Friday, October 26, 2012

Part II: The Electoral College

Now we come to the interactive portion of our analysis. Thanks to the folks
at 270-to-win, we begin our assessment of who wins the Electoral College
and the Election.

Here's the link we'll be using: http://www.270towin.com/
Be sure to hold down your control button before clicking.
If it doesn't work for you, typing, '270,' into your browser
should.

The most defining elections in recent memory, I believe, are the Presidential
election in 2008 and the mid-term elections in 2010 which saw the biggest shift
of power in one election since, "Dewey beat Truman," in 1948. The outcome of
2012 will almost certainly be a mash-up of the two telling us where the balance
of power lies between them.

The map you see colored here reflects what the polls say is the starting point for
each candidate with Obama almost certain to win the Blue states, Romney the Red
states, and the rest being within the realm of possibility for either. It also puts both
candidates at roughly the same starting point, Obama at 201 vs. Romney at 191.

While I do believe the science of polling has improved in the last 20 years, knowing
what I know of past election results in that time we need to make some changes:

Michigan (MI), Pennsylvania (PA), and Wisconsin (WI) should be moved to Blue.

Both Indiana (IN) and New Mexico (NM) should, for the movement be moved to
undecided (Tan).

This is my starting place.  Every state in Blue you see here voted for the Democrat in
EVERY election of the last 20 years going back at least to Bill Clinton's first election
in 1992.  That means none of them, not one, voted for George W. Bush in either 2000
or 2004.  He came close in a few, but never prevailed, and though it isn't impossible
that Romney will win one of them (Michigan, Pennsylvania, or Wisconsin) he only
ever led in Wisconsin this summer shortly after he selected Paul Ryan as his running
mate. Polls currently show the President not just leading but with a majority of the vote
in each of them. What's more, the President won them in 2008 by margins far higher
than even Bill Clinton ever won them let alone Al Gore or John Kerry. Sometimes by
twice as much.

Again holding down the your control key on your keyboard, you can check the
percentage results for 2008 here: http://uselectionatlas.org/RESULTS/

Just move the arrow over each state with your mouse to see.

The Red states are those Senator John McCain carried in 2008. My assessment is that
with the possible exception of Arizona, none of these states is competitive for the
President.  Given the politics of electing the first black President in our history, voting
against that choice required of the voter a fairly conservative point of view. If self-
identified moderates voted for McCain they certainly didn't do so AS moderates.  And
not all of these have a consistent history of voting Republican. SIX of them (Arkansas,
Tennessee, Louisiana, Kentucky,West Virginia, and Missouri) voted twice for Bill
Clinton and with the exception of Kentucky and Missouri they actually gave McCain
a larger share of the vote than they gave to George Bush. And, it means they actually
gave fewer votes to Senator Obama who won, than Senator Kerry, who lost.

Obviously, this map would seem to give the President a huge head start, fewer than 30
electoral votes from 270, while Romney needs 90 more to win. Consistent with the 48%
or so every Democrat won from Clinton to Gore to Kerry these 18 Democratic states
comprise about 90% of what is necessary to elect a President and they would seem to
start each election with probably that much.

In my last post I said that the President was favored to win the popular vote in 2012, but
that no President had ever been re-elected when the economy was in recession as it is now.
It is these facts that cause me to believe that if re-elected, Obama would likely be the first
to do so with a smaller percentage of the vote than when he first took office.

It also brings us to our next changes: North Carolina (NC) and Indiana (IN) back to Red.

One of the most interesting and frankly, helpful dynamics in American politics is that among
those states which are the most highly contested between the parties a consistent ranking can
be made between them. That is, the states can be ranked from most Republican to most
Democratic and with the exception of the most partisan, the states line up in about the same
order in every election. So as the national popular vote moves from one party to the other it's
not difficult to predict which states will flip to the other party first.

Both Ohio and Virgina gave the President a margin of victory matching his national popular
vote.  A fact more than coincidental as we'll see later. But Florida, Indiana, and North Carolina
gave the President smaller margins. In fact, four states denied a majority to either candidate.
Indiana and North Carolina went to the President. Missouri and Montana to Sen. McCain.

If the President were running for re-election in a good economy, it is likely his national popular
vote would about what it was in 2008; probably even a little bigger, in which case it would be these 4 states and maybe Florida being fought over, none of which the President needs to be re-elected. He'd be a shoo-in for re-election.  Instead, as the President's national support has narrowed, it's these states that fall first into Romney's camp.  Polls in all 4 show Romney with a majority of the vote.

And so we move Indiana (IN) back to Red and North Carolina (NC) as well.

Almost immediately, however Romney must confront a basic reality: Moving Florida (FL) to
Blue puts Obama at 271, which means Obama would win and Romney can't afford to lose it.

If Obama wins Florida it will not just be due to the support of its older population, but almost certainly because of the substantial Cuban population which is trending Democratic especially
among the younger generation. The President's support among Hispanics generally in 2008 was
2-to-1 over McCain. He shows the same support among them today. As we consider the Southwest where the substantial Mexican vote is even more Democratic than the Cubans in Florida, this fact must be taken into consideration.  Polls have shown the President consistently ahead in both though Romney almost caught up to him in Nevada right after the conventions. Given Romney's decision not to go with an Hispanic Governor or Senator as his candidate for President I suspect Romney was prepared to concede both to the President early despite the substantial Mormon vote that's also in Nevada.

Thus, we now move both Nevada (NV) and New Mexico (NM) to Blue.

Polls in Florida, though, have been pretty good for Romney. In fact, before the 1st debate in Denver, when it looked as if it might all be over for him, he was still competitive in Florida, Colorado, and New Hampshire as the rest looked like they'd decided on Obama before a single debate. After the first debate Romney overtook the President in Florida and now consistently surpasses the President
there.

So we move Florida (FL) Red. The President is still closer to winning but Romney is closing
only 35 electoral votes away from victory and the President at 17 away from winning.

Shifting Ohio (OH) to Blue again shows just how little margin for error Romney really has, as
Obama again moves to 271.  Even if Romney won all the remaining undecided states, he'd still lose. Ohio is dead center politically and worse for Romney it's unlikely he could win another midwestern state (Wisconsin, Michigan, or Pennsylvania) without Ohio as it has voted consistently to the right of these states for well over half-a-century.  Stipulating that this state could go either way, let's move Ohio (OH) Red for the moment.

The race is now tied at 253 a-piece and we begin to see the real balance of power emerge. Before 2008 states like Virginia and Indiana voted consistently Republican. Arizona and Colorado also leaned that way, and so it became fashionable after 1996 to say that of Ohio, Florida, and Pennsylvania, whoever won 2 out 3 would win the election.

As we can see our map now, that is no longer enough for Romney HOWEVER moving Virginia (VA) either Blue or Red shows again why Ohio and Virginia matched the nation's popular vote.
Assuming Romney can win Ohio, Virginia becomes almost certainly decisive. At this point the President has to either win Ohio or carry ALL of the remaining 3 (Colorado, Iowa, and New Hampshire) in order to win without Virginia. Romney would have to do both. So the same edge
given by the, "keys to the Presidency," would seem to be born out in the Electoral College.

1 comment:

  1. Good, if complicated, analysis! I think you should send this to the Weekly Standard, or to Karl Rove and tell them you'd like to help them do analysis! Maybe they'd bite?? I'm sure you could find their emails, and wouldn't it be fun if they responded! NV, NG!!

    ReplyDelete