The “conventional wisdom” appears to be speculation. Maybe handicapping is a better word? IF this happens, and IF that happens, and IF those other things turn around, then it’s highly likely that (insert candidates name here) will win the upcoming election. IF any of those things do not occur, then we guessed wrong. Sorry folks. Oh well, there’s always another election around the corner.
And in fact, 538 just posted that the advantage is not there. Edit: moreover, that the advantage belongs to Trump.
The CW is usually based on what has been going on recently. Obama used the map to his advantage and if the race had been a near tie, he probably still would have won. But that’s because of the skill of his political team. That skill also enabled him to beat Clinton in delegates despite not winning that many more voters. He took advantage of the caucus states and just out organized her. Likewise, Obama’s team used great microtargetting to make sure key states would e event of a close race. Obama also had a key advantage Clinton won’t have: the black vote. The black vote is key in many swing states and increased turnout in places like Ohio and Florida paid dividends. Clinton is more likely to see the black vote turn out at historical levels so won’t get that extra boost in key swing states. Or maybe they will come out for her and carry her over the finish line.
Generally, the democrat is all but guaranteed 18 states in the northeast, west coast and northern midwest. Plus Hawaii & DC. That will give them about 242-246 EVs as a basement. You only need 270 to win.
Republicans, by contract have closer to 191 EVs as a basement (most of the south, the plains states, some of the southwest).
So basically of the 51 states and territories, the democrats start with 20 and the GOP starts with 23. There are about 8 swing states. The dems need about 25 EVs from those 8 swing states to get to 270, the republicans need 80 EVs from those 8 swing states to get to 270.
The 8 swing states are NV, CO, NM, IA, OH, VA, NC & FL. However CO and NM have been leaning more blue the last few cycles, so has VA. But they aren’t reliably blue the way Oregon, California or New York have been.
Hence the in built bias.
Why wouldn’t there be? The modern GOP has been declaring war on minority voters since 2010.
None of that matters if the GOP candidate wins by 5 points. Then even some guaranteed states start to waver.
If, yes. But when was the last time a candidate won by a 5%+ margin? 1996 I think.
Two elections ago. Obama in 2008.
Yup, 7%. I thought it was closer than that.
Having said that, looking at the map the map still matches the description of ‘safe’ states I was talking about. The Dems have about 20 safe states & territories, the GOP has about 23.
The only safe GOP state that McCain lost was Indiana, despite him losing the popular vote by 7%. McCain did lose all 8 swing states though (NV, CO, NM, IA, OH, FL, VA & NC).
If Trump does well in the popular vote some states you’re defining as safe won’t be, like Pennsylvania or Wisconsin. The electoral map does not always stay the same.
I agree, it changes. In 2004 states like Colorado, Virginia and North Carolina were considered safe. Bush won Indiana in 2004 by 20 points. Obama won all 4 states in 2008. California is reliably blue now, and has gone blue the last 6 elections. But it went red the 6 elections previous to that.
So yeah, I agree things can change. Every 20 years about 1/4 of voters are replaced, on top of that you have changing demographics and people moving around the nation. etc. But unless there is a major realignment politically, I think the states that have gone dem or GOP will remain that way.
It seems the realignment is mostly benefitting the dems. The dems are now winning in the southwest, they are also winning various southern states. It started with VA & NC becoming swing states. But there is speculation Georgia & Texas could become purple in the next 10 years.
But I do not see what changes will lead to the GOP winning more states. Maybe some midwest states like you describe, plus PA, may start to lean GOP because they feel left behind economically. It is hard to say. Those states are also seeing a larger % of the vote be minorities and millennials, which would also pull the states leftward. There aren’t enough working class whites leaning right who feel left behind in the modern economy to make up for all the millennials and non-whites leaning left.
All 538 is saying is that the electoral college won’t save anyone. Of course, if the popular vote changes, it will change the electoral vote. But that’s not the point. The question is based on the information we have now.
And the information we have now is that Clinton has an advantage in the electoral college. She has multiple paths to victory, while Trump has much fewer.
The fact that he has a bigger chance of winning the electoral vote while losing the popular vote is interesting, but it’s not the question. The question there is “how likely is it that we’ll be so close that this is a problem?”
Yeah, this isn’t independent of the popular vote. But no one said it was. It never can be, since the state polls are essentially a part of the national polls.
Examining the state-by-state chances at Rothschild’s site the race for the White House is seen to have converged to a very easy-to-grasp situation.
If Trump gets all the states in which his chance is 40% or more, he’ll have 265 electoral votes. After that there’s a big gap — his chance is 15% or less in all the other states — but*** if he wins even a single one of those states*** (to go with his 40+% states) he wins the election.
Trump has several 15%-chance states to choose from: Wisconsin, Pennsylvania, Colorado, New Hampshire, or (long-shot) Michigan. All five of these states are must-win states for Hillary.
I think sometimes the statement ‘electoral college math favors the Democrats’ is made when what the speaker really means is ‘the Democrats have done generally better than Republicans in recent presidential elections and we expect that to generally continue’.
It’s not necessarily that the details of the EC consistently favor the Democrats. They obviously did not in 2000. OTOH more recently Obama won the popular in 2012 by 3.9% points, but it you assume instead a 1% popular vote margin in favor of Romney due to uniformly 4.9%-points better performance for Romney in every state, Obama still wins 272-266 (Romney only lost FL, OH and VA by less than 5% points). That IMO is what you’d call ‘electoral college math favors party X’ as opposed to just saying ‘party X is stronger in national elections’.
