I hadn't realized that the American Political Science Association's annual convention was going on right now in Washington until I saw Marc Ambinder's note about the panel he'll be joining to discuss the interaction between academics and journalists. Should be very worthwhile. There's not nearly as much interaction as there should be. Per Marc's post, the journalists will be asked to discuss an example of what they've learned from political science that has influenced their work.
I'm not a panelist, but my own example is this Atlantic cover story from a few years on why Karl Rove failed that drew heavily on what political scientists call realignment theory--the idea that a handful of elections in the nation's history mattered more than the others because they created lasting majorities for one party or the other. Think Roosevelt in 1932. Rove was trying to bring about a Republican realignment. He failed spectacularly, and to understand why it helps to know about realignment theory.
I believe I first encountered the idea, and Rove's connection to it, in an article by the historian Eric Rauchway in the Journal of the Gilded Age and Progressive Era (what, it's not on your nightstand?). This led me to the work of political scientists who study realignment theory, including Paul Allen Beck of Ohio State and especially David Mayhew of Yale, whose slim volume Electoral Realignments was particularly helpful (as was Mayhew himself). I think the piece is a good example of how academics can contribute a broader perspective that is often missing from political journalism. And I wish I knew a good way of keeping tabs on what's going on that might be pertinent to my beat, which I'm sad to say I don't.