New York City mayor Michael Bloomberg's decision to change his voter registration to independent has set off a predictable frenzy of speculation about an independent/third party presidential run. Don't believe the hype.
First, can Bloomberg win? Pollster John Zogby claims "his chances are promising" and that Bloomberg "could win it" while Andrew Sullivan warns of a "viable third party challenge." But as I've argued many times before on this blog, the factors preventing a successful third party bid -- which include strategic voting for one of the top two candidates, party loyalty, ballot access, and the Electoral College -- are essentially insurmountable. There's a reason the two major parties have been entrenched since 1860. To have any hope of dislodging them, a third party would need a signature issue that cuts across the axis of major party competition. For Abraham Lincoln and the Republicans, that issue was slavery. (Since then, the parties have successfully adapted and realigned around cross-cutting issues like civil rights and abortion.) No comparable issue exists today for Bloomberg.
In addition, Bloomberg is not exactly an ideal third party candidate. Polls show his appeal is limited. In addition, as Matthew Yglesias notes, as a Jewish ex-Democrat from New York, he's especially poorly positioned to pick up working class/Southern Democrats who might be disaffected with Hillary Clinton, though he could pick up some moderate Republicans who don't like a Thompson-style GOP nominee.
For all of these reasons, Bloomberg is unlikely to exceed 10 percent or so nationally if he does run. But will he realize that he can't win? No doubt many people are currently telling him how great he is and that he should enter the race (including consultants hoping to make millions in ad commissions). Ultimately, however, I think he will pull a Colin Powell -- bask in the adulation, engage in some phony posturing about major parties not talking about "the real issues," and flirt with running, but back away in the end.
However, if Bloomberg does run, he will affect the outcome of the race. Who will he hurt more? Zogby says "my polling shows he would likely take more votes from the Democrat than the Republican." On TNR Online, John Judis argues that Bloomberg would draw from "Blue-State Independents" like John Anderson in 1980, who "provided the margin of victory for Reagan in eleven states." Writing on Tapped, Paul Starr agrees and (bizarrely) suggests an Obama-Bloomberg ticket. And, after considering a number of possible factors, The Atlantic's Marc Ambinder seems to also come down on the side of Bloomberg hurting Democrats, noting that he may appeal to voters on traditionally Democratic issues, that the independents who are most likely to support Bloomberg currently favor Democrats by huge margins, and that Bloomberg might steal Democratic voters in California or the South.
However, an April Rasmussen poll shows Bloomberg pulling almost entirely from McCain and Giuliani in head-to-head matchups with Hillary Clinton, and recent state-by-state Survey USA polling indicates that he currently "siphons enough Republican votes to flip red states Iowa, Missouri, Ohio, and New Mexico blue" (PDF). Finally, David Frum guesses that, after sinking in the polls, a desperate Bloomberg would start running heavily negative ads against "the candidate he dislikes more," who will "almost certainly be the Republican."
Who's right? On balance, Bloomberg seems more likely to peel off Democratic-leaning moderates and independents than any other voters, as Zogby and Ambinder argue. The poll results I cite above are unlikely to hold up due to the logic of strategic voting -- will Republicans vote for Bloomberg over the GOP nominee if it means they are helping to put Clinton, Obama, or Edwards in office? (Note: It's hard to interpret how this will play out. The Anderson 1980 and Perot 1992 precedents are particularly difficult to apply because both of them picked up disaffected ex-supporters of an incumbent running for re-election. The last major third-party challenge in an open-seat presidential race was George Wallace in 1968.)
One final note -- you should use the Bloomberg hype as an opportunity to evaluate political pundits on their understanding of politics and their willingness to hype implausible scenarios. Yglesias, for instance, had the best reaction to the Bloomberg news:
I stand by my one question licensing quiz for political pundits:
Is today the right moment to get excited about a third-party presidential run? If you answer "yes," you need to find some other line of work.
Update 6/22 10:06 AM: Yglesias later wrote something similar to my analysis above about the 1980 and 1992 precedents:
Jimmy Carter was an incumbent president widely judged to have failed in office (recall Ted Kennedy's very strong showing against him in the 1980 primaries). This created a large pool of people ideologically inclined to pick Carter over Reagan, but who also really didn't want to vote for Carter. Much the same could be said about people who defected from George H.W. Bush to Ross Perot; in both cases actual incumbent performance drove disaffection and a proclivity for third party voting.
It seems to me that this kind of dynamic is pretty uniquely associated with incumbency, and probably doesn't apply at all to the 2008 election and certainly doesn't apply to the Democrats in 200.