We’re now into the final three weeks of the 2016 elections (at last!), and as Donald Trump’s campaign implodes and Hillary Clinton solidifies her lead, word has it that the Clinton Campaign is pouring money into strategic swing states with a specific eye towards helping down-ticket races.
Winning the White House is great. Winning with a landslide mandate would be wonderful, but useless if a Republican Congress pretends the mandate never happened. Getting the Senate back would at least facilitate judicial confirmations that have stalled, among other things. And as some forecasters are now revising Democratic pickups in the House to upwards of 25 seats, winning the 30 seats necessary to make Nancy Pelosi Speaker again no longer seems completely impossible.
The forecasters are predicting a narrow Democratic victory in the Senate, but it could be close. And they aren’t all entirely in agreement about how close it may be.
Quick recap for those new to my week aggregator series: Every day I scoop up the forecasters’ overall and state-by-state predictions collected by The New York Times’ Upshot. They use data from five data modelers (the New York Times, FiveThirtyEight, DailyKos, Princeton Election Consortium, and The Huffington Post), three knowledgeable experts (the Cook Political Report, the Rothenberg & Gonzales Political Report, and Sabato’s Crystal Ball), and the PredictWise betting markets. I get the average of the forecasters’ opinions, and once a week I write this diary about Aggregating the Senate Aggregators so that we can see what their consensus opinion is. (On Fridays I also do a diary for Aggregating the Presidential Aggregators.)
Here’s where we stand this week.
The Big Picture
The Huffington Post continues to remain the most pessimistic about Democratic chances of taking control of the Senate, giving them only a 40% chance. But everyone else gives them more than a 50% chance, ranging from 57% for the New York Times and Daily Kos to 77% for FiveThirtyEight. The average for all of them is 61%.
States Up for Election
Democrats are only defending 10 of the 34 seats up for election this cycle, and only one of them, that of retiring Senator Harry Reid of Nevada, is at any risk of flipping to the Republicans. The Republicans hold the other 24 seats, many seized in vulnerable blue states in the Republican wave of 2010. (This is why we have to vote in off-year elections, people!)
I do this chart every week but it’s helpful for people joining us for the first time.
2016 U.S. SENATE ELECTIONS
U.S. SENATE |
NOT UP THIS YEAR |
UP FOR ELECTION |
PROBABLY SAFE |
IN PLAY |
REPUBLICAN CAUCUS |
30 |
24 |
12 |
12 |
DEMOCRATIC CAUCUS * |
36 |
10 |
9 |
1 |
TOTAL |
66 |
34 |
21 |
13 |
* The Democratic Caucus includes two independents who caucus with the Democrats, Angus King of Maine and Bernie Sanders of Vermont. As there’s no reason to believe both won’t continue to caucus with the Democrats, I will continue to lump them in with the “Democratic Caucus” for shorthand.
Republican Seats Probably Safe: Alabama, Alaska, Arkansas, Georgia, Idaho, Kansas, Kentucky, North Dakota, Oklahoma, South Carolina, South Dakota, and Utah
Republican Seats in Play: Arizona, Florida, Illinois, Indiana, Iowa, Louisiana, Missouri, New Hampshire, North Carolina, Ohio, Pennsylvania, and Wisconsin
Democratic Seats With Only Democratic Candidates: California
Democratic Seats Probably Safe: Colorado, Connecticut, Hawaii, Maryland, New York, Oregon, Vermont, and Washington
Democratic Seats in Play: Nevada
Forecaster Predictions in the Number of Senate Seats
Here’s the forecasters’ consensus opinions on Senate control, and how that’s changed over time.
Zooming in on just the 2016 races makes it a little clearer:
As you can see, it appears that some of the leaners on both sides are solidifying, but there’s still a fair amount of volatility in the middle.
Let’s look at the actual numbers over the course of this week.
