Princeton Election Consortium

Innovations in democracy since 2004

Outcome: Biden 306 EV (D+1.2% from toss-up), Senate 50 D (D+1.0%)
Nov 3 polls: Biden 342 EV (D+5.3%), Senate 50-55 D (D+3.9%), House control D+4.6%
Moneyball states: President AZ NE-2 NV, Senate MT ME AK, Legislatures KS TX NC

A hard look at reality, and what you should do

October 10th, 2008, 10:43pm by Sam Wang

(Update: In the comments I answer questions. Also see this analysis. -Sam)

(Update 2: A major bottom line of this post was to give to key races as detailed on my ActBlue page, or on the other side via the National Republican Senatorial Committee. Thanks to all who gave. -Sam)

I’ve noticed a lot of continued speculation about the Presidential race. You ask about the Bradley effect, voter purges, and other detailed topics. It’s all over the comments section, both here and on other sites.

As immersed as I am in the analysis, I am only now noticing that many of you are taking a little time to adapt to current conditions. This post is directed at those of you who are rooting for one side – Democrats and Republicans alike. It is time for you to take a good, hard look at what is going on. Whatever your personal preference may be, a Democratic sweep is coming. The storm is about to make landfall, and we know where. The question is what you should do about it.

Where the Presidential race stands. By the standards of Presidential elections since 1992, Barack Obama is far ahead. For most of this season he has been running about 50 EV ahead of where John Kerry ran at the same point in 2004, which ended in a near-tie. Currently the gap is even larger – it’s nearing Clinton v. Dole proportions. In the face of a down economy and abysmal approval ratings for the Bush Administration, a lead of this size by a Democrat is essentially insurmountable.

This is why John McCain’s tactics have become increasingly savage – it’s his last stand. It is why Obama has started to buy 30-minute blocks of time – he is shooting for a massive blowout. Conservative commentators are jumping ship, writing obituaries for the Republican Party or even coming out for Obama. The writing is on the wall. Every knowledgeable insider on either side knows it.

At a time like this, one impulse is to worry or grasp for straws, depending on who you are rooting for. You might like to speculate on the Bradley effect, in which polls overstate the support for the black candidate. This effect was never more than 2-3 percentage points in the first place, and signs of it disappeared over a decade ago. You might want to know if cell phone users are undersampled. Perhaps, but only by a little, and that’s a population that favors Obama by an even larger margin than the general population. You might want to know if pollsters’ likely voter models are off. This effect isn’t going to be more than a few points, and could well be zero. All of these potential errors are either negligible or suggest that Obama has more support than polls now state. In short, the wind is at Barack Obama’s back. I currently expect a final outcome of Obama 318-364 EV, McCain 174-220 EV.

The last normal game-shifting point in a national campign is the first debate, which worked in Obama’s favor. Of course, it is always possible to imagine an extreme scenario in which the outcome is different. Recent inflammatory words of McCain and Palin do increase the odds of a tragic event. But affecting the likelihood of such a freak occurrence is out of your reach.

Making your efforts pay off. An example of wasted effort at this point is making an additional contribution to either Presidential campaign. I realize that for some of you, this is a difficult proposition. If you are already committed to turning out the vote for your candidate, by all means do so. But if you still have time or money to spare, think about the following argument.

In general, any contribution you make to a strongly leading or trailing candidate makes little difference in the outcome. It’s like voting in Massachusetts or Utah: whether you do or don’t essentially makes no difference in the outcome. The same is true for campaign contributions. In the best of worlds, $100 to Obama-Biden or McCain-Palin would move the national win probability by an infinitesimal amount. Even 0.00001% would be an overstatement.

The place to make a difference is at the margins. Take the Georgia race, in which incumbent Republican Saxby Chambliss is defending his seat. In 2002, Chambliss won office by tarring Vietnam war hero and triple-amputee Max Cleland with an alleged sympathy for Osama bin Laden. Now Chambliss is fighting for his political life, and is in a dead heat with Democratic challenger Jim Martin. If you had the choice of driving voters to the polls in Georgia or in South Carolina, you’d be dead wrong to pick South Carolina. By the same token, a contribution in Georgia, but not South Carolina, might make a small difference in the outcome.

