On July 20th during the GOP Convention, I started to collect the New York Times’ Upshot’s overall and state-by-state forecasts for the 2016 presidential election. I figured that averaging what all of the forecasters were saying would help me clear out some of the noise and keep me from going bonkers from the individual polls coming in.
I wrote my first diary about my process on August 15, and have been posting a weekly Aggregating the Presidential Aggregators series ever since. Since Tuesday is Election Day, this is obviously my last regular Friday update. On Monday I’ll post a final summary combination of this and my weekly Aggregating the Senate Aggregators column, which had its last weekly post on Wednesday.
The NYT Upshot collects overall and state-by-state forecasts from five data modelers (the New York Times, FiveThirtyEight, DailyKos, the Princeton Election Consortium, and The Huffington Post), three knowledgeable experts (the Cook Political Report, Rothenberg & Gonzales Political Report, and Sabato’s Crystal Ball), and the PredictWise.
It’s been a rough week, with it becoming increasingly clear that some FBI agents are trying to pursue investigations against Clinton based solely on fabricated conspiracy theories from biased right-wing sources and leaking reports against their superiors’ commands to Fox News and Rudy Giuliani. Tuesday cannot come fast enough.
The Big Picture
Things have tightened over the last week, but the forecasters still give Clinton an average 87% chance of being elected President on Tuesday. This ranges from 68% from FiveThirtyEight, 84% from the New York Times and PredictWise, 91% from DailyKos, 98% from The Huffington Post, and 99% from the Princeton Election Consortium.
The Electoral College
Up until today, Clinton held a majority of the Electoral College as Solid or Likely Democratic, with odds of 85% or better or winning, plus states worth another 50 Electoral Votes with odds of 65% chance or better of winning. As of today, four Electoral Votes that had been “Likely” slipped into the “Lean” territory.
Let’s look at the last two weeks to see that:
You can barely see the light blue going below the 50% mark. For some perspective, let’s zoom out to see the trends going back to July (the opposite way we usually do this):
For most of the election cycle, Clinton needed to rely on a lot of Lean Democratic states (65% chance or better) to hit a majority of the Electoral College, and during some periods she needed tossups of under 65% chance to get there.
Let’s look at where the Electoral Votes have moved.
2016 ELECTORAL COLLEGE MOVEMENT THIS WEEK
270 NEEDED TO WIN |
10/29 |
10/30 |
10/31 |
11/1 |
11/2 |
11/3 |
11/4 |
SOLID DEMOCRATIC |
227 |
211 |
203 |
201 |
201 |
188 |
188 |
LIKELY DEMOCRATIC |
45 |
61 |
69 |
71 |
71 |
84 |
80 |
LEAN DEMOCRATIC |
50 |
50 |
50 |
50 |
50 |
50 |
54 |
TOSSUP — DEM |
1 |
1 |
1 |
1 |
1 |
0 |
0 |
PURE TOSSUP |
0 |
0 |
0 |
0 |
0 |
1 |
1 |
TOSSUP — REP |
36 |
36 |
36 |
36 |
24 |
0 |
0 |
LEAN REPUBLICAN |
22 |
22 |
22 |
22 |
28 |
36 |
52 |
LIKELY REPUBLICAN |
21 |
14 |
14 |
22 |
18 |
34 |
9 |
SOLID REPUBLICAN |
136 |
143 |
143 |
135 |
145 |
145 |
154 |
Let’s roll it up to see how that impacted the race.
2016 ELECTORAL COLLEGE PREDICTIONS
270 NEEDED TO WIN |
10/29 |
10/30 |
10/31 |
11/1 |
11/2 |
11/3 |
11/4 |
DEMOCRATIC |
322 |
322 |
322 |
322 |
322 |
322 |
322 |
TOSSUP |
37 |
37 |
37 |
37 |
25 |
1 |
1 |
REPUBLICAN |
179 |
179 |
179 |
179 |
191 |
215 |
215 |
It breaks my heart to say this, but for the first time all election cycle, Trump has broken 200 in forecasted Electoral Votes. There was a time when I was hoping he might not make 100.
Let’s look at the individual states.
States That Have Changed Since Last Week
Thirteen states and districts had movement over the course of the week: Alaska, Georgia, Indiana, Iowa, Maine (AL)*, Maine (CD 2)*, Michigan, Minnesota, Missouri, New Hampshire, Ohio, South Carolina, and Virginia.
