First, I really appreciate your bringing this to pen-l; that type of outreach is a big step in the right direction. I think I speak for many others when I say that, overall, I have little reason to regard myself as a "citizen" of the economics profession -- and this exclusion is NOT my doing. This inquiry is a welcome exception. As for specific issues: 1. I am interested in the issue of qualitative case studies. They have completely disappeared from the mainstream econ journals; you find them now only in applied journals -- labor, IO, etc. Yet serious attention to empirical work ought to involve comparing results from a variety of methodologies in order to overcome the shortcomings of each. So why the utter denigration of in-depth case research? What improvements in case methodology have their been while economics was sleeping? What models can be appropriated for high quality case studies in economics? 2. I think there needs to be a serious look at publication bias. I *know* this is a big problem in the area I have worked in most intensively (wage-risk analysis), and I hear furtive conversation about it every year at the meetings. How extensive is it? What types of econometric work are most prone to it? Is there any set of tools or conventions that can counteract it? What progress is being made on making all the relevant data -- not just raw data, but operations on the data, category construction, truncations and exclusions, unpublished runs -- available for replication? We are kidding ourselves if we think that the reason there are fewer research scandals in economics than, say, medicine is that economists are more honest and scrupulous. In fact, there currently isn't any way for anyone to find out. In any event, for every case of falsification there are dozens (or thousands) in which the "right" specifications are selected for submission on the basis of their publishability. Peter John DiNardo wrote: > > Hi. > > The Journal of Economic Perspectives (from the American Economic > Association) is considering a symposium on topic in econometrics. Part of > my job is to get input from a subset of *non--econometricians* on topics > that they might actively choose to read about if published in the JEP. > That is, this is not intended to be interesting to *econometricians* -- it > is intended to be interesting and accessible to JEP readers (a tiny subset > of which are econometricians). I would welcome any suggestions > at [EMAIL PROTECTED] > > John DiNardo