This time the electoral/popular skew seems to have reversed. Harder to say through the margin of error of non-simultaneous polls, but it appears Trump is the one more likely to win while losing the popular vote. The RCP no toss up map is the same 272-266 margin for Democrats (different composition) but with the Republican behind 2.1% in RCP national avg (or 1.1% for several days up to till today). As pointed out, Trump has a ‘step up’ to make of apparently a few points to win any state beyond the 266 where he leads RCP (and that’s not very different from the more ‘sophisticated’ probability measures of some popular blogs). But if he’s really tied in ME, maybe he’s not really behind in NH, and so forth for how IA compares to MI/WI or NV to CO (as Silver pointed out in his piece today).
But Clinton has been the one ahead almost all the time, by more or less, so people are going to (imprecisely IMO) say ‘the electoral college math favors her’.
Even taking your cites at face value I don’t quite agree with your concluding last sentence.
Try this instead:
**IF **Trump wins every single one of his 40% states (states Clinton has a ~60% or less chance of winning) **THEN **Clinton needs to hold onto all those last 5.
But **IF **Trump drops some of his 40+% states (as predicted, since they’re only at 40% for him), then Clinton wins unless she also drops some of his 15% states having more total votes.
That second if says “If Trump does a bit worse than he hopes, Clinton has to do far worse than expected to lose.” It’s kind of hard to see how in a two-horse race both can do worse than expected.
Said another way, the one who does absolutely better will win. Expectations are fun, but actuals decide the race.
What you’re overlooking is that the race rises or falls with high correlation all over the nation.
All fifty states are watching the same cable news; the fifty states will watch the same debates. If there’s an “October surprise” it will surprise all fifty states.
The correlation isn’t perfect. Trump might gain votes in Michigan while losing votes in Florida. But the margin between Trump’s must-win states (Hillary chance ≤ 60%) and Hillary’s must-win states (Hillary chance ≥85%) is huge.
Indeed it was the starkness of this gap (60% is a lot less than 85%) and the fact that the gap occurs exactly at the 269-vote e.v. threshold that prompted me to post.
Setting aside the single votes up for grabs in Maine and Nebraska, what do you think the chances are that Trump will win one of the ≥85% states and simultaneously lose one of the ≤ 60% states? Do you want to place a wager?
Perhaps we’re talking past each other. Or one of us is confused, and I honestly don’t know who.
Agree completely that the 50 states are correlated though imperfectly so.
Right now Trump is expected to lose because of the many states that are only 40% or lower odds for Trump to win. It happens there’s a plateau (or cliff when viewed from the other side) of support for Trump at that level.
If we assume the 50-state public drifts more or less evenly towards favoring Trump as we get closer to the election, he’ll win more and more states. If his support drifts far enough, he might indeed will all those 40% states.
The there’s a band-gap. He currently only has a 15% chance at the next (last really) 5 states. His support needs to jump a bunch for him to win any of the last 5 and therefore the EV. Said another way, if Clinton’s support falls so far that Trump wins his 40% states, there’s a “support level” to use stock-picking terminology, where Clinton would have to fall a bunch more, an improbable distance more, to actually lose the EV.
As you say, the fact this “support level” is right at 269 EVs make for riveting TV.
Perhaps where you and I got separated is that I interpreted you to be saying that *if *things got so bad for Clinton that Trump won all his 40% states, *then *she’d have to perform a miracle to also hold onto all of Trump’s 15% states. That Trump getting all his 40% states almost assured he’d also get at least one of his 15% states. IOW, it sounded to me like you were saying that *if *Trump won his 40% states, *then *he’s all but assured of winning some 15% states too.
My take is that the band gap is real.
But that instead of the gap all-but ensuring a Trump win once he gets past the 40% threshold, I see the 15% states as probably blocking Trump from winning overall. In my view, him getting all the 40% states has some probability X. Him doing well enough to also get some 15% states above all the 40% states has probability Y.
Y is definitely less than X by definition. I think the band gap suggests that Y is much less than X. It sounded to me like you were saying Y is just slightly less than X, or maybe even equal to it.
Evidently.
And we did get separated right where you say: your interpretation that I jumped to a stupid conclusion! 
Reread my post and see if you agree that I never made such a jump. (Perhaps it might be inferred by someone trying to “read between the lines” but we all get bitten by that here.)
One thing I will say: When playing a bridge contract that makes easily against anything but a 5-0 trump split, it pays to give special attention to that unlikely possibility. Hillary’s chances of victory may be excellent (where’s the fingers-crossed icon?) but I hope her campaign gives special attention to those 5 or 6 states where her chance is “only” about 85%.
Your last line is what threw me. “All five of these states are must-win states for Hillary.” I interpreted your tone as implying “… and I (septimus) am really worried she’d be very unlikely to do that under the circumstances of her collapsing support. Yipes! :eek:”
It seems that was me misreading your tone. Sorry.
Overall I hope Clinton wins, and the stronger the better. But I’m far from comfortable about her odds today, nor am I comfortable about what Trumpist authoritarianism means for the future, with or without Trump at its van.
I tend to get hung up on the arithmetic details of the race for 270 evs — arithmetic is something I understand. 
I thought it interesting that 265 evs were within Trump’s grasp, but the next ev might be much harder to get. I posted just to show that “arithmetical curiosity” rather than to discuss actual politics. (The odds by-state fluctuate: this wasn’t the situation 2 weeks ago, and may change again before November 8.)
Actually we seem to be in very close agreement, and have been all along, once the “very unlikely” is deleted from your paraphrasing.
It’s flabbergasting how much support the short-fingered fraudster has achieved, despite that his whole candidacy was more of a publicity prank than a serious effort. One piece of good news is that his base — old white men — are a diminishing portion of the electorate. But they’re a powerful faction and, as you suggest, the havoc they wreak in coming years may be horrific even assuming Trump’s defeat.