Non-Mysteries

With the fundamentals so favorable to Barack Obama, why can't he crack open a bigger lead against John McCain? What's wrong with him? Does he need to change tactics? Or is it, as Andrew Gelman explains, that what the fundamentals predict is a modest victory with Obama getting about 53 percent of the vote. Right now, that's exactly what he's in line for and I expect it's what he'll get.

Note that close-ish elections are probably how the world should be. A lot of people are conservative Republicans. Barack Obama is not a conservative Republican. There's really no reason for those people to vote for Obama, no more how badly the economy may be doing. I think the only scenario in which Obama could really win in a landslide would be one in which Bob Barr starts eating away at McCain's base vote.

Comments

There are also a lot fewer swing voters these days that there were in the 60s, 70s, 80s, and even 90s. Remember all of that 50/50 Nation stuff in 2000?

Posted by joe from Lowell | August 1, 2008 2:49 PM

While this may be true, in light of the failures of the past 8 years, you'd hope that a solid percentage of these "conservative Republicans" would come to the realization that the policies that Republicans have pursued have proven to be outrageously bad for the country at large, and look at what John McCain is proposing to do with open eyes. Many of these people, I would hope, would agree that we need a change of direction.

I'm talking actual legitimate conservatives, not the wing-nuts that are currently holding Bush's approval rating at 25%.

Posted by dannity | August 1, 2008 2:54 PM

Since the Presidency is determined by electoral votes not popular votes, there are in fact quite a few ways for Obama to win a landslide.

Posted by redwards95 | August 1, 2008 2:58 PM

Wow, the two anti-incumbent outliers are 1952 and 1968 -- war elections.

Bad news for McCain.

Posted by MHD | August 1, 2008 3:08 PM

Right now, that's exactly what he's in line for and I expect it's what he'll get.

Obama could still win a fairly decisive EC victory, however, depending on third party activity. I don't expect Obama's victory to be in the same neighborhood as Reagan's in 1980, but the latter only got a smidgen over half the popular vote and yet destroyed Carter in the Electoral College. Personally I think a number of "reach" states for Obama will swing toward him in the final days, and he'll end up taking places like Indiana, Virginia, North Carolina and Montana.

Posted by Jasper | August 1, 2008 3:12 PM

A landslide is what you call Nixon's victory over McGovern. Or Reagan's victory over Mondale. What needs to happen in order for a Democratic candidate to carry 45 or more states?

Posted by MS | August 1, 2008 3:13 PM

If it looks like Obama is going to win, Barr will get lots more votes on election day than his polling will show in the weeks just prior. Republicans will stay in line if they think McCain has a chance but will bolt to Barr in protest and pissiness if they don't see their vote making a difference. I wouldn't be surprised with Obama barely clearing 50% but having an electoral vote blow-out.

Posted by Th | August 1, 2008 3:16 PM

"Close-ish" elections should be the norm for no other reason than what I like to call "the law of really fucking big numbers." That is, when you get into really big sample sizes, like the 100-150 million or so people who will vote in 2008, then it gets harder to create "big" statistical gaps between the 2 choices, and big wins come out looking small if you just look at the gap by %. So, for example, a 7% victory or so might look "close-ish," but in reality it would be the largest non-incumbent victory since George H.W. Bush's 8% win in 1988, and if we count Bush, the Vice-President at the time, as an incumbent it would be the largest non-incumbent win since Reagan's 9% victory in 1980.

Posted by Brien | August 1, 2008 3:16 PM

FTR, I don't really think that chart means anything for this election. Right now it looks like a chart that shows a decent variation with some exceptions. And right now, the chart contains two types of elections: those where there is an actual incumbent, and those where there is only an "incumbent party". Here is the chart with just the actual incumbents:

And here it is with only elections with no incumbents:

In other words, the strong correlation comes from elections where there is an incumbent candidate, while the elections without an incumbent show no particularly strong relationship. Of course there is insufficient data to make a really clear argument based on this data, which is a problem with all of these type of historical arguments.

Posted by Satya | August 1, 2008 3:37 PM

what the fundamentals predict is a modest victory with Obama getting about 53 percent of the vote

Two things. First, as MHD pointed out above, an unpopular war really should skew things more against McCain. Second, Obama is currently running behind the 6-point win you have for him - he's only up 4 at Pollster.com and only up 3 at RCP.

So I don't think that it's incorrect to say that, at this point, Obama isn't doing as well as the fundamentals predict.

Posted by Al | August 1, 2008 3:38 PM

FTR, I don't really think that chart means anything for this election. Right now it looks like a chart that shows a decent variation with some exceptions. And right now, the chart contains two types of elections: those where there is an actual incumbent, and those where there is only an "incumbent party". Here is the chart with just the actual incumbents:

And here it is with only elections with no incumbents:

In other words, the strong correlation comes from elections where there is an incumbent candidate, while the elections without an incumbent show no particularly strong relationship. Of course there is insufficient data to make a really clear argument based on this data, which is a problem with all of these type of historical arguments.

