Much has been made of how many Clinton voters would vote for Obama in the general election and vice versa.  But there's another question that begs to be asked:  "How many of those who voted for a particular candidate did so with no intention of actually voting for that candidate in the general election?"  In Indiana, 10% of the voters identified themselves as Republicans and those broke for Hillary Clinton 53-47.  In North Carolina the number of self-identified Republicans was only 5%, but they broke even more strongly for Clinton: 61-32.  But how many of these Republicans would actually vote Democratic in November?  How many independents would vote Democratic in November?  How much mischief was caused by people voting in the Democratic primary who had no intention of voting for that candidate in November?  In both states' exit polls, voters were asked who they would vote for in a two-way race between Clinton and McCain.  They were also asked in a separate question who they would vote for in a two-way race between Obama and McCain.  Regardless of who the Democratic nominee was, between 23 and 26% of the respondants said they would vote for John McCain or not vote at all.  It is the breakdown of these voters, when analyzed, that is quite illuminating.

In Indiana, 13.8% of those who voted for Hillary Clinton said they would either vote for John McCain or not vote at all in November.  Let that sink in for a second.  Almost one in seven who voted for her in the primary said they would not vote for her in the general election.  For Obama the number was considerably smaller: only 4.8% of those who voted for him said they would not vote for him in November.  Who would vote for a candidate in the primary when they had no intention of voting for him in the general election?  They can only be Republican crossover votes, but more importantly, crossover votes who had no intention of voting Democratic in the fall.  As a group, they broke 75%-25% for Hillary Clinton in Indiana.  

In North Carolina the numbers are even more extreme.  More than one in six (17.2%) of those who voted for Hillary Clinton said they would not vote for her in the general election, while only 3.7% of Obama voters said they would not vote for him in November.  Clinton captured 78% of these voters in North Carolina, which is a similar percentage to what was observed in Indiana.  

It's impossible to tell whether this group is part of Operation Chaos, voting for a candidate who they deem more beatable in November or just registering a second choice. Regardless, since these people had no intention of voting Democratic in the fall, it's reasonable when addressing the "electability" argument to ask what the totals would be without them.  Given that these voters amount to 10% of the votes cast, it's not surprising to find that their impact was considerable.  Counting only voters who said they would vote for their candidate in November if he/she were the nominee, Obama would have won Indiana 52%-48% and his margin in North Carolina would have expanded to 61%-39%.

Have at it.