What your contribution buys. The outcome of the 2008 campaign determines the size of the working majorities in next year’s Congress. Next year, top priorities for any President and Congress will be the war in Iraq, the financial meltdown, health care, and global warming. It will be an unenviable and enormously difficult task. If Obama wins, as I expect he will, what he accomplishes will depend critically on how many votes he has in Congress. Whether you are a Democrat or a Republican, your point of leverage in this process is the Senate, where a minority of 41 can stop a bill from becoming a law.

What I suggest you do. For all these reasons, I have identified three Senate races as being among the most effective places you can put your effort or money. This is my advice to everyone, Democrats and Republicans. The Republican National Committee may be about to act along these lines. These races may become less competitive. Others may come into the picture. But right now they are the best targets, period.

Tonight I’m donating to my side – Merkley, Franken, and Martin – via my ActBlue page. Some of you might want to contribute to the other side via the National Republican Senatorial Committee. I assure you that in either case, the donation will have the maximum possible leverage in what happens in 2009.

Tags: 2008 Election

26 Comments so far ↓

  • Peter Formaini


    Excellent analysis and message.

    The race literally ended last week when McCain took a major political gamble and lost it in completely blowing his attempt to ‘handle’ the economic meltdown. And, of course, tonight’s release of the Palin Ethics Report is the final nail in the coffin – leading perhaps to further erosion in down-ticket races.

    Folks need to donate their hard-earned (and increasingly more precious) resources where they might make a difference! And your selections are right at the top of any objective list (especially the Georgia race).


  • Bruce Maples

    A good reason to donate to Bruce Lunsford! What sweeter win could we get than to finally take out Money-Bags Mitch, the Senate minority leader — who has led the way in blocking the progressive agenda in Congress. Here you go:

  • Sam Wang

    Bruce Maples, by my current definition, races that might become contenders if they move in the Democratic direction are Kentucky (Lunsford) and Mississippi (Musgrove). They are not there yet, so I would say that you are recommending the equivalent of an aggressive investment strategy. Don’t try that on the stock market! If those races get closer, I will certainly announce it.

  • Chris B.

    Excellent advice, accepted and followed. As you say, funds are tight these days, so each got a mere $10, but if this election has demonstrated nothing else, it’s that scads and scads of small donations can go a very long way.

  • Teri

    FANTASTIC post. Great reality check, from a source I really trust.

    I’m a Minnesotan and was thinking of doing some GOTV efforts for Obama. I think I’ll do it for Franken instead.

  • jd

    Good advice. I just used your Actblue page to make three small donations & I’ll do so each week until the election. But I’m also continuing my regular contribution to the Obama campaign. I want the ground game well-funded on election day!

  • Evans

    Interesting take. Good point about the general election, McCain needs to gain traction in what, 10 states? And what’s actually happening? Well, he’s now within the margin of error in Georgia! Georgia, you’ve got to be kidding me, in the words of Dickie V Obama’s a Diaper Dandy!

    As a NC native (and Hagan family friend), I’m curious about whether or not you consider the Hagan-Dole race worth supporting? A recent poll had Dole in the lead within the error, albeit a newer one may indicate that was an outlier.

  • Ginny Mayer

    The prospect of a McCain-Palin win is becoming increasingly terrifying as I see more and more demonstrations of their strategy of courting the lynch-mob vote. I am grateful for the reassurance that all the noise about the Bradley effect etc. is just that, and that Obama will be the next president barring some unforeseen tragedy.

  • Paul

    Yes, great analysis. I agree about 90%.

    I think you may be stating things just a bit too strongly in one respect, Sam: Obama’s victory still relies upon his supporters following through on their enthusiasm.

    So yes, I agree that donations are best directed to the knife-edge races — but it is still tremendously important that we Obama supporters volunteer for the Obama campaign, and that we VOTE.

    These polls, and thus this analysis, rely on enthusiasm translating into action. I would hate for anybody to read this and think that they can sit back and not chip in, or (heaven forbid) not even cast a ballot.

  • W. S.

    I agree that supporting close Senate races should be primary, but continuing to contribute to the Presidential campaign isn’t useless. The margin of electoral-college victory, and even more of the popular vote, is important in defining the national sense of mandate for the victor. Politicians take notice too–as when Democrats voted for Reagan’s tax cuts (unfortunately) because of his victory margin. With a major economic rescue and reform needed, a sense of mandate is essential.

  • desmoinesdem

    I agree, but don’t forget about the state legislative races! You get a massive bang for your buck there.

