The game theory of Arlen Specter

Will writes:

My quick take is that this sucks, because the more choke points in the policymaking process the better. That said, it probably doesn’t change all that much unless Senate Dems can muster reliable intraparty unanimity. A few things that wouldn’t have passed will, and those could be an important few things, but most final votes won’t be different. The one way this hurts the Dems is that it makes a narrative of GOP obstruction less plausible, and if various things go south by the mid-terms, the Republicans can more plausibly say that all of it’s the other guy’s fault.

If the guy is willing to switch parties, he was already in the first place willing to switch policies (if indeed he needed to change his mind at all).  He's suddenly lost of a lot of bargaining power (he had to hold off Pat Toomey, who presumably would have beaten him in the primary) and some of that power has been redistributed to the most conservative Democrats in the coalition.  That could be an improvement.

Note also that Democratic Senators may find it harder to oppose Obama once a policy initiative is announced, so they may work harder behind the scenes, and well in advance, to shape legislation in their preferred directions or simply just kill it off.  In contrast, a Republican veto-voice will be more reactive ex post.

On the marketing side, maybe now the Republicans, being denied the filibuster, will have to come up with some ideas that are actually appealing to voters outside their core constituencies. 

Addendum: Here is Matt's analysis.  And here is the academic evidence that voting behavior changes, following a party switch.