2016 SENATE RACE PREDICTION FORECASTS THIS LAST WEEK
SENATE RACE |
10/13 |
10/14 |
10/15 |
10/16 |
10/17 |
10/18 |
10/19 |
SOLID DEMOCRATIC |
9 |
9 |
9 |
9 |
9 |
9 |
9 |
LIKELY DEMOCRATIC |
1 |
1 |
2 |
1 |
1 |
2 |
2 |
LEAN DEMOCRATIC |
2 |
2 |
1 |
2 |
2 |
1 |
1 |
TOSSUP — DEM |
1 |
1 |
1 |
1 |
2 |
2 |
1 |
PURE TOSSUP |
1 |
2 |
1 |
1 |
1 |
1 |
2 |
TOSSUP — REP |
2 |
1 |
2 |
2 |
2 |
2 |
2 |
LEAN REPUBLICAN |
2 |
2 |
2 |
2 |
2 |
2 |
2 |
LIKELY REPUBLICAN |
3 |
3 |
2 |
3 |
3 |
3 |
2 |
SOLID REPUBLICAN |
13 |
13 |
14 |
13 |
13 |
13 |
14 |
Let’s roll it up to see where things changed.
2016 SENATE RACE PREDICTION FORECASTS THIS LAST WEEK
SENATE RACE |
10/13 |
10/14 |
10/15 |
10/16 |
10/17 |
10/18 |
10/19 |
DEMOCRATIC |
12 |
12 |
12 |
12 |
12 |
12 |
12 |
TOSSUP — DEM |
1 |
1 |
1 |
1 |
2 |
2 |
1 |
PURE TOSSUP |
1 |
2 |
1 |
1 |
1 |
1 |
2 |
TOSSUP — REP |
2 |
1 |
2 |
2 |
1 |
1 |
2 |
REPUBLICAN |
18 |
18 |
18 |
18 |
18 |
18 |
18 |
As with last week, all of the movement has been within the Tossups, leaning this way and that. Let’s see how that affects control of the Senate.
FORECAST FOR THE U.S. SENATE AFTER THE 2016 ELECTIONS
U.S. SENATE |
10/13 |
10/14 |
10/15 |
10/16 |
10/17 |
10/18 |
10/19 |
DEMOCRATIC CAUCUS |
47 |
47 |
47 |
47 |
47 |
47 |
47 |
TOSSUP — DEM |
1 |
1 |
1 |
1 |
2 |
2 |
1 |
PURE TOSSUP |
1 |
2 |
1 |
1 |
1 |
1 |
2 |
TOSSUP — REP |
2 |
1 |
2 |
2 |
1 |
1 |
1 |
REPUBLICAN CAUCUS |
49 |
49 |
49 |
49 |
49 |
49 |
49 |
Argh! I can’t stand how close it is. I’m ready for some trickle-down effects and some Clinton coat-tails.
Here’s what the forecasters are saying about the individual Senate races. Remember, Democrats need a net gain of 4 to bring it to a 50/50 tie with VP Tim Kaine as the tie-breaker vote, or a net gain of 5 to have an outright majority.
CURRENT FORECASTS FOR U.S. SENATE RACES TO WATCH
Senate Race |
All
Pundits
|
Data
Modelers
|
Dem.
Probability
|
Impact |
Illinois |
Likely Dem |
Likely Dem |
91% |
Flip to Dem |
Wisconsin |
Likely Dem |
Likely Dem |
87% |
Flip to Dem |
Indiana |
Leaning Dem |
Leaning Dem |
73% |
Flip to Dem |
Nevada |
Tossup |
Tossup — Dem |
61% |
Hold Dem |
Pennsylvania |
Tossup |
Tossup |
53% |
Flip to Dem |
New Hampshire |
Tossup |
Tossup |
47% |
Hold Rep |
North Carolina |
Tossup |
Tossup — Rep |
38% |
Hold Rep |
Missouri |
Lean Rep |
Lean Rep |
31% |
Hold Rep |
Florida |
Lean Rep |
Lean Rep |
17% |
Hold Rep |
Louisiana |
Likely Rep |
Likely Rep |
8% |
Hold Rep |
Arizona |
Likely Rep |
Solid Rep |
5% |
Hold Rep |
Ohio |
Solid Rep |
Solid Rep |
3% |
Hold Rep |
Iowa |
Solid Rep |
Solid Rep |
2% |
Hold Rep |
Let’s look at how these races are trending.
A little hard to see, so let’s just look at the five closest races.
