One More Day
Tomorrow night (or more likely Wednesday morning), someone is going to look pretty bad. It might be the pollsters who have continually insisted on using a 2008 model for their polls. Or it will be pollsters like us and other analysts who have criticized the 2008 model as unrealistic and exaggerating Obama’s advantage.
But, finally, we’ll get some real answers.
So, to put ourselves out there even more, today we engage in the highly foolish act of prognostication. It would probably be better if we didn’t, but many have been unwise enough to ask me what I think.
So, with great thanks (as always) to WPA partner Bryon Allen who does all the hard work on these forecasts, here is my final WPA election update for 2012.
The Presidential Race
Way ahead or slightly behind
On Tuesday, we’re going to find out that one of two things is true:
- Obama was able to deliver a turnout that looks a lot like 2008 and the polls using a 2008 model or something close to it were right all along, or
- Three consecutive wave elections (2006, 2008, 2010) has caused pollsters to forget how to model a “normal” electorate.
- Turnout rates will look a lot like 2004 which, even when adjusted for demographic changes, still yield an electorate that looks a lot more like 2004 than 2008.
Without the ability to actually see the future, we’ll just have to wait for tomorrow to see which of those outcomes occurs.
But here’s what they mean:
- If the “a lot like 2008” model is right, Obama will be re-elected narrowly with something between 270 and 280 electoral votes (and the media polls will have been right all along).
- If “normal” is still something much more like 2004 group participation rates and partisanship, then Mitt Romney will win with more than 300 electoral votes (and media pollsters will have a lot of ‘splaining to do).
You can guess which one I think will happen. The good news is that none of us now has to wait long to know the answer.
Reading tea leaves
Now, for the very impatient, here are some hints in the latest early voting data distributed today by the AP:
- In Colorado, Republicans lead by two points in early vote turnout.
- In 2008, Democrats lead by two points in early vote turnout.
- That’s a four point swing toward Republicans; but not enough to erase Obama’s 2008 advantage without some big shifts among Independents.
- And, recent polling has shown Obama leading Independents by just two points when he won them by ten in 2008, according to the exit polls.
- So, based on this, Colorado goes Romney.
- In Iowa, Republicans have an 11 point gap in early votes so far, which compares to an 18 point gap in 2008.
- Obama won Iowa in ’08 by more than nine points, so there will have to be an even bigger shift in Election Day votes to change the outcome.
- So, based on this, Iowa goes Obama.
- Ohiois notoriously hard to judge for early voting because the only “party registration” is based on the last primary in which a voter cast a ballot.
- But, based just on that measure and absentee/early ballot requests, Republicans have shaved a 14 point 2008 gap down to a six point gap (an eight point gain) in a state Obama won by less than five points in ‘08.
- So, based on this, Ohio goes Romney.
- In Nevada, Republicans trail by seven points in early voting.
- In 2008, they trailed in the two major counties (which made up 88% of the early votes cast) by 19 points.
- Even accounting for the fact that the rural counties are more Republican, that’s a significant closing in a race Obama won by slightly under 13 points in 2008 and makes Nevada a very close race even assuming Independents don’t shift.
- But Romney leads among Independents by seven points in the latest Las Vegas Review Journal poll (Obama won Independents by 13 according to the 2008 exit polls).
- So, based on this, Nevada goes Romney.
In our analysis (estimation?) that yields three states (Colorado, Ohio and Nevada) that are performing quite differently than the 2008 model and are, based on the current evidence, likely to go for Romney.
Adding North Carolina (not really a swing state this year) to that mix gives Romney 268 Electoral Votes.
This is before discussing states without the same information currently available—states like Virginia (Obama at 48% in RCP average), New Hampshire (Obama at 49% in RCP average), and perhaps even a Pennsylvania (Obama at 50% in RCP average), Wisconsin (Obama at 50% in RCP average) or even Oregon (Obama at 50% in RCP average). All those RCP (Real Clear Politics) averages are based on polls relying on 2008 models.
So, to summarize, if Romney wins any of those states, he becomes the 45th President of the United States.
Status quo is almost certain.
Unless something shocking happens, Democrats will retain control of the Senate and Republicans will retain control of the House.
In the Senate, it really comes down to the candidates and a shift in the map. While there is still a path to 50 available to Republicans, it requires essentially a sweep of the close races. The most likely outcome at this point seems to be Republicans picking up a single net seat, losing a single net seat, or holding steady (by winning two to replace Scott Brown and Olympia Snowe).
In the House, the current RCP generic ballot average has Republicans with a slight advantage. The equilibrium point in House elections tends to be Republicans down slightly on the generic ballot. While redistricting will blunt the opportunity for pick-ups, with this generic ballot reading we expect Republicans to add five to ten seats to their House majority.
Hope you’ve enjoyed these over the past few weeks—this will be my last one for the cycle. We will be preparing our regular post-election “What Happened??” presentation, to try and answer that question. If you’d be interested in having someone from WPA deliver that to a group please contact WPA VP of Marketing, Rob Spicer (firstname.lastname@example.org). Also, please go to www.wparesearch.com/blog and subscribe to our blog by email (upper right corner) if you’d like to read more of our thinking throughout the year. We try to post everything to RedState, but don’t always remember.