Yale has begun an online companion to the Yale Law Journal "The Pocket Part" at http://www.thepocketpart.org/. An attractive looking academic blog, giving an avenue for short pieces that readers can comment on. Yale's blurb says: "While full-length scholarly works remain at the center of legal debate, students, professors, and practicing lawyers are increasingly turning to the Internet to read about, and comment on, developments in the law. With this trend in mind, the editors of Volume 115 have launched The Pocket Part ( http://www.thepocketpart.org), an online companion to the Journal. With The Pocket Part, the Journal hopes to combine the strengths of both print and online media, broadening the readership of traditional legal scholarship while enriching the dialogue among the academy, bench, and bar." The first article is a digest of an article published in issue 1 of volume 115 of Yale Law Journal. Seems to be some self-citation to the Yale Law Journal within the online references, it will be interesting to see if that's coincidental or a continuing practice.
Ronen Perry has an interesting article, "The Relative Value of American Law Reviews: A Critical Appraisal of Ranking Methods" to be published around December 2005 in Va. J. L. & Tech. (volume 10:3) (a pre-publication copy was available on SSRN but had to be removed because of VJOLT's author copyright contract - it will return to SSRN after publication in VJOLT, and VJOLT's articles are freely available online). The key issue for Perry is that: "Mediocre journals with high paginations may rank higher than excellent journals with low paginations." His reflection is that citation-rate criteria (such as citations per x words) better reflect relative academic quality and inhibit editorial incentives to publish more articles of lesser quality.
I think that citation-rate (impact factor) is definitely better for some purposes. Though if you're an acquisitions librarian wanting to know what value a journal has for your library then you're really not too interested in citations/words (or citations/pages or citations/articles) because what you want to know is how many total citations did the publication garner, or even better a citations/cost analysis so that you can determine the cost the library pays for each citation. So for example Harvard Law Review receives about 832 citations per year (using the methodology of the Most-Cited survey) and Virginia Law Review receives about 431 citations. The price (institutional non-profit) is $95 for Harvard and $50 for Virginia, so the number of annual citations per annual dollar spent is 8.76 for Harvard and 8.62 for Virginia. Thus $95 is not an unreasonable price for Harvard Law Review, where a $200 price, for example, would be out of line with its utility as measured by citations to the journal. That said, impact factor is what authors should be looking at as, ideally, it reflects the average quality of the journal, to the extent that can be measured by citation count.
The cites purchased with each dollar for the top 10 ranked law reviews (based on total citation counts from the 1997-2004 data) are:
Harvard Law Review 8.76 (cites per $ cost)
Yale Law Journal 12.89
Columbia Law Rev. 11.0
Stanford Law Rev. 13.02
Michigan Law Rev. 10.22
Fordham Law Rev. 11.43
N.Y.U. Law Rev. 9.08
Georgetown L.J. 9.8
California Law Rev. 8.66
Virginia Law Rev. 8.62
On these figures Yale and Stanford are bargains and could reasonably raise their prices and stay competitive.
Eugene Garfield's concept of "impact factor", developed in the 1960s, currently has a profound influence on science journals and to a lesser extent social-science journals. There is a very interesting article by Richard Monastersky, "The Number That's Devouring Science" in the October 14, 2005 issue of The Chronicle of Higher Education, available at: http://chronicle.com/weekly/v52/i08/08a01201.htm that discusses the problems with impact factor, its potential for fuzzy math, and the complaints of editors and researchers. Nevertheless, as Monastersky acknowledges, there is value to a tool that differentiates the major from the minor journals. One area covered in the Chronicle article is the exclusions that ISI makes in its impact factor calculations. ISI does not include in the denominator of its calculation as to citations per article such entries as editorials and book reviews. The methodology of the Most-cited survey does include them, simply because manually differentiating countable from non-countable items would be too time-consuming. The consequence for anyone relying on the impact factor calculations in the Most-Cited survey is that impact factors are diluted by minor entries that may have few or no citations to them - so if a journal editor decides to write an editorial page, then, unless there's something remarkably citeable in that page, it will reduce the journal's impact factor. Impact factor in the Most-Cited survey may therefore unfairly compare journals that have different compositions of articles, notes, book reviews and other minor items.
For a journal editor then to get a higher score on ISI's impact factor there's a temptation to add in minor items that won't be counted in the denominator but might have an influence by raising the numerator (as citations to these minor items are counted). To get a higher score on the impact factor listing in the Most-Cited survey a journal editor should not publish minor items that will add a number to the denominator but will garner no citations, or small numbers of citations, for the numerator.
As a further note on the calculation of impact factor, ISI and the Most-Cited list calculate the values somewhat differently and impact factor values could not be compared between the rankings. ISI's calculation is based on (citations in survey year to articles published in journal X in the previous two years)/(number of articles published in journal X in the previous two years) - whereas the Most-Cited list uses a seven-year, rather than a two-year period.