Articles and Analysis


One Day to Go and McCain Is Between Barack and a Hard Place

Tomorrow, Barack Obama will become the first Democratic Presidential candidate since Jimmy Carter in 1976 to win an outright majority of the votes cast on Election Day -- and with it a sizeable majority of electoral votes -- making him the next President of the United States.

We make this projection knowing that the gap is closing both nationally and in key states; it is our sense, however, that this trend would have to continue for another 10 days for the election to swing back to McCain.

The following is our rationale for going with Obama:

  • The economic recession/financial meltdown dominated the headlines from mid-September to mid-October. The war in Iraq remains enormously unpopular. Bush's approval ratings are near an all-time low for modern Presidents. And the GOP brand is weak and fractured. As a result of these factors, a majority of this hugely dissatisfied electorate will be voting Democratic to change the direction of the last eight years.
  • October was the worst month for the stock market in 21 years. Yes, last week was an improvement, but the month of October was unkind to John McCain and the GOP. Last Thursday, the government reported that the economy contracted from July through September - the first time consumer spending had decreased in 17 years.
  • With this environment as a backdrop, Obama will pick the GOP lock on the electoral college by winning six states George W. Bush won in 2004--Nevada, New Mexico, Colorado, Iowa, Ohio and Virginia--en route to an electoral vote rout.
  • This election was always about Obama and McCain was never able to paint him as either "unfit" or "unprepared." Nor was McCain able to give people a clear reason to vote for him.
  • In an ironic twist, it was Obama who defined McCain in a negative light rather than the other way around. They started by claiming that he was "confused" four months ago and then painting him as "erratic" in the last 60 days. Of course, team McCain and the candidate himself contributed to this. It will be interesting to count the gross rating points that went behind contrast ads on both sides. My guess is that the Obama campaign might win that count as well.
  • Terrorism and national security virtually disappeared as election issues. These two issues dominated a large part of the national dialogue in 2004 and helped give Bush his re-election victory.
  • New registrants, young voters and black voters are going to break with historical pattern and vote in disproportionately high numbers, giving Obama huge margins in certain states and propelling him to victory over an exhausted and disengaged GOP base.
  • The Democratic ground game will prove to be vastly superior to the Republican operation (money can do that).
  • The turnout will be between 58%-60%, which would be its highest level since 1960. If the total number of voters exceeds 130 million (meaning more than 61% of eligible voters will have voted) then the Obama win could be an electoral landslide because the Democrats have a built-in six-eight point advantage in terms of party identification.

The LCG regression vote model projects that Obama will win by six percentage points tomorrow. We project the following popular vote distribution:

Obama 52.5%
McCain 46.5%
Other 1.0%

Below is our regression projection line. Today's analysis produced the usual curve, which shows McCain losing by 8.6 points. However, if you look at only the last 40 days--which roughly corresponds to the first week that voters digested the impact of the financial crisis (the week of September 25th)--you begin to see more clearly the McCain descent and recent uptick. When we built a separate model for that period, it produced the green line, showing McCain losing by just 6.5 points. McCain pollster Bill McInturff is correct: there has been some movement in the last 10 days. However, it is too little and way too late.

regression nov 3.png

We project that Obama will decisively win the electoral vote:

Obama 311 EVs
McCain 227 EVs

He will accomplish the above by winning the previously-mentioned Bush 2004 states as well as Pennsylvania. The following is our last updated EV projection map and some commentary on specific states:

map nov 3.png

  1. Obama will carry three western Bush states - Nevada, Colorado and New Mexico. However, it is important to note that all three were very close in 2004. Bush only won Nevada by two points and he won New Mexico by just a single point (6,000 votes). The Latino population in Nevada will tilt toward Obama and that will deliver the state. New Mexico nearly went Democrat in 2004 and it will do so this time around due to huge Obama margins in Santa Fe. The demographic shifts in Colorado made it vulnerable for the GOP even without the ideological symmetry with Obama. Colorado has been in the Obama column for a month.

    CO nov 3.png

    NV nov 3.png

  2. McCain lost Iowa the moment he secured the GOP nomination because of his past opposition to ethanol subsidies. Bush only won the state by 10,000 votes in 2004 so it was a toss-up to begin with.

  3. Obama will win Pennsylvania by a sizable margin. Yes, the rural vote will go to McCain, but it will not be nearly enough to compensate for the margins Obama will rack up in the Philadelphia suburbs. In 2004 Kerry won PA 51%-48%, carrying 53% of the Philadelphia suburbs. Obama will perform even better than that tomorrow.

