My database is constructed from EconLit entries for 63 journals in the ten-year interval 1993-2003. The quality-weighting of journals is based on the Kalaitzidakis et al., EER 2002, index which reflects citation counts, adjusted for things like the annual number of pages and the age of a journal.
The scoring method is as follows. For every entry (an author + article), the score is the page share (pages*2/(1+n) where n is the number of authors) weighted by journal quality and normalized. Journal quality is the log of the Kalaitzidakis et al. index, which runs from 0 to 100. Suppose we have two articles of the same page length and with the same number of co-authors, one in AER (which has index value 100.00) and the other in Applied Economics (index value 2.00). With logs, the ratio is 2 to 0.3, so seven AE publications are approximately equivalent to one AER publication. It's somewhat arbitrary, but the original ratio of 50 to 1 seems rather too punitive. The quality-weighted page share is divided by 100. I like this normalization because a very high quality publication has approximately a value of 1 (e.g. two members of the same department publishing 35 pages in Econometrica together get 0.5 each). The department's total (which is shown in the rankings) has an interpretation: it's the equivalent of "great papers" produced over the ten year period.
There are approximately 65,000 entries to begin with, but thanks to two "exclusion rules" I work with less than 50,000 entries: (i) every article shorter than three pages goes (they're likely to be errata, corrigenda, comments, book reviews, etc. - which are a nightmare to remove explicitly), (ii) authors are excluded if they have not published since 1997 (in which case the affiliation is probably not current) or if they have published only one article in the entire period. The latter is partly for convenience (it's often difficult to fill in missing information for such people) and partly because they are either apparently inactive, so not really relevant for our purposes, or possibly graduate students whose affiliation is not permanent. In the database proper, affiliation is determined by the department cited in EconLit for the most recent publication in one of the 63 journals. Of course, there is backlog and some outdated matches result. This can seriously distort the rankings if a small department has recently gained or lost someone prolific.
All JEL codes are treated equally in assigning entries to subdisciplines and minor fields. For example, if an article has four descriptors that belong to three different fields as defined by this site, then the university to whom the author belongs gets a quarter of the entry's score in two fields and half the score in the other. Hence, although articles may have as few as one JEL code or as many as six, an entry's total weight across all subdiscipline rankings is always 1. Precise definitions of fields in terms of associated JEL codes are given at the bottom of each page and should be consulted.
Network Rankings. "Network" scores are the combined scores of all authors who either belong to a department or co-authored with someone in the department. Hence they reflect the quality of the wider research community a department interacts with. Frequency of collaboration is not captured - every co-author is counted once, even if there were multiple collaborations. (Of course, department members who collaborate are not double-counted.)
One of the advantages of this measure is that it effectively has more datapoints, so that I can report meaningful ranks for more institutions in each subdiscipline. Note that subdiscipline scores measure the total output of linked co-authors and department members in the field. Therefore, a department which has no output in a field may still have a nonzero network score, say if someone working in micro theory has a co-author who also writes in macroeconomics.
Average Productivity Rankings. This is a "size-adjusted" version of the regular rankings. To calculate the "average equivalent papers" for a department, I rank its authors by individual scores and compute the average for the top 15. If a department has more than 15 authors, say 20, the remainder (5) is not counted in the ranking; if it has less than 15 authors, say 10, I effectively add (5) dummy authors with zero scores. Hence the ranking punishes departments whose publishing faculty falls short of a minimal critical size (in this case, fifteen). But, unlike the regular ranking, it does not punish departments further for being smaller than others.
The table also reports the average for all authors in the department - but one should not read too much into this number. For departments with many publishing authors, it is automatically biased toward mediocrity, although an active faculty is clearly a good thing. One should keep in mind that "author count" is not the same thing as a "faculty size count." I do not have data about faculty size (and if I had, I would not find them interesting). The average productivity ranking is not given for subdisciplines because I believe that more expertise in a specialty is valuable regardless of how many authors embody it. Hence the regular subdiscipline rankings are most appropriate.
Included journals (in alphabetical order): American Economic Review / American Journal of Agricultural Economics / Applied Economics / Canadian Journal of Economics / Contemporary Economic Policy / Economic Inquiry / Economic Journal / Economic Letters / Economic Record / Economic Theory / Econometric Theory / Econometrica / Economica / European Economic Review / Explorations in Economic History / Games and Economic Behavior / International Economic Review / International Journal of Game Theory / International Journal of Industrial Organization / Journal of Applied Econometrics / Journal of Banking and Finance / Journal of Business Economics and Strategy / Journal of Comparative Economics / Journal of Development Economics / Journal of Economic Behavior and Organization / Journal of Economic Dynamics and Control / Journal of Economic History / Journal of Economic Literature / Journal of Economic Perspectives / Journal of Economic Theory / Journal of Econometrics / Journal of Environmental and Economic Management / Journal of Financial Economics / Journal of Financial and Quantitative Analysis / Journal of Human Resources / Journal of Industrial Economics / Journal of Institutional and Theoretical Economics / Journal of International Economics / Journal of Labor Economics / Journal of Law and Economics / Journal of Law, Economics, and Organization / Journal of Mathematical Economics / Journal of Monetary Economics / Journal of Political Economy / Journal of Population Economics / Journal of Public Economics / Journal of Risk and Uncertainty / Journal of Urban Economics / Land Economics / National Tax Journal / Oxford Bulletin of Economics and Statistics / Oxford Economic Papers / Public Choice / Quarterly Journal of Economics / RAND Journal of Economics / Review of Economics and Statistics / Review of Economic Studies / Scandinavian Journal of Economics / Social Choice and Welfare / Southern Economic Journal / Theory and Devision / World Bank Economic Review / World Development.