At one point, it almost looked like we might have to write off retiring Senator Harry Reid’s seat in Nevada (though pollsters have a notoriously difficult time with the Latinx vote there and it’s never a good idea to discount Democratic candidates in a close race). But now State Attorney General Catherine Cortez Masto (D) has made a huge surge and the data modelers’ consensus is that she has a better than 60% chance of holding the seat. That saves Democrats from needing to flip a fifth seat to take control.
The Likeliest Outcomes
Not all data modelers are entirely in agreement in how they weight the polls. Two of them have pretty different predictions in what they see as the most likely outcomes.
As of today, the New York Times sees a 50/50 split as the likeliest outcome with a 21% chance. They also see Hillary Clinton as the most likely winner of the White House, which would make Tim Kaine Vice President. As such, his tie-breaking vote would give Senate control to the Democrats. Their next highest probability is a 20% chance that the Republicans retain control with 51 Senators, followed by a 17% chance that the Democrats control with 51 Senators, a 14% chance that the Republicans control with 52 Senators, and an 11% chance that the Democrats control with 52 Senators.
In contrast, FiveThirtyEight is much more bullish on the Democrats’ chances. Their highest probability is that Democrats pick up six seats and control with 52 Senators, which they give a 15.3% chance. That’s followed by a 14.8% chance that the Democrats will have 53 seats, a 14.1% chance they’ll have 51, a 12.3% chance that it will be a 50/50 tie, an 11.5% chance that the Democrats will pick up eight seats and control with 54 Senators, and only then a 9.1% chance that the Republicans will control with 51 seats.
Given the disparity, I thought it would be interesting to show what the different data modelers give as the odds of the Democratic candidate winning the battleground Senate races. I’ve bolded the ones where the modeler predicts the Democrat will win.
Data Modelers' Predictions of the Democrat Winning in Battleground 2016 Senate Races
Race |
Avg |
NYT |
538 |
HP |
PW |
PEC |
DK |
Illinois |
91% |
94% |
91% |
81% |
89% |
93% |
95% |
Wisconsin |
87% |
85% |
95% |
98% |
86% |
73% |
84% |
Indiana |
73% |
68% |
73% |
50% |
73% |
90% |
83% |
Nevada |
61% |
64% |
61% |
50% |
60% |
80% |
53% |
Pennsylvania |
53% |
43% |
63% |
50% |
59% |
45% |
55% |
New Hampshire |
47% |
44% |
64% |
39% |
54% |
40% |
41% |
North Carolina |
38% |
31% |
51% |
32% |
49% |
27% |
38% |
Missouri |
31% |
24% |
62% |
6% |
42% |
17% |
33% |
Florida |
17% |
19% |
30% |
2% |
20% |
17% |
11% |
Louisiana |
8% |
8% |
17% |
1% |
14% |
6% |
0% |
Arizona |
5% |
7% |
9% |
1% |
12% |
1% |
0% |
Ohio |
3% |
1% |
5% |
1% |
7% |
1% |
0% |
Iowa |
2% |
2% |
3% |
1% |
6% |
1% |
1% |
Net Dem Pickups |
+5 |
+4 |
+8 |
+2 |
+6 |
+4 |
+5 |
The Huffington Post sees the Democrats picking up only two seats, allowing Republicans to retain control, though three other races are pure 50/50 tossups that could still give Democrats control.
The New York Times and Princeton Election Consortium currently see Democrats picking up a net gain of four seats, giving them a 50/50 tie with Vice President Tim Kaine providing the tie-breaker.
DailyKos sees a net gain of five seats, which would allow the Democratic majority to conduct business without needing Vice President Tim Kaine to constantly be casting tie-breaking votes right and left. And that, incidentally, is what the average consensus also has.
The PredictWise betting markets are predicting that the Democrats will get a net gain of six seats — they see Democratic pickups in Pennsylvania and New Hampshire that some of the others are less confident about. That would give them a total of 52 seats.
And FiveThirtyEight’s model currently sees Democrats picking up a net gain of eight seats, eking out a win in North Carolina — and surprisingly enough, they are even more confident about a win in Missouri. A total of 54 seats would provide a tiny cushion for the challenging number of seats Democrats will be defending in 2018.
Only 20 more days until the election!