Posted by Satya | August 1, 2008 3:39 PM

I'm not so sure. It's comforting to rely on "fundamentals" because it suggests that the campaigning actually doesn't matter and so when Obama is campaigning poorly, as I think he is now, it's a relief to think it's all set in stone by conditions outside of the candidates' control.

The problem with this analysis is that this is an unprecedented election. We've never had a black candidate before. Who knows what it means? We'll see.

But in any case, the question remains: what is Obama's campaign plan? I hope he has one. But right now it seems to be: sit back, wait, be positive, and hope the fundamentals win the election for you. This seems like a poor strategy to me. McCain has opened the door to negative ads. Why doesn't Obama run some?

Posted by Z | August 1, 2008 3:40 PM

Let's not forget that Obama organized pretty impressively for the primaries and he is putting together his field operations right now. If he puts together a ground game as well as he put together caucus goers and targeted primaries, the polls today won't mean much in Nov. Of course, he could throw it all away with a lousy VP pick. I would hope he goes with someone who would carry on his programs for 8 years after him like Clinton planned with Gore. Forget ticket balancing, go for 16 years of consistent policies. Be audacious!

Posted by Th | August 1, 2008 3:54 PM

"Fundamentals" be damned. We've never had a black major party candidate for president before.

The Atwater/Rove fundamentals around October will notice, "Say, that guy is BLACK!" and all the paid shills will say stuff like "I don't know if the country is ready for a BLACK president." Or "Can we risk our soldiers not serving a BLACK president." Etc.

Posted by Jeffrey Davis | August 1, 2008 3:56 PM

Thank you, Matthew.

Here's more: 538 currently predicts an Obama win percentage of 68.3%, 50% - 47% in the popular vote 307-231 in the electoral college.

http://www.fivethirtyeight.com/

Please explain to Ross Douthat why we are not quaking in our boots. (You may also wish to explain covert Republican racism for him and advise him that Josh Marshall is not a "paranoid idiot.")

http://rossdouthat.theatlantic.com/archives/2008/07/dana_milbank_cryptoracist.php

Look, I understand that liberals are frustrated at Barack Obama's inability to pull away in the polls, despite all the favorable tailwinds he's enjoying and the fact that John McCain is running a staggeringly inept campaign. I'd be frustrated too! But that isn't a reason to make yourself sound like a paranoid idiot.

Posted by bob | August 1, 2008 4:09 PM

I find it amusing that the current Republican talking point on the poll numbers basically boils down to: "Our party is widely loathed and our candidate sucks, but Obama is only winning by a few percent. He should totally be kicking our asses!"

OK, point taken. But is this line of argument really going to convince a single person in America to vote for John McCain?

I can imagine the campaign ad...

Husband: "Gosh, honey, why is a despicable lout like John McCain only losing by 4%?"

Wife: "There must be something horribly wrong with this Obama guy. Maybe we should vote for the pathetic old sap just to be on the safe side."

VO: "I'm John McCain, and I (wheeze) approve this message.

Posted by LaFollette Progressive | August 1, 2008 4:13 PM

LaFollette, I agree. I think the same non-logic is at work in that celebrity ad: "Do you really want to vote for someone so popular that crowds of people chant his name?" No, perish the thought.

Posted by tomemos | August 1, 2008 4:35 PM

I would just like to encourage Matt to continue to predict a close victory for Obama. Predicting victory for McCain would be even better.

Posted by NoahB | August 1, 2008 4:43 PM

Here's more: 538 currently predicts an Obama win percentage of 68.3%, 50% - 47% in the popular vote 307-231 in the electoral college.

Who cares? "538" is a one-man-and-his-dog amateur pollster. Actually, not even a pollster. Just a guy with a blog.

Posted by Mixner | August 1, 2008 6:08 PM

Or is it, as Andrew Gelman explains, that what the fundamentals predict is a modest victory with Obama getting about 53 percent of the vote. Right now, that's exactly what he's in line for and I expect it's what he'll get.

So you've finally accepted, have you, that your "devastating game-changer" (Maliki's "endorsement" of Obama's "timetable") was neither devastating nor a game-changer? Or are you still clinging to the hope that it will somehow provide a massive boost for Obama? Any Day Now.

By the way, Gallup's latest daily tracking poll shows another 1-point drop for Obama. That's five days in a row Obama has lost ground. He's now tied with McCain, 44% to 44%.

Posted by Mixner | August 1, 2008 6:16 PM

There's really no reason for those people to vote for Obama, no more how badly the economy may be doing.