  4. Obama will win Virginia by four points by swamping McCain in northern VA, particularly Loudon County. He will be the first Democrat to win the state since Lyndon Johnson.

    VA nov 3.png

  5. McCain is going to win Indiana. Bush won by 21 points but its proximity to Illinois and the economy have made it a toss-up. However, the GOP base has come home in the final days.

  6. The two candidates will split the mega-battleground states of Ohio and Florida, with Obama taking the former and McCain the latter. Ohio has been hard hit by the economy and Bush only carried the state by two points in 2004. It will be close but should end up in the Obama column. Florida could really go either way but our sense is that McCain - with the help of Governor Crist and votes in the I-4 corridor - will pull out a very narrow victory.

  7. Missouri and North Carolina will be the closest states to call but both should end up in McCain's column. Both are tough calls with several polls showing it dead even, but our sense is to go with history. In North Carolina Dole will lose but McCain should win. Missouri will give McCain a narrow win and some redemption.
  8. MO nov 3.png

Finally, here is how we see the Senate and House races:

Democrats will increase their majority status in the Senate by 8 seats to 59. We are projecting that incumbent GOP incumbent senators Smith, Stevens, Coleman, Dole and Sununu will all lose. In the House we project a 31 seat gain for Democrats.



Excellent analysis but I disagree with you on the following points:
1) Obama will win Georgia. All of the recent polls have consistently pegged AA turnout at approximately 25%, while the reality on the ground is that 60% of the 2004 total voter turnout has already voted early, and AA turnout has held steady at 35%. The pollsters underestimated AA turnout in Georgia and Obama will pull an upset.
2) North Carolina, Indiana, Florida, and Missouri all are within the MOE. When races come down to the MOE, history shows that it is the GOTV that breaks the tie. Being that Obama's GOTV operation is far superior, he will inch by McCain in these states.
Obama projected electoral votes:
McCain projected electoral votes:

You heard it here first :)



Excellent analysis but I disagree with you on the following points:
1) Obama will win Georgia. All of the recent polls have consistently pegged AA turnout at approximately 25%, while the reality on the ground is that 60% of the 2004 total voter turnout has already voted early, and AA turnout has held steady at 35%. The pollsters underestimated AA turnout in Georgia and Obama will pull an upset.
2) North Carolina, Indiana, Florida, and Missouri all are within the MOE. When races come down to the MOE, history shows that it is the GOTV that breaks the tie. Being that Obama's GOTV operation is far superior, he will inch by McCain in these states.
Obama projected electoral votes:
McCain projected electoral votes:

You heard it here first :)



Thanks for the projections. I copied it off and will hold it dear as results come flooding in. I am very proud of u.s citizens.



oops sorry about the miscalculation:
Obama projected electoral votes:
McCain projected electoral votes:


Iowa City:

And thus, Missouri's 100 year reign as a presidential bellwether will end.


Professor T:

Good analysis. I agree with most of your conclusions, but I think Ohio will go Republican, and I think Pennsylvania will be closer than expected, with Obama winning there by 2 points or less.

I think there may be a southern surprise with both North Carolina and Georgia going to Obama. Both states have large black populations that have historically voted in low numbers. If that does change, it could be enough to give Obama both states. This result is more likely because both states have high-profile senate races with relatively unpopular GOP incumbants. There may well be some synergy at work in both states, a fact that would help Obama and the Democratic senate challengers.

Thanks for the good work. No matter what the outcome tomorrow, we will learn a lot about polling from this process. And we'll have a new President, too.


Hope Reborn:

311 seems to be a narrow EV considering the large victory in popular vote...



I find it irresponsible for a site such as this to trot out the tired Republican rhetoric that the race is tightening. There is nothing to suggest that either nationally or in most battleground states there has been anything more than short-term noise.

I may not be a polling expert, but the trend lines presented here run counter to the claim that McCain is making progress.



In re: Virginia, it's spelled "Loudoun" County, and I think you meant to write Fairfax County. Loudoun is most definitely GOP-leaning, although there have been some demographic shifts there. Fairfax County has HUGELY blue-shifted in the past decade, by comparison to Loudoun and Prince William. Fairfax, not Loudoun.


Hope Reborn:

538.com is showing now Mccain with a 1.9% chance of victory, with 311 coming in as the most likely scenario...

Seems that you guys are coming together too... I like 311, alot of room for Upside Surprise and Landslide talk if that's the expectation going into the night.



