Saturday morning update:
As with the previous vice presidential picks I've outlined below, Joe Biden looks likely to have little if any direct effect on vote preferences. In a new ABC News/Washington Post poll completed last night, 13 percent of registered voters said having Biden on the ticket would make them more likely to support Obama, while about as many, 10 percent, said it would make them less likely to do so. But most by far – 75 percent – said it would make no difference in their choice.
This kind of finding supports the axiom that the top of the ticket drives the vote. For a v.p. pick to be determinative, voters would have say, in effect, that Obama and McCain are so similar they can’t decide between them. Few get there.
Presumably the Democrats’ hope is that Biden will shore up Obama’s weakness in experience and foreign affairs. In our previous polling registered voters have divided evenly on whether or not Obama has enough experience to serve effectively as president – a lot of people to lose on a basic qualification for office.
At the same time, despite his long tenure in office, Biden’s not particularly well known. His favorability rating, the most basic measure of a public figure’s popularity, last was tested in a Gallup poll in April 2007. Twenty percent had a favorable opinion of him overall, 25 percent unfavorable, and 55 percent couldn’t say one way or the other.
If he isn’t well-known nationally, one other polling tidbit does shows the extent of Biden’s tenure: Gallup first polled on him 21 years ago, in April 1987. Seven percent of Democrats and Democratic-leaning independents in that poll said they knew something about him.
Obama, at the time, was a 25-year-old community organizer for a Chicago housing project.
We'll be reporting in detail on our new poll tomorrow morning.
Original post follows:
Recent history can offer bit of a diversion from the Obama veepwatch: In the past few campaigns, how has support for presidential candidates been impacted by their vice presidential picks?
The answer: essentially, not at all.
John Kerry had 48 percent support among registered voters in an ABC/Post poll completed June 20, 2004. He named John Edwards as his running mate July 6. In our next poll, completed July 25, Kerry had 46 percent support. Net change: -2.
In 2000, Al Gore had 37 percent support in a poll we finished the day before he picked Joe Lieberman and 40 percent support a few days later. Net change: +3.
The same year, George W. Bush had 44 percent support just before he named Dick Cheney as his running mate and 49 percent just after. Net change: +5.
And in 1996, Bob Dole’s support was 34 percent just before he picked Jack Kemp, and 32 percent just after. Net change: -2.
Polling, as I’m fond of saying, isn’t laser surgery. Small, generally insignificant changes tend not to amount to much, and it’s dicey at best to attribute them to specific events. It’s dicier still when they’re inconsistent. Average -2, +3, +5 and -2 and you get a grand 1. (These polls also had some slight changes in the opposing candidates' numbers, but unlike convention bounces, it's hard to make a plausible argument for evaluating v.p. picks on the margins.)
Data from 1988 and 1992 are conflated with the conventions, making it a tougher read. But we do have a clean look from Gallup polls in 1984, at the time of Walter Mondale's attempt at a transformational selection in Geraldine Ferraro. His support went from 37 percent a few weeks before naming Ferraro to 39 percent a few days after. Net change: +2.
All this fits in with a few other things we know about the impact of vice presidential nominees, including this, which I’ve reported previously, from our last poll: When we tested 17 individual issues for their importance to vote choices, the candidates’ vice presidential selections came in dead last.
These picks can tell us something about the presidential candidates, and as such can be revealing. But as far as directly influencing vote preferences… we don't see it.