Democrats are down to about a 25 percent chance to retain the Senate, according to the latest FiveThirtyEight forecast. A 25 percent chance isn’t nothing. The average National League hitter hit about .250 last year.
But what would a Democratic Senate win look like? Harry Enten and I pondered that question earlier. There are two basic scenarios.
First, Democrats could win in a squeaker. If the races behave more or less independently from one another, Democrats could keep the Senate by narrowly winning a number of states where they are only modest underdogs in the polls.
Alternatively, the Democrats could win in a shocker, with the polls proving to have a systematic bias against them. These things happen, although — as we’ve emphasized repeatedly — the polls have just as often had a systematic bias against Republicans.
We’ll consider the “squeaker” vs. “shocker” distinction in more depth in our final Senate update early Tuesday morning. But there’s another thing that would characterize a Democratic win. It would, very probably, require some electoral chaos.
We’ve talked before about the strong possibility the Senate’s outcome will not be determined on Tuesday or even by later this week. Instead, resolution may require one or more forms of “overtime”: Greg Orman deciding who to caucus with if he wins in Kansas, runoffs in Georgia or Louisiana, a slow vote count in Alaska and possibly one or more recounts.
Even as Republicans have consolidated their overall position in the polls, overtime remains more likely than not. While our model gives Republicans a 75 percent chance of winning, it shows only a 30 to 35 percent chance that a Republican victory would be called by midday Wednesday.
But if Democrats win, it will almost certainly require overtime. We show just a 5 percent chance the election will be called for them on Tuesday or early Wednsday.
Let’s do some arithmetic. (You can follow along on the ‘math’ tab on our Senate interactive.) Democrats need 50 seats. They start with the 34 they’re holding over from the current Senate. In another 11 states, they’re at least 95 percent likely to win according to our model; that gets them to 45.
Let’s also give Democrats North Carolina and New Hampshire, where they’re narrowly ahead in the polls. Democrats aren’t certain to win these, but if they don’t, all of this math won’t matter. These are must-win states for Democrats; in our simulations, they lost the Senate 98 percent of the time they lost New Hampshire and 97 percent of the time they lost North Carolina.
That gets Democrats to 47 seats. They need three more.
Iowa and Colorado are their next best choices if they want to avoid chaos. Although there have been disputes about where the polls stand in those states, the circumstances in Iowa and Colorado are otherwise normal — no runoffs, no major third-party candidates. They’ll be called on Tuesday or early Wednesday morning unless they’re so close as to be in recount territory.
That gets Democrats to 49 seats. They need just one more, and here’s where they run out of good options.
Mark Begich could win in Alaska. But that’s one of our chaos states. In the past two Senate races there, the outcome took 15 days to call. Begich would not only have to win, he’d have to win by a definitive margin — and that’s unlikely.
Kansas? If Orman wins, it’s a chaos state.
Georgia? It’s a chaos state unless one candidate gets past 50 percent of the vote to avoid a runoff. The Republican David Perdue is close to 50 percent in some polls, and he could easily cross that threshold. But a winning majority in Georgia on Nov. 4 has become a real long shot for Democrat Michelle Nunn. At this point, she’s probably hoping to facilitate a runoff and take her chances later on.
Louisiana? Chaos state. Democratic chances of holding the Senate have become grim enough that they might have to count on Sen. Mary Landrieu to pull through. But Landrieu is even more likely than Nunn to require a runoff; our model now gives Landrieu less than a 2 percent chance of winning an outright majority on Tuesday.
Arkansas? Kentucky? They aren’t chaos states, but the polls have broken against Democrats in both states, and there’s no time left to make up ground. Our forecast gives Sen. Mark Pryor just a 6 percent chance of winning in Arkansas and Alison Lundergan Grimes just a 3 percent chance in Kentucky.
So Democrats are very unlikely to win a “clean” victory on Tuesday or Wednesday. The messier things get and the longer they take to resolve, the better their chances.
Of course, Democrats have benefited from a little chaos before. After the 2008 elections, they gained a 60-seat super majority in the Senate only after Al Franken won a lengthy recount in Minnesota and Sen. Arlen Specter of Pennsylvania switched parties. In 2001, Sen. Jim Jeffords of Vermont defected from the GOP and began caucusing with Democrats, giving them majority control.
If Democrats do hold the Senate, it is likely to be very narrowly — probably either 50-50 or 51-49, counting two to three independents and Vice President Biden’s tie breaking vote. But as the Jeffords example proves, a 50-50 or 51-49 majority in the Senate is not so safe: Any party switch or any special election can potentially flip it.
A chaotic Democratic win — or a chaotic Republican one — could beget more chaos later on.