By now most folk will have heard about the paper by a group of UMASS-Amherst economists that showed that a famous paper with claims about the relationship of growth and debt, which (together with the book aimed at the public) played a non-trivial role in providing the academic justification for austerity policies, by famous economists Reinhart and Rogoff were based on mistakes in the treatment of data. (Here is a response by Reinhardt & Rogoff, which is remarkably lacking in generosity to their critics and also lacking in humility about the magnitude of their error. It reads as if it was written by lawyers rather than academics-in-search-of-truth.)
George Gale pointed me to this news-story, which recounts how a PhD student, Thomas Herndon, "first started looking into Reinhart and Rogoff's work as part of an assignment for an econometrics course that involved replicating the data work behind a well-known study." Beyond the significance that politics may be guided by flawed science, here is a part of the story that transcends partisan politics; economics (like many other sciences) has terrible practices in sharing and replicating data (here's a link to recent scholarship on this.) My economist friend, David Levy, has been publishing about these issues for at least two decades (here's a link to a paper in Social Epistemology), so I am not optimistic that we will see a change in ethos in the field any time soon.
Update: Recall this earlier story about Jeffrey Sachs.