Which Party Will Win the Senate?

September 26, 2022 Which Party Will Win the Senate?

Early voting started in Virginia last Friday, but neither of our Senators, Warner or Kaine, are facing the end of a term.

Of the 100 seats in the Senate, 35 of them are up for election this cycle.

Of those 35, 21 are currently held by Republicans and 14 by Democrats.

Despite that seemingly Democratic advantage, Republicans and Democrats have about the same number of competitive seats. According to the latest Cook Political Report ratings, Republicans are defending three seats in the “lean R” category, one in the “toss-ups,” and one in the “lean D”, while Democrats are defending three “lean D” and two “toss-ups.”

So if you look only at toss-ups, Republicans would seem to have a two-to-one advantage. But since Democrats have to win only one of those three, the advantage is reversed (because Vice-President Harris is the tie-breaker if the Senate splits equally, as it is right now).

The three toss-ups are: Nevada, Georgia, and Wisconsin. Until last week, Cook considered Arizona one too, but they changed its status to “Lean D” last week. According to 270toWin, there’s also a fifth: Pennsylvania.

And some consider Ohio, North Carolina, and New Hampshire to be in play too. (Occasionally, even Florida and Colorado are mentioned, but that seems a bit far-fetched to me.)

Of those 8 races, to win control of the Senate, Republicans need to win 5.

To keep control of the Senate, Democrats need to win 4.

Looking at the polls, odds look good for Democrats. According to the conglomerate averages at FiveThirtyEight on September 21:

• in Pennsylvania, Fetterman is up by 9.5 points.
• in Arizona, Kelly is up by 8.7 points.
• in New Hampshire, Hassan is up by 6.6 points.
• in Nevada, Cortez Masto is up by 2.9 points.
• in Georgia, Warnock is up 2.1 points.
• in Wisconsin, Barnes is even.
• in Ohio, Ryan is down by .2 points.
• in North Carolina, Beasley is down by .3 points.

Using to the Cook spreadsheet, the polls predict these winners:

So if the election had been held on September 21, the Senate would have gone either 51/49 or 52/48, with Democrats gaining one or two seats (definitely Pennsylvania and possibly Wisconsin). Though given the closeness of North Carolina and Ohio (both are essentially even with differences of only .2 and .3), either of those could have tipped Democratic too, for as much as an unlikely but possible 54/46 majority:

But are the polls right?

I keep reading anxiety-driven articles (“Are the Polls Wrong Again?” “Will the Polls Overestimate Democrats Again?” “Pollsters fear they’re blowing it again in 2022“) insisting that there is no Democratic polling bias—just the relentless fear of one.

Is there a way to counter that fear?

According to Politico’s Stephen Clermont: “When voters dislike both parties, they will vote Republican unless Democrats compel them otherwise.” So he recommends looking not at the polling difference between candidates, but at just the Democrat. Based on the 18 Senate and presidential races from 2014-2020, if the Democrat polled:

• 49% or higher, the Democrat won.
• 48%, the Democrat won almost two out of three (63%).
• 45-47%, the Democrat won about one out of five (19%).
• 44% or lower, the Democrat never won.

I see at least two problems with this approach. First, merging three percentages (45, 46, 47) allows for a lot of wiggle room for one of the most critical questions (when are the odds 50/50?), and, second, are 18 races really enough of a data set to draw meaningful conclusions? Also, I’m not sure Clermont’s claim (“When voters dislike both parties, they will vote Republican unless Democrats compel them otherwise”) is true generally, and “dislike” may not capture attitudes toward some of this year’s MAGA extremists. As McConnell said, the GOP is suffering from “candidate quality.”

Still, this seems like a useful exercise. According to the conglomerate polling averages at FiveThirtyEight on September 21:

• in Pennsylvania, Fetterman: 50.9%
• in New Hampshire, Hassan: 49%
• in Arizona, Kelly: 48.8%
• in Georgia, Warnock: 47.9%
• in Wisconsin, Barnes 47.6%
• in North Carolina, Beasley: 45%
• in Ohio, Ryan: 44.5%
• in Nevada, Cortez Masto: 44.2%

The top two are clear Democratic winners, and since 48.8% is pretty close to 49%, I’d call that three. Democrats need just one more. Since both 47.9% and 47.6% round up to 48%, both should have close to two-to-one odds, with either one likely to produce the Democrats’ necessary fourth win. If they do win both, that tilts the Senate 51/49, with the remaining three seats likely going Republican.

Back on the Cook spreadsheet:

And that’s the most pessimistic prediction. Even though Cortez Masto’s 44.2% is at the bottom of the list, she’s leading in polls by about 3 points, and FiveThirtyEight gives her 54% odds of winning, the same as Warnock who, according to Clermont, should win.

According to FiveThirtyEight predictions, Democrats should gain one seat for a 51/49 majority:

Though 54% odds on either of the two tightest races are only slightly better than a coin toss, for Republicans to take control of the Senate, they need to with BOTH coin tosses. That’s possible, but not likely.

And, again, that’s not according to the polls. If you look at just those numbers (now updated as of Sunday September 25th), the Democrats would win four additional seats for a 54/46 majority:

That’s the kind of too-good-to-be-true prediction that has Democrats so paradoxically terrified that all positive predictions must be false. Since two of those predictions are close to even odds (51 and 53), and two are only slightly better (55 and 56), I wouldn’t assume anything about those outcomes. But even if Democrats lost all four, they would still keep control of the evenly divided Senate (because of Warnock, who is predicted to win with two-to-one odds).

Last January, I wrote this blog entry: “Which Party Will Win the Senate in November 2022?“

Because Arizona, Nevada, and Georgia all lean Republican, the key question then was:

“Can Democratic Senators Kelly, Warnock, and Cortez Mastro hold their seats in unfavorable territory? If any two lose, the Republicans control the Senate. If only one of those Democratic senators loses, control of the senate then hinges on Pennsylvania. Because Pennsylvania’s Republican senator is retiring, the seat is wide open, with over a dozen candidates competing in both party primaries. Personally, I’m betting the current lieutenant governor, John Fetterman, wins both the Democratic primary and the general election.”

Nine months later, Kelly and Fetterman both look like safe bets. And while Warnock and Cortez Mastro remain unknowns, they are joined by Wisconsin’s Republican incumbent Johnson.

Republicans have to win all three tossups, plus two additional battleground states, Ohio and North Carolina, in order to win the Senate.

Democrats have to win only one of those five.

As I said last January: “The answer is of course unknowable. It will probably remain unknowable right up to the election. All we can know right now are the factors that will shape the outcome.”

While that remains true, the factors look very good for Democrats.

• Leave a comment
• Posted under Uncategorized