Certainly no war in Iraq, more soldiers dying in Afghanistan, sheer incomptenance at the White House.

Obama and McCain are practically twins!

Posted by Jeff | August 1, 2008 7:20 PM

Christ, another stupid post from Matt.

"Gee, there are a lot of conservatives - so that's why Obama can only win by three points."

Jesus, and The Atlantic hired this kid to write a blog?

If Matt can't figure out why a young, smart Democrat can't beat a senile old Republican like McCain with better than three points, get the fuck out of the blogging business - or go blog on some college campus Web site where you belong.

Posted by Richard Steven Hack | August 1, 2008 7:23 PM

Good grief...these polls mean what when they are only registering 85-90% of those called? When those 10-15% make up their mind (or choose to speak to a pollster) and it's still close, then start worrying.

Poll fascination before the candidates are seen together on the same stage is nonsense to the extreme.

Besides, a 3% win would have delighted both Clinton (twice) and Bush (twice). A one electoral point win still makes you Prez.

Posted by G Davis | August 2, 2008 12:29 AM

Re: First, as MHD pointed out above, an unpopular war really should skew things more against McCain.

The Vietnam War was pretty unpopular, and affected a lot more people due to the draft, and yet Humpghrey came very close in 1968-- in fact he might have taken the presidency if Wallace had not bled off a lot of the Southern Democrat vote.
Foreign policy plays second fiuddle in American elections.

Posted by JonF | August 2, 2008 7:24 AM

I'm pretty sure Obama will have a huge victory over McCain, but I'm absolutely certain that this victory will take place on November 4th, and not in August.

Posted by npr | August 2, 2008 8:20 AM

"Close-ish" elections are NOT how the world should be. Close-ish elections devalidate their winners. As always and still, all that anyone needs to know is shown on the county map of the 2000 election. The county map of the 2008 election will be identical, except for a bit of noise. It depicts an ungovernable country, which will remain ungovernable until it undergoes change of unprecedented kind and/or degree.

Posted by Frank Wilhoit | August 2, 2008 1:22 PM

Keep in mind that Obama won the Primary mostly due to his ground game and his incredibly disciplined campaign staff, particularly his delegate numbers crew. One thing we can count on is that Obama's people know the electoral numbers inside and out and that they will be boosting votes in key districts in order to capitalize on the electoral college. McCain's people will have to be on top of the numbers or the will be as stupified by Obama's sweep as was Hillary's campaign and supporters.

As during the primaries, I expect to see elction bloggers noting sitings of Obama vans everywhere getting out the vote. The campaign started its countrywide voter registration drive early and it will amplify significantly after the convention. Be prepared to see Obama volunteers canvassing the country registering voters and getting them to the polls in November.

It's the ground game stupid - and if Obama's got anything, he's got ground game.

Posted by Tane | August 3, 2008 12:21 PM

I don't care if he wins by 1 percentage point, as long as Sen. Obama wins. He can win and then show all those racists that he is the right man for the job. When the voter is satisfied, then the voters who didn't vote for him will start saying what a good decision they made.

Under no circumstances can that war monger win and take us into another war in Iran. He is not even personally a nice man witness all the things that his 'friends' and people he has known in his past have said about him and they weren't the lunatic fringe either. Just look at the MIA issue. Incredible insensitivity on his part.

Others have said this, but I'm going to repeat and rephrase it: This election is not an election for old men period. We have enough old men in congress right now who have done nothing but feather their own nests at the expense of the American taxpayer.

Become an activist - make a difference. Join the LWV in your state.

Lenore R.

Posted by lenore Rapalski rapalski | August 3, 2008 12:23 PM

I don't care if he wins by 1 percentage point, as long as Sen. Obama wins. He can win and then show all those racists that he is the right man for the job. When the voter is satisfied, then the voters who didn't vote for him will start saying what a good decision they made.

Under no circumstances can that war monger win and take us into another war in Iran. He is not even personally a nice man witness all the things that his 'friends' and people he has known in his past have said about him and they weren't the lunatic fringe either. Just look at the MIA issue. Incredible insensitivity on his part.

Others have said this, but I'm going to repeat and rephrase it: This election is not an election for old men period. We have enough old men in congress right now who have done nothing but feather their own nests at the expense of the American taxpayer.

Become an activist - make a difference. Join the LWV in your state.

Lenore R.

Posted by lenore Rapalski rapalski | August 3, 2008 12:23 PM

I agree with Matt that landslide elections are less likely to happen now. When people on both sides vote on abortion and things like gay rights, it makes it hard to switch sides, even if your candidate sucks. Many of these people might vote a third party, though. This year I think Barr could get way more votes than Nader got in 2000.

Posted by Lolis | August 8, 2008 1:36 PM

http://www.fudousanget.com

Posted by 不動産投資 | August 13, 2008 9:49 PM