Just how bad things are for Bush, or, Evan Bayh for Vice President
A snapshot of Indiana elections at the national level, 1988-Present
In 1988, with no 3rd party candidate running, George H.W. Bush defeated Michael Dukakis by 20%, 60-40
In 1992, with a 3rd party candidate running who stole more votes from the Republicans than the Democrats (exit polls from 2000 indicate Perot voters went for Bush by a 65-35 margin), Bush won Indiana 42-36. Assuming Perot's 22% vote was split 65-35 between Democrats and Republicans , Clinton would have lost by 12%, 56-44.
In 1996, with Perot running again, Dole beat Clintion 47-41. Without Perot in the race, Dole likely would have beaten the wildly popular Clinton by 8%, 54-46.
In 2000, George W. Bush trounced Gore in Indiana, by 16% 57-41, with Nader taking 2% of the vote.
In 2004, a SurveyUSA poll of registered voters puts Bush ahead by only 6%, 51-45, with 5% undecided. In fairness, those undecided voters are likely former Bush voters who are sitting on the fence, and thus are more likely to vote for Bush than Kerry (or even Edwards). So let's call it an 6-10% lead, that Bush would win somewhere between 53-47 and 55-45.
Now, granted, we are probably in the "trough" of Bush's popularity at the moment, but this underscores the point: Indiana, which was Bush's 12th easiest win and his only Midwestern win, is at the moment close enough that swaying the minds of 3% of Indiana voters would turn the state towards the Democrats for the first time in ages. In an of itself, that fact is astounding.
It might be tempting for the eventual nominee to nominate Evan Bayh, the former governor of Indiana, who is currently the chairmain of the Clintonian Democratic Leadership Coalition. However, it would be a mistake to choose a running mate based on the polls at the "trough" of Bush's popularity. More likely, the eventual nominee should wait until things calm down in the late summer primaries before figuring out whether to pick from the crowd of whispered names from VP.