7 Days: The Breaking Point
I’m going to spend today doing nothing but taking the 12 electoral college projections I regularly track and pushing them to absolute results with no tossups. This is the 2012 Presidential election pushed to the breaking point.
The only state I don’t feel comfortable calling myself is Virginia, and here’s why:
Nate Silver is the best man in the game and has Barack Obama a 51.4% chance to win. At this point, it’s probably not polls driving that 1.4% so much as other variables like state unemployment, voter registration, and things like that. I won’t argue with it. But I also won’t argue with anyone that wants to hand the state to Mitt Romney, either. And since I have to give it to someone, I’m using Silver’s projections as my mental tie breaker. The rest I’m calling myself.
All projects will have Mitt Romney easily winning Indiana even though Obama won it in 2008. That was more of a fluke than anything. Indiana is not even a North Carolina that is still red, but destined to be a purple perennial tossup, with NC perhaps eventually ending up light blue eventually.
Nate Silver/538
Nate’s models have no tossups, so there’s nothing for me to do here except report his results. I’ve listed the generally accepted tossup states below just to illustrate who Nate has winning each of them. Obama should win 295.4 electoral votes to Romney’s 242.6.
Obama tossups: Nevada (6), Colorado (9), Iowa (6), Ohio (18), Virginia (13), New Hampshire (4).
Romney tossups: North Carolina (15), Florida (29).
Winner: Barack Obama, 295-243.
Talking Points Memo
TPM’s projection begins at 261-206, Obama leading, with 71 tossup votes from Colorado, Michigan, New Hampshire, Virginia, and Florida. Michigan is not a tossup. Romney hasn’t lead a poll there since August 16th, and the last ten polls look like this:
Obama +3, +3, +2, +7, +3, +10, +7, +6, Tie.
The tie is obviously an outlier. So Obama gets Michigan without any analysis necessary. Obama automatically gets Virginia as well, for the reason I explained above.
That leaves New Hampshire, Colorado, and Florida. Of the last five polls in each state, Obama leads in both New Hampshire and Colorado by 2.2 points, while Romney leads in Florida by 1.2 points. Obama takes NH and CO, and Romney takes FL. Obama should win the electoral college 303 to 235.
Obama tossups: Michigan, Virginia, New Hampshire, Colorado.
Romney tossups: Florida.
Winner: Barack Obama, 303-235.
Pollster
Pollster’s project begins at 277-206, with Barack Obama already having won enough states to win the election. There are four tossup states accounting for 55 electoral votes: Colorado, Florida, Virginia, and New Hampshire.
Most of these projections will end up being pushed the same until you get to maps with a ridiculous and unjustified number of tossups. Romney wins Florida, Obama wins Colorado and New Hampshire, and gets Virginia by tie breaker. That pushes Obama’s total to 303 and Romney’s to 235.
Obama tossups: New Hampshire, Colorado, Virginia.
Romney tossups: Florida.
Winner: Barack Obama, 303-235.
National Public Radio
NPR’s election scoreboard is new and has little historical data, and otherwise is a bit conservative, much like CNN’s map. Obama has 237 electoral votes and Romney has 206. They consider Florida, Ohio, Virginia, Wisconsin, Colorado, Iowa, Nevada, and New Hampshire to be tossups.
Like Michigan, Wisconsin is not a tossup. Here are the most recent ten polls:
Obama +4, +1, +3, +2, +6, +3, +2, +5, +6, Tie.
Wisconsin is obviously not tied, and not a tossup. Obama will win Wisconsin as easily and he will win Michigan. But there are a few other states to look at that the other projections haven’t touched yet. Here are the last ten polls for each of Ohio, Iowa, and Nevada:
Ohio: Obama +1, Tie, Tie, Obama +3, +2, +5, Tie, Obama +2, +4, +2.
Romney hasn’t lead a poll in Ohio in over a month and Obama’s average lead there this late in the race is looming large, and it’s a bigger lead than he had over the last ten polls (+1.9). That means Obama’s lead is growing in Ohio, and is already of important size. If Romney had a shot at Ohio this year, it was after the Denver debate. But whatever Romney had, he lost. So it goes to Obama, the man who has lead or tied 14 straight polls and has an average lead of 2.6 points in the last five polls in the state.
Nevada: Obama +5, +3, +3, +7, +8, +2, +2, +4, +3, +1.
Mitt Romney’s last lead in Nevada was in mid-April, and he’s not going to win the state this year.
Iowa: Obama +2, +4, Tie, Obama +3, +8, Romney +1, Obama +1, Tie, Obama +2, +4.
Iowa is tight, but the averages tell the story. Obama’s average lead in the last ten polls is 2.3 points, and in the last five polls it’s 1.2 points. Iowa is one of the few states in my battleground tracker where Mitt Romney is gaining ground (the others are North Carolina and Florida, where he’s already leading.) But he’s still behind. The state is close, but it’s not tied. And the trend is Romney flattening out and Obama moving up, just every so slightly.
New Hampshire: Tie, Romney +4, Tie, Obama +1, Romney +1, Obama +9, Obama +3, Romney +2, +2, Obama +3
Pretty close, and that +9 for Obama is probably an outlier. The last ten and last five averages are Obama +2.2 and Obama +0.7, respectively. To squash that outlier, the last 10/5 medians are a tie and Romney +2. The reason I don’t see this as a tie or a Romney lead is because New Hampshire was showing huge leads for Obama before the Denver debate. Between April 9th and September 30th, Obama had lead or tied in 18 polls, to Romney’s 3. Obama held leads as high as 15 points.
I just don’t buy that a state churning out results like that has magically become tied. Not when so many of Romney’s good results come from right-leaning pollsters like Rasmussen and ARG.
