Over the last few elections the "Limbaugh effect" has gone from grassroots conspiracy theory to Obama-campaign talking point. On a campaign call today, John Kerry said that "If it was not for Republicans taking Democratic ballots, [Obama] would have won."
Did mischievous Republicans make the difference in Indiana?
There are a couple of ways to look at the math. Clinton won the state by about 18,000. Exit polls show that 10 percent of Democratic primary voters were Republicans, 54 percent of whom went for Clinton. Since about 1.3 million people voted total, that means about 68,000 of them were Republicans who voted for Clinton, compared to about 58,000 who voted for Obama. So if Republicans hadn’t been allowed to vote, as Ben Smith points out , Obama would have gained 10,000—not enough to catch up. Likewise, for "Limbaugh Democrats" to have made the difference, they would have to make up a quarter of the Republicans who voted for Hillary (18,000/68,000 = 0.26). That seems unlikely.
More compelling is an examination of Clinton voters who said they’d vote for McCain in the general. The Obama campaign points out that 16 percent of Democratic primary voters said they’d prefer McCain over Clinton in a general election matchup—and 41 percent of those voters actually voted for Clinton in the primary. That means about 7 percent of the Democratic electorate—about 83,000 voters—voted for Clinton in the primary and said they’d vote for McCain in the general even if Clinton were the nominee .