Is it too early to say Romney has lost?

I think it is generally correct that there are more Republicans in the blue states than Democrats in the red states, but this distribution actually supports my claim and not yours.

The notion you express here is that more Republican votes are cancelled out by a winner-takes-all EC system than Democratic votes. But cancelled out votes are still counted by pollsters in the current framework. If I vote in Texas as a Democrat, I am still counted by the pollsters as a likely voter and so the current popular vote totals account for me. And we know that the current distribution of popular vote is basically split 50/50.

The variable we don’t know about is what will happen if Texas and California become competitive. If they move from 60% participation to 80%, what will their totals look like? People have examined this more comprehensively, but to summarize, what we know is that the stay-homes who would vote in a competitive election tend to be disproportionately of the party that is destined to win. Thus, in red states the people staying home are disproportionately Republicans and in blue states its the opposite. Thus, when every vote suddenly matters, you get more Democrats than Republicans.

Well, yes and no. Democrats have an advantage in that they get more “free” votes that they don’t really have to campaign for. But GOP voters definitely have an out-sized EC-per-vote share.

The reason I think the Tea Party wing of the GOP is likely to still support the EC even if Romney loses a split is because a non-EC world likely generates a more moderate GOP, one that can get large shares of Northeast and West Coast voters that they currently don’t really care much about.

Take a look at the map of solid red states - they tend to be smaller in population (except for Texas) and thus have more electoral votes per person than the bigger states. A pure popular vote system would reduce the influence of those small, low population states.

That’s quite easy to do when you either:

(1) Assume Democrats will, at least, double their turnout from 2008 or
(2) Assume Republican turnout will be depressed in 2012 more than it was in 2008

Neither of which are quite likely to happen.

While that’s fun, and allows various election analysis sites to shade Virginia light blue today or tomorrow, I *know *that ain’t right. I hope he wins Virginia, but I’d lay money Obama won’t take it by 4%.

Just to be clear…are you referring to the fact that the EV count is based on number of representatives (wholly proportional to population, except maybe in Wyoming, Vermont, and Alaska), PLUS number of senators? That is, it’s the addition 2 extra EV votes in each state which causes the skew of which you speak?

I vaguely recall that CBS NYTimes might be on the few pollsters that actually does weight by party ID, but I can’t seem to find that information in their release. Do you have it handy, or are you just assuming they do so?

I’m not him, but yes. That’s why smaller states get over-represented. A person in a 3 or 4 EV state has their vote count much more than a person in a 15+ EV state.

How much would you like to bet on that?

Got it, thanks.

I wonder if anyone has seriously proposed changing the EV system to eliminate the extra two votes per state. It seems like a relatively easy fix, at least compared to the politically/procedurally daunting task of eliminating the EV system altogether.

And you know Nate Silver is doing this because…?

Oh wait, he isn’t. From his latest post:

Hey, I’m just an anonymous blowhard with an internet account. You don’t actually expect me to put real money behind my opinions, do you?

:slight_smile:

Funded by FOX News is silly overstatement, but it would be an absolute utter shock if the conservative blogosphere didn’t erupt over this. People on the left still bring up Gore’s loss, and the left’s blogosphere is far more individualistic than the right’s.

What exactly are you objecting to with this statement?

FWIW, here’s the unweighted frequency from NYTimes/CBS/Quinnipiac for Ohio:

343 GOP
396 Dems
334 Indies
That’s a 53-response gap in the partisans. Assuming the independents split 55/45 Romney (which I think is generous), you still have Obama winning by about 2% without any weighing at all.

Says you. It’s also possible that demographic changes (as well as opinions of the parties) have led to more Democrats and fewer Republicans, relative to population, in the electorate, or some other explanation.

But the vast majority of the polling for any particular state is saying the same thing. It’s a lot more likely that those polls, as a whole, are pretty close to the truth, then that they’re all way off.

If the polls said Romney was ahead, I doubt you’d brush them off. But they don’t, so you do.

Conservatives, especially far-right Tea Party types, support the EC. They will even if it’s their ox that is gored (to the extent that Romney can even be called their ox). They know that, by and large, it is good for them that our system gives outsized weight to rural states, and extremely outsized weight to middle-right swing states.

The folks that would push for a national popular vote are more likely to be big-city conservatives (maybe a Will or Brooks), not Tea Party types. Hell, browse freerepublic for a day and you’ll find folks that want to do away with popular election of Senators.

Does “the right” lobby to get rid of the Senate when the Democrats are in control?

I’m sure there will be many who will challenge the legitimacy of Obama’s presidency if he loses the popular vote, or they’ll challenge the idea that Congress should cooperate with him (as if they don’t already do that), but they won’t try and get rid of the EC. That concept of federalism is firmly rooted in “the right’s” ideology.

And yes, I think the idea that FoxNews would get behind such a movement is absurd.

And yet the Tea Party is pretty much a creation of Fox News. Why would this be any more absurd?

No idea about weighting, because it’s not important to what I said, but crosstabs are here.

FL is D+7 at 37D/30R/29I
OH is D+8 at 37D/29R/30I
VA is D+8 at 35R/27R/35I

To compare to 2008:

FL was D+3 at 37D/34R/29I
OH was D+8 at 39D/31R/30I
VA was D+6 at 39D/33R/27I

Unless you want to argue that Obama is going to get, or exceed, his 2008 turnout, good luck defending that.

(And that’s without getting into the enthusiasm gap or the double digit swing in Independents.)

Almost anything is possible; that doesn’t mean it has a good chance of happening. Swing states are deemed as such for a reason. It’s highly unlikely they would buck national trends and see a sudden and rapid increase in Democratic affiliation whereas the exact opposite has occurred outside of the country. I’ve been over this before, but data tells us that Democratic affiliation across the U.S. is down whereas Republican affiliation has shown minimal change or, in some cases, large increases.

(Looking at Florida alone tells me there’s no way in hell there will be a D+7 split in the electorate this year.)

Unless, of course, they’re based on a faulty assumption.

Unluckily for you I’m pretty consistent, where I’ve said I expect something closer to a D+3 or so split nationally and state turnout which is somewhere between 2008 and 2010 turnout.

I don’t agree with that.

Because it goes against conservative principles.