I do not think it is going to far to say the election may be decided in the next two nights. As I have written here before, it is more common for races to be decided in the summer than in the fall. As of this moment, tracking poll information indicates that Obama has not received a bounce from either the first night of the Convention or from the Biden announcement. As I noted here previously, even Kerry got a bounce out of naming Edwards. In '96 Dole got a significant bounce from naming Kemp (and from the first two nights of the convention), and both Bush and Gore in 2000 saw some movement after naming their running mates (data on the flip).
The reason for the lack of a bounce thus far is obvious: everyone is waiting for Hillary and Bill. Right now the overwhelming media narrative of this convention is of a party that is badly split.
This may be counter-intuitive, and you can dismiss it as wishful thinking, but the very lack of a bounce to date to me suggests the opportunity at hand. What television hungers for is drama. When Bill and Hillary speak tonight and tomorrow the drama will be very real. And that is why what they say will be of critical importance.
A generation ago Ted Kennedy spoke to the Democratic Convention and delivered the one of the most moving speeches I have ever seen. The situation was not unlike the present (people keep drawing the wrong analogies to 1980) - the Democratic Party was so badly split in 1980 that Carter was actually below 50% among Democrats in one summer poll. Kennedy was hardly conciliatory to Carter in that speech. But it WAS effective, and it did help Carter temporarily put the Democratic Party back together.
You can argue that for Clinton the task is somewhat different tonight: surely she must put to rest the sense that her support for Obama is simply for show. On another level, though, her task is to show how HER causes would be destroyed by a McCain Presidency. When Kennedy spoke in 1980 he barely mentioned Carter. But he made to sure to attack Reagan.
If she makes this case effectively, and Bill re-enforces it tomorrow night, the election may be substantially decided. If they do not, we can expect a very close and hard fought election.
In past cycles there is usually at least one spot poll that tests the horse race the day after the nominee names his running mate. Opinion Research polled for CNN and found a dead even race, suggesting little movement for Obama. Rasmussen and Gallup also appear to have found little movement to Obama.
Some past snap polls:
In 2004 NBC took a poll the day Edwards was named showed Kerry up 8, a significant increase from polling taken the week prior.
Two subsequent polls showed less movement: Zogby and Time showed Kerry up 2 in polls taken the two days after Kerry was named.
In 2000 Gallup polled the day of Cheney's announcement and found Bush leading by 4, a two point increase from a week earlier (though Bush went from 45 to 49). Polling taken in the two days after Cheney's announcement showed significant movement in two polls: Bush lead by 9 in a Fox Poll and by 14 in a CNN poll, both taken 2 days after the announcement, though Newsweek and NBC found only a five point lead for Bush in the same period.
The snap poll for Lieberman's announcement suggested significant movement: Gallup found
Gore down only one. The previous Gallup poll taken a week earlier had Gore down 16. As noted in this post, however, the 2000 Gallup polling should be generally disregarded - it was terrible. In fact a week later Gallup found Bush leading by 13. Most of the other polling on Lieberman didn't get completed until 3 days after he was named.
In 1996 Dole named Kemp the day before the convention, and went from down 22 to down 12.
That race was completely different from this one, and I would not expect to find something similar.
Here is a summary of the bounces that candidates have received from naming their running mates: