The Race is Tightening! (Well, sort of)

Ok, not really. But there are potentially – I say potentially – interesting developments afoot, which has delayed (along with teaching) my presentation of the final polling results before today’s election.  Now, I don’t want you to overreact to this, because in the end I don’t see how they change the final outcome of this race. Barack Obama should win.

But John McCain is closing in on, or pulling away from, Obama in every battleground state, with the exception of Pennsylvania and possibly Georgia  Now, let’s be clear here. I am not saying that he is going to win the Electoral College vote!  Indeed, even if he wins every one of these states, which today looks less preposterous than it did a week ago, he still will fall short in the Electoral College. Obama’s firewall remains Pennsylvania, with Virginia serving as a backup. I don’t see McCain winning either of those states.

Also keep in mind that these state-level survey results become essentially meaningless if Obama’s much-hyped GOVT proves as effective as some media outlets are suggesting. Keep in mind that Gallup estimates 64% turnout, up 4% from 2004, including a whopping increase in African-American voters.  If this plays out, it may render the state-level surveys irrelevant.

But without presuming anything about turnout, the results show that McCain has made across the board gains in almost EVERY BATTLEGROUND STATE.  I’m not sure why the media is not making a bigger deal of this, but the polling data is there to see. What interests me, as a political scientist, is how McCain can appear to be gaining ground at the state level, but not in national polls. Keep in mind that in some of these states the gains are within the polls’ margin of error, so it might simply be random noise.  But one would think that pure statistical noise would not manifest itself in every battleground state as gains for McCain – some of them should favor Obama.

As always, let’s look at the numbers.  Just to be sure, I’ll give the current polling average in both RCP and Pollster.com, since they use slightly different methodologies, with the person ahead in parentheses.  And I’ll show the trend in the RCP graph for the last week (Pollster doesn’t show a trend graph that is as easy to use).

RCP AVERAGE POLLSTER AVG.   McCain RCP GAIN in one week

North Carolina 48.4-48 (M)                 48.8-44.4 (O) McCain +2

Indiana             47.8-46.4 (M)              48.1-46.9 (M)  McCain +1.7

Florida             49.0-47.2 (M)              48.8-47.1 (O) McCain +4.5

Ohio                 48.8-46.3 (O)              49.4-46.3 (O)  McCain +2.8

Georgia            49.8-45.8  (M)             49.5-46.6 (M)  McCain NO RCP GRAPH

Missouri           48.5-47.8 (M)              48.5-47.4 (O)  McCain +1.3

Virginia 50.2-45.8 (O)              51-4-51.6 (O)  McCain +2.9

Pennsylvania     51-43.7 (O)                 51.6-44.4 (O) McCain +4.5

So it appears the late deciders are breaking for McCain – heavily in some cases. Lest I appear to be named Matt Drudge, let me reiterate: the bottom line is that Obama is ahead in most of these states, although his lead is well within the margin of error.  Moreover, most of McCain’s gains are within the margin of error as well. But the most recent polls in almost all these states show McCain narrowing the gap, or even taking the lead – that’s an interesting development worth keeping an eye on to see if it shows up in other states.

Now we need to keep this in perspective.  I see no evidence at the national level of a last minute surge toward McCain.  Can McCain lose the popular vote and still win the Electoral College by stealing these battleground states?  The math doesn’t work for him.  Without Pennsylvania, he falls short.  He needs to pick up another state or two, depending on the electoral size – and I don’t see any other states for him to win.

But he’s making it interesting.

By the way, many of you have asked if I will do an Electoral College projection.  I won’t, because it’s theoretically uninteresting.  Anyone can add up some polling data at this stage and put together a reasonable estimate for the final total – you don’t need my expertise at all to do this. It’s akin to standing atop Mt. Worth, watching the sky go grey over Lake Champlain, and predicting rain tomorrow in Ripton. You don’t need to be a meteorologist to do that – you just need to be observant. So it is with predicting the Electoral College vote the day before the election.  I would just be cluttering your inbox with extraneous information that you can easily put together yourself.  I’m sure Chuck Todd and the various talking heads have presented their own models.  No need for me to add to the clutter.

Remember, I made my prediction in September. I see no reason to change it.

Festivities begin in about three hours at the Grille. Meanwhile I’ll be parsing the polls a bit more, looking for any national trend, and taking a second glance at some western states: Nevada, Colorado and New Mexico that might make for a long night if McCain could steal two of them.

A final reminder: typically there are three waves of exit polls, with each successive wave calibrated to adjust for what the consortium is seeing in turnout. Drudge usually leaks the first one in the early afternoon – IT WON’T NECESSARILY BE ACCURATE!

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>