Dan Drezner has a very pessimistic item on the lobbying power of political scientists. It's good, and I agree with most of it. I do think, however, that political science does have one advantage that virtually no other interest group shares: political scientists just happen to have tons of personal connections with Members of Congress and Hill staff.
The less important factor: many political scientists get to know politicians and their staffs in the course of doing their jobs -- we study them, we invite them to speak to our classes, some of them use our expertise, and more. Most of these interactions are casual and meaningless, but some of them turn into longer-term relationships.
The more important factor: dozens and dozens of political scientists have actually worked on Capitol Hill or on congressional campaigns or both. Some of them as Congressional Fellows, but many of them prior to going to grad school, or (in the case of electioneering) during or after grad school. That translates into dozens and dozens of personal relationships, no?
How many of them are there, and how strong are their relationships with their former bosses and former co-workers? I have no idea! Obviously having the occasional political scientist actually in Congress isn't enough to prevent this sort of thing.
I guess I should note one more thing: there are plenty of Congressional staffers (and campaign professionals) who have at least political science MAs, and a fair number of people who finished or almost finished doctorates. Now, some of them are hostile to political science, no doubt. But surely some of them are disposed to be friendly to political science.
Anyway, all of this should translate into access -- or potential access, at least -- allowing political science to fight well above its weight class. But it's only apt to be successful if it's organized, and as far as I know it really isn't. Getting that organized and effective would seem to me to be the most promising avenue for advocating for the profession, during a time when it's really needed.