I would add to Mary's remarks that, though the UK was mentioned in an earlier posting, it does not operate on quantitative analysis of publications, but on the basis of assessing 4 named publications per person submitted. The panel could (but probably would not) decide that an unpublished discussion paper was so good that it counted for more than a paper in Econometrica. Panels will, no doubt, be influenced by rankings in their evaluations but, as I understand it, there are no formal rules for this. I believe this is different from the system in some other countries. Roger Backhouse