I hope you guys are right and I am being just way too cautious but I still have nightmares about 2000.

I don't know about Ohio. I am going with Obama at 291 EVs. All of the Kerry states, he brings back the Gore states of Iowa and New Mexico. Colorado puts him over the top, Nevada gives him a small cushion and, early the next morning, he takes Virginia.

Democrats pick up 8 Senate seats (to 57) and have to still deal with Lieberman.


Gary Kilbride:

I agree with all your numbered points, with the exception of McCain winning Florida. I can't see a Democrat winning nationally by 5-7 points and losing Florida in 2008. The state simply isn't that far red in relation to the country itself.

Unfortunately, I think you are too high in the estimates of Democratic gains in the House and Senate.



At least 2 of the 5 toss-ups that you paint red (NC, GA, FL, MO, MT) will turn blue on Nov 4. My gut feeling is that it will be NC and FL. Yet none of it really matters once PA and VA are called for Obama, at which point McCain has no winning sequence available to him. Obama with 364 EVs.



59 Senate seats is good news for those of us who think Lieberman should be stripped of his chairmanship and refused to caucus with Dems. He gambled and lost - and so did his constituents.


KMart Dad:

I don't disagree with this analysis except that I do think Obama will win FL. The early vote has given him a strong head start, and his ground game is superior.

So I would say Obama (52%) 338, McCain (45%) 200. 2% other.

Also, Chambliss & Martin will be forced into a runoff in GA. That should be interesting.



woah, hold the phones, this thing ain't over! i am an obama supporter, but its awfully brazen for you to declare victory like this.


Vicente Duque:

The most difficult to predict, because of being very tight :

Bets at Intrade to win $100 ( taken 1 hour ago ):

North Carolina (15EVs) .. Obama 61.0 McCain 40.5
Missouri (11EVs) ......... Obama 58.0, McCain 50.0

Toss Up States for McCain :

Indiana (11EVs) ........... Obama 40.0, McCain 62.2
Georgia (15EVs) .......... Obama 27.5, McCain 77.0

Obama will lose Georgia, no doubt. Most Probably he loses Indiana.

I just found the papers of the Brookings Institution that explain why the Southwest is becoming Democrat and Obama Friendly. Those documents are wonderful, deep, profound. The best I have read. But it is a mountain of documents and graphs.

I am very angry with myself for not having found them before.

I am trying to make synopsis of the knowledge of the Brookings Institution in my sites :


bets here ( updated everyday ) :

Vicente Duque



311 seems a bit wussy to me. It assumes the Mackster wins all of the tossups, which I don't buy - why would you assume that? I'll go out on a limb and give Obama ND (314) and either NC or GA (329). That's my conservative pick. If he rings the bell in FL or MO he goes into the 340's. But I'll stay conservative and make it 329.



I was out for my morning run and ran past three polling places here in Broward County, Florida. At just after 7:00 AM, the lines snaked outside and around the buildings. Borward county going heavy Democratic shouldn't be a surprise, but I've never seen turnout like this ever. I've also been pretty skeptical of the "historical" proportions turnout predictions.

Anyway, I'll have to file this experience away somewhere, because I did notice that people pulling into and out of the parking lots for the polling places were paying absolutely no attention whatsoever to other traffic or me.

One last note for Steve...I love the analysis, but can't help but obsess over your first graph. By using linear regression on past polling results (even with quadratic terms), you assume that past changes predict future changes. I don't buy it. If anything, trends tend back toward the middle after some period of time, not continue on toward infinity in the same direction. That's why the revised "green" line feels better qualitatively. Just a nit to an otherwise great analysis. I'll keep this handy while I'm watching tonight.



Georgia is going to be close. I think it could go either way. Weeks ago, the Georgia Obama HQ people told me that if they get within 3%, they thought their superior GOTV and enthusiasm will probably carry the day. Yesterday some polls showed them within 1%.

Here in DeKalb Co. Georgia the Obama HQ was a mob scene on Sunday & Monday. There were tables of phone-banking in the parking lot, cars parked on the grass, and carloads of canvassers who distributed thousands of door-hanging leaflets.

I do not know what the McCain HQ was like. They were supposedly going to do a 72-hour GOTV marathon. That's kind of obsolete in the era of voting before election day. I doubt they out-marathoned the Obama people.


Post a comment

Please be patient while your comment posts - sometimes it takes a minute or two. To check your comment, please wait 60 seconds and click your browser's refresh button. Note that comments with three or more hyperlinks will be held for approval.