Why doesn’t the median voter truly rule American Elections, as the classic Downsian model predicts? While uncertainty, intensity, abstentions, alienation, and other factors certainly play a part in the failure of American elections to converge on the median voter, less noticed (but perhaps more important than all other factors combined) is that the issue dimension is orthogonal to a second determinant of elections: valence. Valence is just a fancy word for “likeability.” Candidates well-liked by voters (on a host of non-political factors like appearance, dress, personality, character, charm, etc.) can ‘afford’ to diverge from the median positions because voters will discount those heresies in favor of their favorable judgement of the candidate’s personality.
Obama’s best asset this fall is likely to be his valence advantage over Romney. As this Gallup poll demonstrates, Obama has a significant edge on the valence dimension over Romney going into the election. While we might expect this to close as Romney consolidates his base and moves into the general election phase of his campaign, it is likely to remain a significant Obama advantage over Romney…and indeed could be the key to his relection.
With that in mind, the recent article by Robert Costa on National Review Online floating the possibility of a Mike Huckabee VP nod is particularly interesting. While other VP choices provide swing state, Tea party cred, block vote, and other traditional advantages in considering a VP selection, few have the potential to match Huckabee in sheer likability. His friendly affect and demeanor and good sense of humor has been particularly useful in his radio and TV gigs and was a big factor (in addition to his evangelical cred) in his solid run in the 2008 presidential primaries. The article touches on this asset, but really it is Huckabee’s best case as a VP choice. He has the potential to shore up Romney’s biggest weakness versus Obama in 2012…at least to the extent a VP candidate can have any effect.