College-level Math 1, Electoral edition
Original post made by Tom Cushing on Sep 26, 2012
The good news for the Romney ticket is that they have amassed an insurmountable lead in square-mileage more than 2 million-to-1 million, with less than 400,000 really in-play. Their bad news is that, despite concerted state legislative attempts at redefinition, the term "voter" neither extends to the vegetable or mineral contents of that area, nor is it defined anymore as "white males who own property" (cf. 1776). No, voters must be living, breathing, non-felonious citizens; corporations do not (yet) qualify, either.
The electoral college was established in the US Constitution to represent a mixed system of electing the Prez and the VP each state is allocated a number of electors equal to its senators (2/state, regardless of population , or size) and its Congressfolk (apportioned by population). Thus, California has 55 electors, whereas seven small-or-empty states and DC have only three each. The electors are pledged to cast their ballots for the winners of their states' popular vote, except for Maine's four and Nebraska's five who are not so tethered. 270 votes half of 538, plus one is the magic threshold of victory.
This system means that states whose voters are reliably GOP or Dem can be counted as 'safe' by the campaigns, and largely ignored except as sources of funds. Thus, you rarely see the candidates in our fair environs, or Texas or NY, except with their hands-out. Conversely, the so-called battleground states, whose populations are more fickle, receive disproportionate attention (now, and probably also in governance anticipating the next election cycle).
With that in-mind, it's also possible for us to simplify our calculations: reliably GOP states, largely in the south and mid-continent total 200 votes, whereas safe Dem states (both coasts, upper Mideast) total 237. Since neither Party controls a majority, those 8 swing states FL, OH, NC, VA, WI, IO, NV and NH are in-play. BTW, the info that follows is largely drawn from 538.com, a website that attempts to aggregate and weight the outputs of various individual polling organizations, according to their methods and historical reliability. It is also licensed to the NY Times, which may be enough to send some readers scurrying to the temporary comfort of Rasmussen. So be it 538 founder Nate Silver's statistician chops are well-established, and I'm seeking accuracy.
Currently, those battleground states shake-out as follows:
Florida (29 votes): currently 48/46 Obama, projection 50.4/49.0, probability* 79%
Ohio (18): currently 48/44 Obama, projection 51.2/47.5, probability 63%
North Carolina (15): currently 47/46 Romney, projection 50.6/48.7, probability 67%
Virginia (13): currently 48/45 Obama, projection 50.9/48.3, probability 73%
Wisconsin (10): currently 50/44 Obama, projection 52.4/46.8, probability 87%
Iowa (6): currently 48/44 Obama, projection 51.1/47/5, probability 74%
Nevada (6): currently 49/45 Obama, projection 51.6/47.1, probability 82%
New Hampshire (4): currently 47/44 Obama, projection 51.7/47.2, probability 77%
* Probability is a measure of confidence in the winner projection
Silver concludes that the likeliest electoral college split 312/226 Obama, at a win probability of 79.7%, with the popular vote also going his way, 51.2/47.7%. Since the respective conventions, the trends have also favored Mr. Obama, with his chance of winning having risen 7% in just the last week. Obviously, there have been repeated blunders by the campaign of the candidate many supporters are counting-on to be a competent manager of his business.
All that said, it is striking just how close the numbers are in those critical swing states, and much action particularly the debates lies ahead. Florida and Ohio are crucial Mr. Romney simply cannot win without carrying at least one of them. Indeed, if Ohio and WI go blue, then Mr. Romney must take all the rest, in which case we end up at 265 Obama to 273 Romney. There are even ways that those NE and ME electors become crucial, and a 269/269 is statistically possible, although at a 0.3% probability.
We'll take another look at these numbers much closer to November 6, as well as looking at the likely future balances of power in the House and the Senate. Once again, stay tuned!
on Sep 27, 2012 at 9:55 am
It's certainly within ethical standards to shop for the best polling data for your candidate, but if you're "seeking accuracy" then a source closely tied to the NYT is not your best choice.
on Sep 27, 2012 at 12:23 pm
I suggest everyone read Dick Morris' analysis of the poll numbers. Quite revealing. The MSM and this columnist are in the tank for Obama who thinks the death of Americans is a "bump in the road." Romney bumper stickers guarantee your car will be vandalized--thus so few seen on the roads. The sleeping giant is alive and well.
on Sep 27, 2012 at 2:03 pm
Darn, I really do want a Romney/Ryan bumper sticker on my car, but I have to admit that I worry about who will ultimately trash our car or even hurt our children when they're driving it.