methods of voting that Trump castigates as fraud, it’s reasonable to think his vote total would have been much closer to his 2016 showing — which, for an incumbent president with all the advantages of the office, isn’t so impressive at all"; and this zinger: "The last president to lose an election in a two-way race, Gerald Ford, claimed 48 percent of the vote in 1976, compared with Trump’s 47. No one remembers Ford as having demonstrated an iron grip on his party on his way out the door."
It seems likely that many of the voting reforms of 2020 will be retained in future years, simply because people enjoy the convenience of not voting on Election Day, and because Trump's complaints do not line up with people's experiences.  Florida, for example, continued to run an extensive mail-in voting operation (which I used with ease), and Trump won my state by more than three percent.  If so, vote totals will be higher in all future elections.  If not, then 2020 will be an outlier, but we should not view Trump's 74 million votes this year as a big deal either way.

Yes, I am still terrified that there are over seventy million Americans -- or, for that matter, even 70 Americans -- who would vote for Trump.  That is not today's question, however.  If Bai's analysis is correct, then 2020 did not witness the sudden emergence of 11+ million new Trump supporters.  It is quite likely that Trump's supporters, like other candidates' supporters, included many people who would have voted for him in 2016 but found the process inconvenient or worse.  Even fire-breathing MAGA people can be fired for not showing up at work on Election Day, so there is no reason to think that the actual vote total in 2016 represented everyone who would have voted for him.

Is this conclusion appealing to me because it supports my status quo ante?  Yes, but not because "it shows I was right all along."  I was perfectly willing to be wrong and to figure out why.  And even so, my new explanation -- that the Trump virus can infect new people just as shockingly as it infected people four years ago -- seems like an important possibility to bear in mind.

No, the reason that Bai's explanation resonates so strongly is because it undercuts -- in a completely objective way -- the idea that Trump's increased vote total this year constitutes proof that he is a historically unique force to be reckoned with.  He is dangerous (more so every day), but his too-large base of support almost surely did not meaningfully expand in 2020.
Calling our situation "only as scary as in 2016" is not as good as "no longer scary," but it is better than "even more scary today."  In a world where I am looking for reasons to be optimistic without being pollyannish, this is a win.


CEP said...

One wonders how these totals would look absent well over a decade of demographically/partisanly biased voter suppression efforts that are not reflected in the vote totals. Consider, for example, several hundred thousand Florida voters who "should have been" reenfranchised, but who were successfully prevented from registering to vote in the first place (by efforts properly labelled a "poll tax" by the US District Courts, but treated as a nonjusticiable political question in all but name by the Eleventh Circuit).

I'll just go off muttering under my breath about Baker v. Carr and the time for all deliberate speed being (long) past.

Michael A Livingston said...

Yes and no. Probably Trump’s votes were inflated by mail-in voting, as were Biden’s. But that’s the point. The theory was that a higher turnout, especially among minorities, would lead to a larger Democratic win. But it didn’t: it largely canceled out, and indeed Republicans improvised at the down ballot level. So it’s hard to predict what happens in the future, either way. Overall I would say it’s a pleasant surprise for the R’s.