At the end of a long day, I was browsing the internet and came across this, which I thought was interesting and shared on facebook. Generated a bit of discussion with a friend of mine doing an extremely complicated physics PhD, who argued that the study of economics was compromised by ideological bias and dismissed the idea that it was scientific. Obviously foolish to argue about what science is with a scientist, but I gave it my best shot, which might just be interesting enough to put here as well (also to make up for the radio silence...)
I would hope that most economists subscribe to the idea that their analysis has to make reference to evidence and that it should interpret the evidence in a plausible way. Bias inevitably creeps in terms of the evidence you select and the analysis you choose to apply -- but you can still evaluate whether those two things are done properly, coherently. More generally, since human behaviour in aggregate is more variable: we can do more things more quickly than, say, trees (and we have more interest in and data about ourselves), isolating stable cause and effect relationships is inevitably more difficult. But that's why it's a social science, not a natural one. I would hope the methodologies or at least the mindset in which questions are approached make them similar enough to justify using the word 'science' for both.