Now add them up:
Obama tossups: Ohio, Nevada, Iowa, Virginia, Wisconsin, New Hampshire.
Romney tossups: Florida.
Winner: Barack Obama, 303-235.
CNN
CNN has one of the most conservative (not in a political sense) maps of anyone, probably second only to Real Clear Politics. They’ve called 237 votes for Obama and 206 for Romney as safe, and they haven’t budged from that all year unless I missed a switch on North Carolina. Their tossups are Nevada, Colorado, Iowa, Wisconsin, Ohio, Virginia, Florida, and New Hampshire.
Obama automatically gets Virginia (tie breaker) and Wisconsin, because Wisconsin is not a tossup. Per previous analysis of recent state polling, Obama also gets Nevada, Colorado, Iowa, Ohio, and New Hampshire. Romney gets Florida, the only toss state he has a lead in. (CNN has Romney already winning North Carolina, as it should.)
Obama tossup: Nevada, Colorado, Iowa, Wisconsin, Ohio, Virginia, New Hampshire.
Romney tossup: Florida.
Winner: Barack Obama, 303-235.
Time Magazine
Time’s projection appears based on analysis done by the Associated Press, and has enough states allocated to Barack Obama to win the election without any changes, 271-191. Their tossup states are the familiar Colorado, Nevada, New Hampshire, Virginia, and Florida. But not Iowa or Ohio. And they consider North Carolina to be a tossup.
Having a new state to look at, here are the most recent ten polls out of North Carolina: Obama +1, Romney +1, +4, +2, +6, Obama +3, Romney +1, +8, Tie, Romney +6.
Not a tossup. Again, based on previous analysis of the other states, we get..
Obama tossups: Colorado, Nevada, New Hampshire, Virginia.
Romney tossups: Florida, North Carolina.
Winner: Barack Obama, 303-235.
Real Clear Politics (T)
Conservatives have been clinging to RCP’s map (with tossups) about as much as they’ve been pretending that Gallup is the only credible pollster in the world. But RCP has another map, one without tossups, that isn’t so favorable.
Let’s call the map with tossups RCP-T, and without, RCP-NT.
RCP-T has reverted to its previous projection of 201-191, Obama high, with a laughable 146 tossups, nearly as many electoral votes as Mitt Romney has. There’s the standard crop of Colorado, Nevada, Iowa, Ohio, Virginia, Florida, and New Hampshire. There’s the questionable pick of North Carolina, the silly picks of Wisconsin and Michigan, and the downright dumb pick of Pennsylvania. No other projection that I know if considers Pennsylvania a tossup, and here’s why:
PA: Romney +4, Obama +4, +4, +7, +4, +10, +5, +3, +5, +6.
The one Romney lead came from Susquehanna, a GOP pollster on behalf of George Allen’s Senate campaign. In other words, it’s the equivalent of an internal campaign poll which is usually worthless. A previous poll by Susquehanna that wasn’t paid for by George Allen’s campaign showed Obama up by 2 points.
Pennsylvania is not a tossup.
Obama tossups: Pennsylvania, Colorado, Nevada, Iowa, Ohio, Virginia, New Hampshire.
Romney tossups: Florida, North Carolina.
Winner: Barack Obama, 303-235.
Real Clear Politics (NT)
And the no tossup results from RCP which can’t be pushed were Obama 290, Romney 248. The only quibble is Virginia, which I won’t contest because it’s close enough to go either way. But this just illustrates how little Virginia matters in this election. Even with Virginia, Obama still wins easily, and there are no other states as close as Virginia except perhaps New Hampshire and it’s 4 electoral votes. The next closest after that is Iowa, and its 6 votes. Give Romney Virginia, New Hampshire, and Iowa, and he still loses 280-258.
electionprojection.com
This site can’t be pushed either because it has no tossups. It has Obama winning 290-248, while losing Virginia. That is the only state that removes this site from the 303-235 consensus.
electoral-vote.com
This site also has no tossups with Obama winning 332-206. It gives Virginia to Obama, but also gives him Florida. The only explanation I have for that is that this map was last updated in September when Obama was leading in Florida on the back of his convention domination. But that isn’t the case now and I don’t think that allocation is very defensible. Like electionprojection.com, only one state removes EV.com from the 303-235 consensus.
Josh Putnam
Putnam is an assistant professor of political science at Davidson College and maintains his own regularly updated projection, which is identical to that of EV.com. Obama wins the electoral college 332-206 because of winning Florida, which isn’t likely, despite this projection being updated on the 24th.
Princeton Election Consortium
PEC’s map only considers Virginia to be a tossup, yet gives it to Obama for a 291-247 win. I’m not sure how they got to that result because of a very poor site design and an interactive map that won’t load. But it’s roughly in line with other projections.
270towin.com
Another projection with too many unjustified tossups. A race is not a tossup simply because it’s close. It’s a tossup if it’s too close to call, and only two states — New Hampshire and Virginia — really make sense. This map begins with Obama leading 217-191, with Nevada, Colorado, Iowa, Wisconsin, Ohio, Pennsylvania, Virginia, North Carolina, New Hampshire, and Florida as tossups.
As previously discussed, Wisconsin and Pennsylvania are not tossups and so go to Obama. Neither is Florida and North Carolina, which go to Romney.
Obama tossups: Nevada, Colorado, Iowa, Wisconsin, Ohio, Pennsylvania, Virginia.
Romney tossups: Florida, North Carolina.
Winner: Barack Obama, 303-235.
So here’s the summary with no tossups:
As I’ve been saying all along, this race isn’t as close as the mainstream media is pretending it is (or they simply don’t understand) or as national polls show it to be.
**Note: this story was written on Saturday, October 27th, and has not been updated with weekend battleground polling