STATES AND DISTRICTS WITH ELECTORAL COLLEGE FORECASTING CHANGES THIS WEEK
State |
10/29 |
10/30 |
10/31 |
11/1 |
11/2 |
11/3 |
11/4 |
AK (3) |
Likely R |
Likely R |
Likely R |
Likely R |
Likely R |
Likely R |
Likely R |
GA (16) |
Lean R |
Lean R |
Lean R |
Lean R |
Lean R |
Likely R |
Likely R |
IN (11) |
Likely R |
Likely R |
Solid R |
Solid R |
Solid R |
Solid R |
Solid R |
IA (6) |
TU — R |
TU — R |
TU — R |
TU — R |
TU — R |
Lean R |
Lean R |
ME-AL (2) |
Likely D |
Solid D |
Likely D |
Likely D |
Likely D |
Likely D |
Likely D |
ME-2 (1) |
TU — D |
TU — D |
TU — D |
TU — D |
TU — D |
TU |
TU |
MI (16) |
Likely D |
Likely D |
Likely D |
Likely D |
Likely D |
Likely D |
Likely D |
MN (10) |
Solid D |
Likely D |
Likely D |
Likely D |
Likely D |
Likely D |
Likely D |
MO (10) |
Solid R |
Solid R |
Likely R |
Solid R |
Solid R |
Solid R |
Solid R |
NH (4) |
Likely D |
Likely D |
Likely D |
Likely D |
Likely D |
Likely D |
Lean D |
OH (18) |
TU — R |
TU — R |
TU — R |
TU — R |
TU — R |
Lean R |
Lean R |
SC (9) |
Solid R |
Solid R |
Likely R |
Likely R |
Likely R |
Likely R |
Solid R |
VA (13) |
Solid D |
Solid D |
Solid D |
Solid D |
Solid D |
Likely D |
Likely D |
Ohio and Iowa continue to drift away, and may well be the only states that voted for Obama twice to turn red this cycle. Of the states above, Virginia and New Hampshire are part of Clinton’s blue firewall, states she’s counting on to ensure she gets to 270 even if other tossups like Florida don’t go her way. She’s still given a 95% chance of winning Virginia, which would still put it in Solid Democratic territory (Cook, Rothenberg, and Sabato knocked it down to “Likely Dem,” which brought down its score). And though there were some tight poll numbers out yesterday in New Hampshire, she’s still given an 82% chance of winning the state.
Tossup States
Here’s where the 13 battleground states have trended since the conventions.
That’s too hard to read, so I’ve broken it into three tiers: the swing states that are part of Clinton’s “blue firewall” that make up her minimum 270 Electoral Votes, the buffer states that could pad her majority and give her some leeway should there be a crack somewhere in her firewall, and the improbable bonus states that would only tilt her way in the event of a blue wave.
Tier 1
Who would have thought that Virginia would have been the most solid of the battleground states? Colorado gave us a scare last month when it briefly dipped below 70%, but it went right back up again. Pennsylvania has benefited from a lot of high profile visits recently and looks pretty steady. New Hampshire has taken a sudden dip but still remains about 80%.
Tier 2
These are “buffer” states that Clinton doesn’t need to win if she holds on to all of her Tier 1 states, but is nevertheless leading in. Winning them helps give her a buffer, gives her claims to a mandate (which the GOP wouldn’t acknowledge even if she broke 400 Electoral Votes), and gives her breathing room if something should happen with any of her Tier 1 states.
All have bounced up and down, dipping below 50% around the time of her 9/11 illness and taking a slight hit recently with the FBI leaks, but she still has a 66% chance of winning Nevada’s 6 Electoral Votes, a 70% chance of winning North Carolina’s 15 Electoral Votes, and a 73% chance of winning Florida’s 29% Electoral Votes. That’s a padding of 50 Electoral Votes with a 66% chance or better of winning.
Tier 3
Some of these states were at varying times within reach for Clinton, and others were always pipe dreams unless there was a real Democratic wave.
Ohio and Iowa may well be slipping away could well be the only two states Obama won twice to not go for Clinton (with Maine’s Second Congressional District also being a possibility). Arizona looked like a tantalizing prospect before the FBI went rogue (its label went awry, but it’s the bluish-purple bar in the middle there that suddenly spikes up before settling down again).
But it’s likely Clinton won’t win any of these, but who knows? Nevada has a notoriously hard-to-poll Latinx population that has helped Democrats beat expectations, and that could conceivably happen in Arizona or other states. Perhaps Republican women have been lying to pollsters (one poll found a huge disconnect between how men thought their wives would vote versus how women planned to vote and how they thought their husbands thought they would vote). Perhaps many will vote for Clinton, either because they can’t bear Trump’s misogyny or in solidarity with the first major party female candidate, or both.
Either way, none of these states are necessary for a Clinton victory and all — even Ohio and Iowa at this point — would be a shocking upset if she won them.
I’ll leave it at that for now. On Monday, I’ll post a final update of where things stand, state by state, as well as an update for the Senate races and a first ever-forecast for House races as well.