The conventional wisdom among the cognoscenti seems to be that a vice presidential candidate doesn’t normally effect the outcome of an election. I’m not sure how we can know this for sure. It’s not like we can run elections in alternate universes with different vice presidential candidates in order to determine that there were any appreciable differences in electoral outcomes.
I wouldn’t expect a vice presidential pick to gain or lose a candidate a large number of votes but I wouldn’t expect the effect to be completely negligible either. I’ll concede that the margin of victory in most states in most elections is usually large enough to make whatever effect the vice presidential candidate has moot, but in recent elections we seem some razor thin margins. In 2000, five states were decided by margins of 0.5% or less. In 2004, three states were decided by margins this small. This year seems to be shaping up to be very close as well. Who’s to say, if Obama ends up squeaking by in a state like Florida, that it wasn’t Biden who pushed him over the top?