    Here are Five reasons to get involved in state legislative races.

  • jim

    I live in Lakewood [Denver suburb] Colorado, the neighborhood of 210 houses normally has a dominance on Republican yard signs, this year there are 3, one of those is placed behind a bush [irony] so that it can’t be seen from the street. I think there is going to be a huge stay-at-home factor for McCain. People may show up as supporters in the polls but they will not get out and vote. That does not mean that Obama people should take it easy, it would be a disaster to come this far and not bury the ball in the back of the net!

  • Glenn

    Wow. My lungs have filled with air imported from Princeton. I am grateful, Mr. Wang (I came to you via a link posted by a commentator at HuffPost).

  • NotStacy

    I live in Dallas. Dallas is a blue city, that should go Obama, within a candy-apple red state. I will be working for my Democratic candidate for the State House, Emil Reichstadt, and donating to Rick Noriega for Senate. Local efforts at all levels of the democratic ticket will help everyone on the ballot.
    I’d also add that the DNC is a good place to send a few bucks. When the RNC shifts from McCain to flood close Senate races with $$, the DNC won’t be able to compete.

  • Sam Wang

    Thank you all for your comments and your contributions. I’ll answer here.

    Voter fraud as a factor. This is highly unlikely. Unlike 2000 or 2004, the current race is not close. For fraud, challenges, and/or technological problems to affect the outcome, a race needs to come down to a few percentage points in one or a few states.

    It is possible to estimate how many votes must be invalidated, stolen, or suppressed to alter the outcome. Look at The Power Of Your Vote on the right, which lists key margins that would need to be overcome (except for NJ, which is just there as a joke). Using the turnout numbers for 2004 to convert these to votes, it would be necessary to shift or steal about 1.7 million votes. That’s not going to happen.

    Other Senate races. Other Senate races were left off because they are not currently on a knife edge. In North Carolina, Hagan (D) is ahead of Dole (R-i) by 3.0+/-1.1% (N=3 polls). In Kentucky, McConnell (R-i) is ahead of Lunsford by 9.0+/-2.6% N=3). Someone even asked about Noriega (D) vs. Cornyn in Texas. If this one gets close it’s a real canned-goods scenario. As significant as these races might be, none of them is a critical place where you have a lot of leverage.

    House and local races. Some of you pointed out the value of House races and state legistature races. I agree wholeheartedly. However, at these levels, polling data are quite sparse. I advise you to simply apply the principle I have outlined, which is to look for close races and give a good, hard push.

    Again, I want to emphasize my original point. I am addressing the type of question that organizations like the DCCC, the DSCC, and the NRSC most solve: how to identify the place where limited resources can be most efficiently deployed. Other races are important, sometimes extremely so. But if a race tilts strongly in one direction then it’s not the optimal place to apply effort.

  • Josh


    I’ve enjoyed your comments / analysis immensely. I just happened to pick up a New Yorker (October 13th issue) and Packer’s article, “The Hardest Vote” is a kind of Kohut effect analysis (Kohut’s article about ‘non-polled undetected racism’) basically suggesting that unlike those who want to appear to others (and perhaps for a moment or two, to themselves) as non-racist (a kind of Bradley effect, or in Jay Smooth’s words, “Tolerance Fatigue”), here are conscious unadulterated racists who are out of the polling loop, and lots of them.

    So, it would be interesting to run a racist coefficient array (i.e. different probable outcomes) demographically (i.e. so called blue-ish states) and see what, if any significant effect we could predict considering as well past more localized elections where there was a sharper colour line).



    Article reference:

    Quote from article:

    Obama has had particular trouble with the prized demographic group that once delivered the Presidency to Roosevelt and his successors. Anecdotally, and in polls, unusually large numbers of working-class voters seem to remain undecided or determined to sit the election out, as if they couldn’t bring themselves to vote Republican this year but couldn’t fathom taking a chance on Obama. Roger Catt, a retired farmer and warehouse worker, who lives in a small town near Eau Claire, Wisconsin, characterized the choice this way: “McCain is more of the same, and Obama is the end of life as we know it.”

    And another somewhat contrary quote:

    “At one point, he had doubted that Obama stood a chance in Glouster. “From Bob and Pete’s generation there are a lot of racists—not out-and-out, but I thought there was so much racism here that Obama’d never win.” Then he heard a man who freely used the “ ‘n’ word” declare his support for Obama: “That blew my theory out of the water.””

