In response to Jonah’s request . . .
For starters, I take the Daily Kos/Research 2000 polls with a grain of salt. (Or a shaker. Or a mine.) Research 2000 is a professional firm, and I’m sure they do their best, but they usually tend to depict a best-case or near-best-case scenario for the Democrats, at least recently.
Last year, their last poll in New Jersey showed Chris Christie winning by 1 percentage point; he won by 4 percentage points on Election Day. Their last poll in Virginia showed Bob McDonnell winning by 10 percentage points; he won by 18 on Election Day. Their last poll in Massachusetts showed a tie; Scott Brown won by 5 percentage points on Election Day. I’ll let those who study polls in greater depth hash out why this is so, but I generally presume that the actual results will not look quite as good for Democrats as the final R2000 poll does.
In January, Rasmussen had Bayh only up by 3 percentage points over Hostettler and trailing in a hypothetical matchup with Pence. If somebody wants to argue that Rasmussen’s robo-pollsters skew right, it’s a free country, but he’s done pretty well in capturing the sense of momentum for GOP candidates since the tail end of last year.
On Election Day 2008, when Obama was carrying the state, Indiana’s electorate split 36 percent Democrat, 41 percent Republican, 24 percent independent. Obviously, Democratic turnout will probably be worse in 2010 than 2008; and independents have turned against Democrats by a wide margin in Virginia, New Jersey, and Massachusetts. There’s no reason to think that the pattern wouldn’t continue with Indiana.
What’s more, 13 percent of Indiana’s Republicans voted for Obama; no way that Bayh would score that well with the national mood as it is.
In short, Bayh had a tough, though not impossible, road ahead in his reelection bid. Judging from his comments, he has determined that even with his pile of cash and the advantages of incumbency, it wasn’t worth it.