Election 2022: My Final US Senate Predictions ...

Why so early? Two reasons:

1. It's not fun waiting until the last minute, like the day before election day. Yes, there can be "October surprises," but it's not really much of a prediction if you're using exit polls.
2. Early voting has already begun in many if not most states, and it's going gang-busters. The bulk of the vote in those states will likely be cast in the next week or so (a note on that below, though).

My predictions for what were considered "toss-up" states earlier in the year:

Arizona: Incumbent Democrat Mark Kelly in a walk. This state never really looked like a "toss-up" to me. It was always "leans hard Democrat." The GOP's nomination of Trumpy Peter Thiel protege Blake Masters sealed the deal. An early October poll from OH Predictive Insights has Kelly up 46-33 over Masters, with Libertarian Marc Victor at 15% and only 7% "unsure." Yes, the third party vote tends to diminish as election day draws near, but the FiveThirtyEight polling average (which includes only Kelly and Masters) has Kelly up by 5.7 points.

Georgia: This one's slightly more complicated. FiveThirtyEight has incumbent Democrat Raphael Warnock up on Republican challenger Herschel Walker by 3.7% on average; RealClearPolitics has Warnock up 2.4%. So, close. BUT! In Georgia, a majority is required to avoid a runoff, and neither candidate is polling at 50% ... and Libertarian Shane Hazel is doing God's work. He forced a runoff two years ago, and last week he debated Warnock with Walker pulling a no-show. I expect Warnock to get more votes than Walker, but for the race to go to a runoff. And I expect Warnock to win the runoff, for two reasons. The first is that the Democrats will pour big money into the runoff; the second is that Walker is clearly a lost cause scandal machine and Republicans likely won't invest in trying to save him, especially with (as we shall see) doing so offering them no chance of getting the Senate majority they were hoping for.

Nevada: Both FiveThirtyEight and RCP have the race between incumbent Democrat Catherine Cortez-Masto and Republican challenger Adam Laxalt marked as pretty much a statistical tie with Laxalt up by 0.2% and 1.2% respectively. I predict a Cortez-Masto win.  It's a "get out the vote" game here. Democrats are traditionally better at GOTV. GOTV efforts are more effective in urban areas, and the Democratic vote is concentrated in Las Vegas/Clark County while the Republican vote is dispersed in smaller towns and rural areas. And, again, early voting (see note below).

Pennsylvania: No incumbent, open (previously Republican) seat. Yes, Democrat John Fetterman had a stroke earlier this year and has residual "auditory processing" problems. Yes, Republican Mehmet Oz is making the most of the "health issue." Yes, RCP has Fetterman up by only 2.4 points (FiveThirtyEight says 4.4). I predict that Fetterman wins this one. The people who would let Fetterman's health negatively influence their vote weren't going to vote for him anyway. The people who were going to vote for Fetterman are MORE motivated and sympathetic, since he's clearly not a vegetable or anything. Fetterman's a Pennsylvania mayor/lieutenant governor. Oz is a carpet-bagging TV quack with a foreign-sounding name who edged out the Trump base's preferred candidate in the primary. And, again, early voting.

Wisconsin: I didn't expect this one to be a nail-biter. I expected incumbent Republican Ron Johnson to go down hard to Democratic challenger Mandela Barnes. The polls say differently: FiveThirtyEight's average has Johnson up by 2.9%, RCP's by 2.8%. The power of incumbency is strong, there is no in-person early voting, and a witness signature (not especially inconvenient, it doesn't have to be notarized, but any inconvenience reduces vote-by-mail "turnout") is required for absentee ballot mail voting. I predict Johnson will pull off re-election. I could be wrong. If I am, I suspect it will be because really terrible weather shows up that depresses rural voting more than urban voting.

So: To get a Senate majority, Republicans need a net gain of one seat. If all the "non-toss-up" states go as expected, my prediction has the Republicans losing one seat. If Cortez-Masto wins in Nevada, then even managing to pull Herschel Walker through a runoff in Georgia won't get them to a majority.

Outlier:

Ohio: The polls have this race -- an open seat -- statistically tied. FiveThirtyEight has Republican JD Vance up 0.8% over Democrat Tim Ryan. RCP has the margin at 2.8%. Vance has proven to be a pretty poor candidate, but he's running for a previously Republican seat, as a born-again Trumper, in a state that Trump carried 53% to 45% versus Joe Biden in 2020 and 51% to 43% versus Hillary Clinton in 2016. It looked like things might go a different way earlier in the year, but now I predict that Vance is going to the Senate.

OK, so let's talk about my emphasis on early voting, especially voting by mail.

At the moment, it helps Democrats the most. That wasn't always the case. Republicans used to be all for it because, among other reasons, it makes it easier for their rural base voters to "turn out."

But, you guessed, it, Trump. He convinced a lot of Republicans that voting by mail was just a scam to help Democrats and discouraged them from doing it. Go to the polls on election day, he said. Cast your ballot personally, he said.

So, let's make some not terribly unreasonable estimates here: 50% of likely voters are Republicans. 30% of those (that is, 15% of the voter pool) still distrust voting by mail two years after Trump's opposition to it may well have cost him a presidential election. And 10% of those voters -- that is, 1.5% of potential Republican votes -- end up not voting.

Overslept and didn't have time to vote before work. Babysitter called, sorry, can't cover for you after work. Blowout fight with the spouse, went to the bar to drink the troubles away. Flat tire on the way to the polling place, screw it. Just. Didn't. Really. Feel. Like. It.

In a close election, 1.5% is the ball game. Early mail voters can put their ballots in the mailbox tomorrow if they forgot to yesterday. Election day voters who don't show up to vote on election day are done.

So, yes, I think it makes a difference, and that difference is in favor of Democrats right now, because Trump is the gift who keeps on giving.