  • Frank


    Although I agree with you that Obama very probably has this wrapped up, I have two reservations, and these are related to each other.

    First, has him winning 224-158 in states in which he or McCain crossed the finish line, that is, exceeded 50% when undecided and other candidates are included in the total. He needs 29% of the remaining 156 electoral votes, and he is leading in 83%.

    Second, the Bradley-effect paper that you cited may be confusing periods and cohorts. In the 18 Senate or Governor elections involving a Black candidate in 10 election years from 1989 through 2006, it found that the average Bradley effect went from 5-6% to zero in nearly the first seven years (five election years). The conclusion was that the Bradley effect is a thing of the past: a product of “a specific period in political history.” However, maybe history will repeat itself with Presidential elections, especially considering their greater gravity. The good news is that your meta-margin (6.2%) exceeds the initial Bradley effect.

    Continued appreciation for your site.

  • Myra

    You said “Recent inflammatory words of McCain and Palin do increase the odds of a tragic event. But affecting the likelihood of such a freak occurrence is out of your reach.”

    I do think that people can make efforts to calm down the rhetoric. If that effort is made by our political and religious leaders, as well as by ordinary people (on blogs, for example), it might make a difference in the emotional atmosphere in this country.

    You also said that voter fraud is not likely to be a factor. Which is true, but for an individual voter who is deprived of his or her vote, it is a factor. Which is why we need to publicize the following: If voter fraud or vote suppression is suspected: call the Election Protection Hotline
    at 866-OUR-VOTE (1-866-6887-8683)
    Their Web site is
    They also need volunteers.

  • Eddie

    I agree very much with the idea that folks will have a much bigger political punch by contributing to certain congressional elections at this point than parts of the presidential one, and I really admire the work that you offer on this site, but I do still feel that your personal confidence of an Obama win is overestimated. Hopefully Obama will win though, and it’ll be no matter!

    Here is an interesting Salon article that I just read on the subject:

  • ftmoon

    Thank you very much for this wonderful site and the Easy ActBlue link. I enjoy your commentary and analysis very much.

  • John Dinwiddie

    You seem to be a sign of life. I’ll be watching
    for appearances from your website. What
    makes you so calm and collected it, and can
    it be sprayed from a crop duster?

    Anyhow, thanks for sounding sane and thoughtful.

  • M.T.

    Obama supporters, regardless of where they reside, need to vote and to encourage other Obama supporters to vote. This is true even in “safe” states such as California or Illinois where Obama is almost certain to win. The greater the popular vote, the greater of a mandate his election will be. Moreover, nothing can be taken for granted anywhere. There are the risks of election fraud, intimidation, malfunctions or just plain old error in counting votes. Even more importantly, Kerry won more popular votes in 2004 than any presidential candidate in history, except for one person — George W. Bush. The Republicans got more supporters to the voting booths on election day. Turn out was, and remains, critical.

  • Dave

    Sam, great site!

    I disagree with your strategy of focusing exclusively (or almost exclusively) on races that are currently extremely close. There is just too much time left, and in races such as Georgia or North Carolina, donations might create or prevent a scenario in which a candidate could take advantage of events on the ground to make a moderately close race into a very close one (thereby attracting much more money from the Senatorial committees and others).

    If events cause a three point swing in senate races in general (in either direction), donors who want to focus only on the closest of the close races will find that they have been giving to exactly the wrong campaigns. The races they gave to would not fit that definition on election day. Earlier money means more than later money, and it also makes sense to do some fishing in case there is a repeat of the VA senate race from last cycle.

    Of course, I don’t expect my $100 or whatever to make the difference, any more than I expect my vote to make the difference (when was the last senate race or house race with a margin of one vote?) But our voices can make a difference (and yours in particular, Sam), so I think it makes sense to expand the definition of how close a race is to warrant attention.

    Just my two cents. Again, great site!

  • William

    This is off-topic but I’d like to say congratulations to your colleague Paul Krugman for his shiny new Nobel Prize!

  • RichW

    Oregonians should also remeber that they get a state tax credit of up to $50 for political contributions. So if you haven’t made any contributions to date, I suggest you send $50 to the Merkley campaign. If you pay Oregon taxes, you will get every penny bak when you file your return.

    “That One is The One” ’08