How the electoral vote looks now

As I said a few days ago, I would turn my attention from national polls to the battleground states and the electoral vote. So I Googled “battleground states” and the first thing I found was an article at the Daily Kos presenting the findings of political statistician Nate Silver (who used to be with the New York Times, I don’t know if he still is). The article says that as of today no battleground states are leaning Romney. Silver says Obama currently has electoral vote advantage of 313 to 225. There are 538 electoral votes distributed among the 50 states and the District of Columbia. 270 electoral votes are needed to win the election.

At Real Clear Politics, the electoral vote situation is as follows. 221 are solid, likely, or leaning for Obama; 191 electoral votes are solid, likely, or leaning for Romney; and 126 are tossups.

RCP’s tossup states are:

Colorado (9)
Florida (29)
Iowa (6)
Michigan (16)
Nevada (6)
New Hampshire (4)
North Carolina (15)
Ohio (18)
Virginia (13)
Wisconsin (10)

As a sample, Florida is very close, all the polls show the candidates within two points of each other. In Colorado the RCP average has Obama 2.3 points up on Romney. Iowa is very close. In Michigan, the RCP average has Obama 2.4 up. And so on.

- end of initial entry -

Dave T. writes:

This is why I don’t think Romney is going to win in November. Ever since Romney secured the nomination in the spring, I’ve been loosely following the polling coming out of key battleground states such as Ohio and Florida (must-wins for Romney), and its been consistently close with a slight edge to Obama in the aggregate. Unless the Undecideds break for Romney by a wide margin in the final stretch of the campaign, which some conservatives are predicting, the presidential election is already over as far as I’m concerned.

Joe S. writes:

North Carolina is going for Romney, Virginia almost certainly will, and Wisconsin is very likely for Romney-Ryan. The polls are always two to four points too Democratic at this stage—the closer to the election, the “likelier” and more Republican the poll respondents get.

Whoever wins Ohio and Florida is a very heavy favorite. If Romney wins Florida and loses Ohio he still is a clear favorite; if he loses Florida and wins Ohio it’s 50-50.

John G. writes:

Real Clear Politics is just an aggregate of the major polls, including those that are not the least bit reliable. CNN/ORC and ABC, among others, have been shown as consistently oversampling Democrats. right now, the electorate is about evenly split between Democrats, Republicans, and Independents (35/35/30), but some polls have been of a sample of 45 percent Dems, 25 percent GOP, and 35 percent Independents. Others have shrunk the GOP share and increased Independents (still favorable to Obama). One poll a few months ago showed an 11 percent lead for Obama, but after reworking the Dem/GOP/Ind split to reflect reality, I got Romney up by a few.

It isn’t just a conservative prediction that undecided voters will break for Romney, its a historical trend that, at this point in the election year, most undecided voters will go against the incumbent (2/3 to 3/4). It’s not guaranteed, of course, but it is likely. A lot can happen in two months, so we’ll have to see. A better measure is the approval rating. Basically, the incumbent’s share of the vote will likely be nearly identical to his approval rating.

If you apply the above two (polling bias and undecided voters) to each swing state, Romney actually looks pretty good, and would likely win many of them if nothing changes. And while the national popular vote has swung Obama’s way the last few days, the swing state votes are unchanged.

Obama’s approval soared recently with Gallup. Interestingly, this not only coincides with the Democratic Convention, but comes immediately after they were threatened with labor law investigations. Even without that, while relatively reliable, they do show a slight leftist bias (inexplicably switching to all adults rather that registered voters to measure the reaction to Ryan’s VP selection, for example).

Unlike Gallup, Rasmussen did detect a bounce for Romney. They are also seeing one for Obama. Friday’s jobs report should bring Obama’s numbers back down. Even the MSM isn’t covering anymore, and seems to be admitting that unemployment is down due to people giving up rather than finding work.

All that being said, while I’m not ready to give up, I’m darn nervous. Particularly concerning is that the Dem Convention lit a fire with their voters, and the GOP has, for now at least, lost their enthusiasm advantage. The fact that the Dems can come right out and use Communist language, support abortion on demand, etc., and not send many voters running away is frightening. We truly are fighting for the survival of our way of life.

LA replies:

If by “we” you mean the Republicans, can you say that “we” are fighting for the survival of our way of life? No. The Republicans have not identified, much less opposed, much less fought against, what the Democrats stand for.

John G. replies:

Oh, no. “We” certainly does not mean Republicans. The GOP is simply the only option in the voting booth. “We” refers to traditional conservatives, those of us who understand the actual (rather than liberal/Orwellian) meaning of words like “freedom” and “equality.” Expanding on the last paragraph of my above post, its amazing how much our society is coming to resemble both Nineteen Eighty-Four and Brave New World—and how much the Democrats come right out and say it’s what they want.

And even with a Romney victory, we will have a lot of work to do.

I occasionally regret having children. If I were a single man I could make peace with the direction we are heading, even take guilty pleasure at how bad things will be for our adversaries should they be victorious in eliminating us; but not as a father. I truly fear for my children.

Alexis Zarkov writes:

Based on current polling, I calculate an Electoral College win probability of 20 percent for Romney, if the election were held today. He would get in the vicinity of 239 Electoral College votes, far short of the 270 needed to win. Here’s how I did it.

Using the Electoral Map data at Real Clear Politics, I assigned win probabilities on a state-by-state basis as follows: (a) 98 for “solid;” (b) 70 for “likely;” (c) 55 percent for “leans;” (d) 50 percent for “tossup.” All numbers are percentages. Thus Romney will win California’s 55 electoral votes with a two percent probability, Texas’ 38 electoral votes with 70 percent, and so fourth. I then calculate the win probability for all possible outcomes of the electoral vote which range from zero (he loses every state) to 538 (he wins every state). Then I sum all the Romney win probabilities for the range 270 to 538. In theory one must examine all possible state-by-state election outcomes, and that’s gigantic—two raised to the power of 51, which is two and a quarter quadrillion cases. Nate Silver and others use Monte Carlo simulation to sample from that gigantic universe. However, one can use the method of generating functions and avoid all that complexity, and do an exact calculation very easily. I don’t understand why he doesn’t do this. Sam Wang over at Princeton Election Consortium does exactly this, but he doesn’t explain it very well. Wang has put in a tremendous amount of work getting the state-by-state win probabilities, where I pretty much guessed at what “solid,” “likely,” “leans,” and “tossup” mean. He calculates that Romney will get 229 Electoral votes. Very close to my estimate. Nate Silver over at his New York times blog site, FiveThirtyEight, also gets a 20 percent win probability for Romney. Silver uses a very elaborate procedure, and has put in a lot of work. I spent about 30 minutes, most of that typing in the numbers. Writing the code to do my calculation took only a few minutes.

Finally I note there are some potential problems with this approach. Silver and Wang are both assuming that the win events in each state are all statistically independent of one another. I’m skeptical. If we found out the Romney won New York, that would certainly tell us something about the win probability for California before the polls closed. In any case, all this depends on how much one thinks polls reveal and predict what the voters will do in November. Color me skeptical.

Posted by Lawrence Auster at September 08, 2012 10:30 AM | Send