As a follow-up on last week’s post, here’s Paul Krugman on Nate Silver’s new FiveThirtyEight:
I’d argue that many of the critics are getting the problem wrong. It’s not the reliance on data; numbers can be good, and can even be revelatory. But data never tell a story on their own. They need to be viewed through the lens of some kind of model, and it’s very important to do your best to get a good model. And that usually means turning to experts in whatever field you’re addressing.
A tentative suggestion: It seems to me that when Krugman says “model”, philosophers of science might prefer to say “mechanism”. I don’t think Krugman wants you to use a particular way of representing reality (a model); he wants you to analyze data with reference to the actual entities and interactions in the system under study (its mechanism).
Raphael Scholl 
Could you elaborate on the notion of "mechanism" you are thinking of in this context? Do you have some kind of causality in mind (so that in certain cases some correlations the data might show point to causal links)?
Following much current work in the philosophy of science, I am thinking of mechanisms as entities and interactions organized such as to produce phenomena of interest. An important initial paper in this is by Machamer, Darden and Craver, here: http://www.jstor.org/stable/188611. And another important one is by Bechtel and Abrahamsen: https://mechanism.ucsd.edu/research/explanation.mechanisticalternative.pdf. These authors were thinking about biology, but it seems to me that the basic idea of a mechanism can be transferred to economics – although there is of course much room for debate about the details.